Clinical Quality Improvement Metrics (CQIM)

Slide 1

Welcome to this short video demonstration on understanding the policy behind the Clinical Quality Improvement Metrics, also known as CQIMs. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, referred to as MSDS.

Slide 2

The better births report, released in 2016, recommended the development of a nationally agreed set of indicators to help local maternity and neonatal systems to track, benchmark and improve the quality of services through regular review by multi-professional teams.

Since then, trusts’ performance across a small number of Clinical Quality Improvement Metrics and national maternity indicators is routinely benchmarked through the national maternity dashboard.

Maternity services should review the available data to draw out themes and trends and identify areas of concern including consideration of the impact of inequalities. Analysis of meaningful data can be helped by ensuring high-quality submissions to the Maternity Services Data set by trusts.

Supporting links: NHS England (2016): Better Births: Improving outcomes of maternity services in England – A Five Year Forward View for maternity care

Slide 3

To support this policy of clinical quality improvement, NHS Digital developed 12 key metrics, rates.

CQIMApgar: This metric measures the proportion (rate per thousand) of full term, live or still born, singleton babies who have been recorded with an Apgar score (at 5 minutes) between 0 and 6.

This metric only includes babies who have been recorded with a valid Apgar score at 5 minutes.

CQIMBreasfeeding: This metric measures the proportion (percentage rate) of babies who have been recorded with a first feed of donor or maternal breast milk.

This metric only includes babies who have been recorded with a valid first feed status.

CQIMBreastfeeding6to8Weeks: This metric measures the proportion (percentage rate) of 9 week old babies who have been recorded as fully or partially breastfed between 6 and 8 weeks.

This metric also includes babies who have been recorded only in the Community Serviced Data Set (CSDS).

CQIMPPH: This metric measured the proportion (rate per thousand) of women that had been recorded with post-partum haemorrhage of 1,500ml or more.

This metric included all women who gave birth to a live or still born baby.

Slide 4

CQIMPreterm: This measures the proportion (rata per thousand) of women that had a live born, singleton baby which was delivered between 22 and 37 weeks gestation.

This metric only includes women who delivered a baby recorded between 22 and 45 weeks gestation at birth.

CQIMRobson01: This metric measures the proportion (percentage rate) of Robson group 1 women who have been recorded with a delivery method of caesarean section.

This metric includes all Robson group 1 women.

CQIMRobson02: This metric measures the proportion (percentage rate) of Robson group 2 women who have been recorded with a delivery method of caesarean section.

This metric includes all Robson group 2 women.

CQIMRobson05: This metric measures the proportion (percentage rate) of Robson group 5 women who have been recorded with a delivery method of caesarean section.

This metric includes all Robson group 5 women.

Slide 5

CQIMSmokingBooking: This metric measures the proportion (percentage rate) of women that had information recorded at their booking appointment which identified them as a current smoker.

This metric only included women whose smoking status was recorded at the time of their delivery.

CQIMSmokingDelivery: This metric measures the proportion (percentage rate) of women that had information recorded at the time of their delivery which identified them as a current smoker.

This metric only includes women whose smoking status was recorded at the time of their delivery.

CQIMTears: This metric measures the proportion (rate per thousand) of women that had a full term, live or still born, singleton, vaginally delivered baby who were recorded with a 3rd or 4th degree tear during delivery.

This metric does not include women who gave birth by a breech delivery.

CQIMVBAC: This metric measures the proportion (percentage rate) of women that had delivered their first baby by caesarean section, who went on to have a second, full term, baby delivered vaginally.

This metric only includes women whose delivery was recorded with a valid delivery status.

Slide 6

In support of the 12 CQIM rates, there are another 40 data quality (DQ) metrics which ensure that the CQIM data quality is of good enough quality to be published. These DQ metrics ensure that each key CQIM rate is only calculated when the relevant data from the submitting organisation is of sufficient quality and completeness.

The CQIM DQ metrics assess the completeness of the data submitted to MSDS for the following areas of maternity care:

• The number of births or the number of bookings.
• The mother’s smoking status at delivery or booking.
• The mother’s history of live and still births.
• The mother’s delivery complications (post-partum haemorrhages, critical incidents, tears).
• The mother’s method of labour onset.
• The fetus’s presentation at onset of delivery.
• The baby’s gestation at delivery.
• The baby’s method of delivery.
• The baby’s Apgar score at 5 minutes.
• The baby’s first feed type.

Slide 7

NHS England publishes these measures as part of the Maternity Services Monthly Statistics. The publication series can be found at the link (here).

There are additional videos in this series which offer guidance on navigating the Maternity Services website and finding the published data (here) and there are also videos which give further details on how each of the CQIM rates and accompanying DQ metrics are built (here) as well as for other measures derived from the Maternity Services Dataset (here).

Slide 8

In the published data you will find the 12 key CQIM rates referred to by the following names:

CQIMPreterm, CQIMPPH, CQIMRobson01, CQIMRobson02, CQIMRobson05, CQIMSmokingBooking, CQIMSmokingDelivery, CQIMTears, and CQIMVBAC. These are CQIM rates which are all related to the woman, at either her pregnancy booking or delivery.

Slide 9

While CQIMApgar, CQIMBreastfeeding and CQIMBreastfeeding6to8weeks are all related to the baby.

The published data will show the numerator, denominator and rate (or rate per thousand) of each CQIM metric, where the DQ metrics indicate the data is of sufficient quality to do so. It will also show the result of each metric – as either a ‘Pass’ or ‘Fail’.

Slide 10

The published data will also include the 40 CQIM data quality metrics for each of the provider trusts. These will be shown with a numerator, denominator, rate and a result of ‘Pass’ or ‘Fail’.

The DQ metrics are numbered – CQIMDQ02, CQIMDQ03, etc. and the definition of each DQ metric, and which CQIM rates they are applied to, can be found in accompanying metadata.

Slide 11

This brings us to the end of the video.

Thank you for watching this video demonstration on the policy underlying the Clinical Quality Improvement Metrics – the rates and supporting data quality metrics, that are built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS.

We value your feedback, please use the email address here if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also know as CQIM, for proportion of babies born at term with an Apgar score less than 7 at 5 minutes. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month and the two months prior, the proportion of babies born at term with an Apgar score <7 at 5 minutes.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for all singleton babies born in a given month and the two months prior, with a gestational length at birth between 259 and 315 days (37+0 and 45+0 weeks), where the APGAR score is between 0 and 10

For the numerator, it is very similar to the denominator. We will look for all singleton babies born in a given month and the two months prior, with a gestational length at birth between 259 and 315 days (37+0 and 45+0 weeks), however in the numerator we look for an APGAR score between 0 and 6

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published meta data to explain how the measures are built. This is available by following the bottom link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at CQIMApgar, the proportion of babies born at term with an Apgar score <7 at 5 minutes.

The metadata that details the build for CQIMApgar describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure

The timeframe for CQIMApgar is a given month and the two months prior. For example, if the given month was January 2021, then the measure would be counting babies in January 2021, December 2020 and November 2020.

Slide 6

We want to be able to uniquely identify each baby, we do this through data submitted to MSD401 baby demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (BABY). This is primarily generated using the Master Person Service (MPS) Person Index logic.

We also use MSD401 baby demographics to get the following 3 pieces of information:

• The Person Birth Date Baby to ensure the child is born in the time frame
• Gestation Length (at birth) to ensure gestation is greater than 37 weeks and less than 45 weeks (between 259 and 315 days)
• Pregnancy outcome to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 7

We use the table MSD405 Care Activity Baby to get the APGAR SCORE 5 minutes after birth

Slide 8

We use the table MSD301 Labour Delivery to get the BIRTHS PER LABOUR AND DELIVERY to identify singleton births, with 1 meaning singleton.

Slide 9

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

Slide 10

We will start by looking at the denominator, all single babies born with a gestational length at birth greater than 37 weeks and an APGAR score between 0 and 10 in a given month and the two months prior. Let’s see which data items we use to do this.

Slide 11

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.

We next consider only the babies who are singleton by checking if Births per labour and delivery is equal to 1. This is also a derived field.

Then we look for Births with a PREGNANCY OUTCOME of 01, 02, 03 or 04 to ensure the baby was born.

Next, we consider only babies with a GESTATION LENGTH (AT BIRTH) that is greater than 37 weeks.

And then consider only babies with an APGAR SCORE of between 0 and 10.

We confirm the records are in a given month and two months prior using the Person Birth Date (Baby) and the reporting period start date for each record.

This is our denominator, the number of single babies born with a gestational length at birth between 259 and 315 days (37+0 and 45+0 weeks) and with an APGAR score between 0 and 10 in a given month and the two months prior.

Slide 12

Next let’s look at how the numerator is calculated. We will look for all singleton babies born in a given month and the two months prior, with a gestational length greater than 37 weeks, however in the numerator we instead look for an APGAR score between 0 and 6. Let’s see which data items we use to do this.

Slide 13

We use the cohort of babies we found for the denominator, then retain only those where APGAR SCORE is between 0 and 6.

Slide 14

A few key points to take away are

• We take the latest data that is available for a given month and the two months prior to it.
• The denominator counts babies who:
  • are singletons
  • With a gestational length at birth between 259 and 315 days (37+0 and 45+0 weeks)
  • And with an APGAR score of between 0 and 10
• The numerator counts how many of those babies had an APGAR score of between 0 and 6.

Slide 15

The final step is to create the measure as a percentage,

Slide 16

We simply divide the numerator by the denominator and multiply by 100 to get a percentage

Slide 17

CQIMApgar is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ14 has a pass rate of greater than or equal to 70%

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 18

The description of the CQIM DQ measures can be found on this slide, this information is also in the metadata file.

Slide 19

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for the proportion of babies born at term with an Apgar score <7 at 5 minutes is built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for the proportion of babies born at term with an Apgar score of less than 7 at 5 minutes. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, referred to as MSDS.

Slide 2

Today we will look at the 4 data quality measures that support the CQIM measure, CQIMApgar, the proportion of babies born at term with an Apgar score of less than 7 at 5 minutes.

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.

This video will look at the DQ measures supporting CQIMApgar and show some examples of how these are currently displayed in the published data.

Slide 4

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built. This is available with each monthly data release, available by following the second link.

Many of the data items which appear in these DQ measures also exist in CQIMApgar, the measure to which they are applied. A video describing the MSDS measure, CQIMApgar, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5

There are 4 data quality measures that support CQIMApgar, the proportion of babies born at term with an Apgar score of less than 7 at 5 minutes. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.

The first DQ measure is CQIMDQ14. It calculates women giving birth in MSDS, as a percentage of HES average monthly deliveries, in the current 3 month reporting period.

The second is CQIMDQ15. It calculates the percentage of singleton babies with a valid gestational length at birth, that is between 154 days and 315 days, in the current 3 month reporting period.

Slide 6

The third DQ measure is CQIMDQ16. It calculates the percentage of singleton babies with a gestational length at birth between 259 days and 315 days in the current 3 month reporting period.

The fourth is CQIMDQ24. It calculates the percentage of singleton babies, live or stillbirth, with an Apgar score recorded between 0 and 10, and a valid gestational length at birth between 259 days and 315 days in the current 3 month reporting period.

The timeframe for CQIMDQ14, CQIMDQ15, CQIMDQ16 and CQIMDQ24 is the current 3 month reporting period, or in a given month and the 2 months prior. For example, if the given month was January 2021, then the DQ measures would be counting women and babies in November 2020 to January 2021.

Now, we will look at all the fields in MSDS that underpin these measures.

Slide 7

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE.

Slide 8

We want to be able to uniquely identify each mother and baby; we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER) to identify the mother, and one called PERSON ID (BABY) to identify the baby. These fields are primarily generated from the NHS Number(s) submitted for mother and baby.

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers:

• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame
• PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.
• The GESTATION LENGTH (AT BIRTH) to identify births where the gestation is in the relevant range.

Slide 9

We use the MSD301 Labour and Delivery table to ensure the number of births associated with each labour and delivery is 1, for a singleton birth. This is done by ensuring the field BIRTHS PER LABOUR AND DELIVERY = 1. This field is derived from LABOUR AND DELIVERY IDENTIFIERs which are submitted by the provider.

Slide 10

We use the MSD405 Care Activity (Baby) table to identify babies with an Apgar score between 0 and 10. This is done by ensuring the field APGAR SCORE is between 0 and 10. This field is derived from MASTER SNOMED CT OBSERVATION CODE and OBSERVATION VALUE which are fields submitted by the provider.

We will see how these items are used in the next slides.

Slide 11

We will start by looking at the denominator for CQIMDQ14, the number of deliveries in HES (using published 2018-19 data) pro-rata-ed over the number of days in the current 3 month reporting period, the given month and the 2 months prior.

Let’s see how we do this.

Slide 12

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link here.

Then, we count the number of days in the reporting period, that is in a given month and the 2 months prior.

We calculate the denominator as the number of days in the current 3 month reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19.

This is the denominator, the pro-rata-ed number of deliveries in HES.

Slide 13

Next, let’s look at how the numerator for CQIMDQ14 is calculated. We will be looking for all women who gave birth in a given month and the 2 months prior, where there was a live or stillbirth. Let’s see which data items we use to do this.

Slide 14

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next look for a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’ to include only live and stillbirths.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our numerator, the number of women who have a live or stillbirth in the current 3 month reporting period.

Slide 15

The final step is to create the CQIMDQ14 measure as a percentage.

Slide 16

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 17

We will start by looking at the denominator for CQIMDQ15, the number of singleton babies born in a given month and the 2 months prior. Let’s see how we do this.

Slide 18

We use PERSON ID (BABY) to count unique babies born in the time frame. This is a derived field. We next look for a BIRTHS PER LABOUR AND DELIVERY of 1 to include only singleton births.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of singleton babies born in the current 3 month reporting period.

Slide 19

Next, let’s look at how the numerator for CQIMDQ15 is calculated. We will be looking for the number of singleton babies born with a valid gestational length at birth between 154 days and 315 days (between 22 and 45 weeks) in a given month and the 2 months prior.

Let’s see which data items we use to do this.

Slide 20

We use the cohort of babies we found for the denominator, then retain only those recorded with a gestational length at birth between 154 and 315 days. Again, we use PERSON ID (BABY) to count unique babies born in the time frame.

To find babies with a valid gestation length at birth we look for anyone in the denominator who has a GESTATION LENGTH (AT BIRTH) between 154 and 315 days.

This is our numerator, the number of singleton babies born with a valid gestational length at birth between 154 days and 315 days (or between 22 and 45 weeks) in the current 3 month reporting period.

Slide 21

The final step is to create the CQIMDQ15 measure as a percentage.

Slide 22

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 23

We will start by looking at the denominator for CQIMDQ16, the number of singleton babies born with a valid gestational length at birth between 154 days and 315 days in a given month and the 2 months prior. Let’s see how we do this.

Note that this denominator is the same as the numerator for CQIMDQ15.

Slide 24

We use PERSON ID (BABY) to count unique babies born in the time frame. This is a derived field. We next look for a BIRTHS PER LABOUR AND DELIVERY of 1 to include only singleton births.

We confirm the records have a valid gestational length at birth, by ensuring the GESTATION LENGTH (AT BIRTH) is between 154 and 315 days.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of singleton babies born with a valid gestational length at birth between 154 and 315 days in the current 3 month reporting period.

Slide 25

Next, let’s look at the numerator for CQIMDQ16, the number of singleton babies born with a valid gestational length at birth between 259 days and 315 days in a given month and the 2 months prior. Let’s see how we do this. This numerator has a very similar definition to the denominator for CQIMDQ16.

Slide 26

We use the cohort of babies we found for the denominator, then retain only those recorded with a gestation length at birth between 259 and 315 days. Again, we use PERSON ID (BABY) to count unique babies in the time frame.

To find babies with a valid gestation length at birth we look for anyone in the denominator who has a GESTATION LENGTH (AT BIRTH) between 259 and 315 days.

This is our numerator, the number of singleton babies born with a valid gestational length at birth between 259 and 315 days, in the current 3 month reporting period.

Slide 27

The final step is to create the CQIMDQ16 measure as a percentage.

Slide 28

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 29

We will start by looking at the denominator for CQIMDQ24, the number of singleton babies born, as a live or stillbirth, with a valid gestational length at birth between 259 days and 315 days in a given month and the 2 months prior.

Slide 30

We use PERSON ID (BABY) to count unique babies born in the time frame. This is a derived field. We next look for a BIRTHS PER LABOUR AND DELIVERY of 1 to include only singleton births.

We then look for a PREGNANCY OUTCOME of 01, 02, 03 or 04 to include only live and stillbirths.

We confirm the records have the correct gestational length at birth, by ensuring the GESTATION LENGTH (AT BIRTH) is between 259 and 315 days.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of singleton babies born as a live or stillbirth, with a valid gestational length at birth between 259 and 315 days, in the current 3 month reporting period.

Slide 31

Next, let’s look at how the numerator for CQIMDQ24 is calculated. We will be looking for all singleton babies born as a live or stillbirth, with an Apgar score recorded between 0 and 10, and a valid gestational length at birth between 259 and 315 days, in a given month and the 2 months prior.

Slide 32

We use the cohort of babies we found for the denominator, then retain only those with an Apgar score recorded between 0 and 10. As in the denominator, we use PERSON ID (BABY) to count unique babies in the time frame.

We keep records found in the denominator, where the recorded APGAR SCORE is between 0 and 10, to find babies with a valid Apgar score at 5 minutes.

This is our numerator, representing the number of singleton babies born as a live or stillbirth, with an Apgar score recorded between 0 and 10, and a valid gestational length at birth between 259 and 315 days in the current 3 month reporting period.

Slide 33

The final step is to create the CQIMDQ24 measure as a percentage.

Slide 34

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 35

Now, we look at the pass thresholds for the four DQ measures supporting CQIMApgar. A provider must pass all the data quality measures to be included in the published data. The provider’s rate for CQIMDQ14, CQIMDQ15, CQIMDQ16 and CQIMDQ24 must all be 70% or more.

We will see how the values reached by the DQ measures affect what is shown in the published data.

Slide 36

Let’s look at the data published for January 2022.

RDE – East Suffolk and North Essex NHS Foundation Trust – has passed all four of the DQ measures associated with CQIMApgar. RDE reached 85.5% for CQIMDQ14, 99.6% for CQIMDQ15, 95.1% for CQIMDQ16 and 97.8% for CQIMDQ24.

This means that CQIMApgar figures are published for RDE, and this provider will contribute to the national and sub-national figures included in the January 2022 published data.

Slide 37

Now, we will look at another trust using the January 2022 data.

R1H –Barts Health NHS Trust – has passed three of the DQ measures associated with CQIMApgar. R1H reached 80.4% for CQIMDQ14, 92.3% for CQIMDQ16 and 98.5% for CQIMDQ24. It failed the DQ measure CQIMDQ15, with only 61.6%.

This means that CQIMApgar figures are not published for R1H. Instead, the denominator and numerator are shown as ‘0’ and the Rate per Thousand as ‘Low DQ’ for CQIMApgar and will not contribute to the national and sub national figures included in the January 2022 published data.

Slide 38

Let’s look at how applying these data quality measures affects the larger picture. 120 providers submitted data for January 2022 births. As the CQIM DQ measures span a time frame longer than just one month, 124 providers were considered for CQIMDQ14, 15, 16 and 24.

In January 2022:

116 providers passed CQIMDQ14, recording at least 70% for this DQ measure. 8 providers failed the measure or did not submit any data for this measure at all. 117 providers passed CQIMDQ15, 120 providers passed CQIMDQ16, and 110 providers passed CQIMDQ24.

Note that the pass rates of these measures could alter in the future.

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.

Slide 39

This brings us to the end of the video.

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, the proportion of babies born at term with an Apgar score of less than 7 at 5 minutes, are built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, the proportion of babies with a first feed of breast milk. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month, the proportion of babies with a first feed of breast milk.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look at the number of babies born in a given month that had a recorded first feed type.

For the numerator, we take the babies in the denominator and retain only those where the baby had a recorded first feed type of maternal or donor milk.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25 available from the top link here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, available by following the bottom link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at the proportion of babies with a first feed of breast milk ‘CQIMBreastfeeding’.

The metadata that details the build for CQIMBreastfeeding describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure. We will look at each table in turn.

For reference, the timeframe for CQIMBreastfeeding is a given month. For example, January 2021

Slide 6

We want to be able to uniquely identify each baby, we do this through data submitted to MSD401 Baby Demographics.

From this table we get the following 3 pieces of information:

• Person ID (Baby) - NHS Digital gives each baby a unique identifier derived through items submitted by all providers. We use this to uniquely identify each baby.
• The Person Birth Date (Baby) to ensure the child is born in the relevant time frame
• Baby First Feed Breast Milk Indication Code, to ensure the first feed is either ‘01’, meaning maternal breast milk or ‘02’ meaning, donor breast milk.

Slide 7

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

We will see how these items are used in the next slides

Slide 8

We will start by looking at the denominator, Number of babies with recorded first feed type in a given month. Let’s see which data items we use to do this.

Slide 9

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.

Next, we check that that birth occurred in a given month by confirming the REPORTING PERIOD START DATE and PERSON BIRTH DATE (BABY) are within that month.

Finally we confirm that the BABY FIRST FEED BREAST MILK INDICATION CODE is ’01’, ‘02’ or ‘03’ as these are all valid codes.

This is our denominator, the number of babies with recorded first feed type in a given month.

Slide 10

Next let’s look at how the numerator is calculated. We will look at the number of babies that received maternal or donor milk for their first feed in a given month. Let’s see which data items we use to do this.

Slide 11

We use the cohort of babies we found for the denominator, then retain only those with a BABY FIRST FEED BREAST MILK INDICATION CODE of ‘01’ or ‘02’, meaning breastmilk or donor milk.

Again, we use PERSON ID (BABY) to count unique babies in the time frame.

Slide 12

A few key points to take away are

• We take the latest data that is available for a given month.
• The denominator counts babies who:
  • Were born in a given month
  • With a valid first feed first code

• The numerator counts how many of those babies had a first feed that was maternal breast milk or donor breast milk.

Slide 13

The final step is to create the measure as a percentage,

Slide 14

We simply divide the numerator by the denominator and multiply by 100 to get a percentage

Slide 15

CQIMBreastfeeding is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

Both CQIMDQ08 and CQIMDQ09 have a pass rate of 70%.

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 16

The description of the CQIM DQ measures can be found on this slide, this information is all in the metadata file.

Slide 17

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure the proportion of babies with a first feed of breast milk is built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for babies receiving a first feed of breast milk. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, often referred to as MSDS.

Slide 2

Today we will look at the 2 data quality measures that support the CQIM measure, CQIMBreastfeeding, the proportion of babies receiving a first feed of breast milk.

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.

This video will look at the DQ measures for CQIMBreastfeeding and show some examples of how these are currently displayed in the published data.

Slide 4

We will look at the data items in MSDS and HES (the Hospital Episode Statistics database) that contribute to these DQ measures.

All of the MSDS data items can be found in the technical output specification version 2.0.25 available from the top link here.

There is also published metadata to explain how the measures are built. This is available alongside each monthly data release, available by following the middle link.

These DQ measures also require information derived from HES, which was published as part of the annual maternity publication ‘NHS Maternity Statistics, England, 2018-19’ and can be found through the third link here.

The MSDS data items used in these data quality thresholds also exist in the measure they are supporting. A video describing the MSDS measure, CQIMBreastfeeding, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the final link.

Please note, all links in this video can be found in the transcript below.

Slide 5

There are 2 data quality thresholds that support CQIMBreastfeeding, the proportion of babies receiving a first feed of breast milk.

The first is referred to as CQIMDQ08. It calculates the percentage of babies whose first feed type was recorded. It takes all babies born in the current 1 month reporting period, whose first feed was recorded, as the numerator, and all babies whose birth was recorded in that same reporting period, as the denominator.

The second is referred to as CQIMDQ09. It calculates women giving birth - as recorded in MSDS - in the current 1 month reporting period, as a percentage of HES average monthly deliveries. It takes all mothers who gave birth in the current 1 month reporting period – as recorded in MSDS - as the numerator, and uses an estimation of HES average monthly deliveries as the denominator.

Providers must pass both of these measures at at least 70%, for their data for CQIMBreastfeeding to be published and included in national and sub national figures.

The time frame for CQIMDQ08-09 is the given month only. For example, if the given month was July 2021, then these CQIM DQ measures would be counting women and babies in July 2021.

Now, we will look at all the fields in MSDS that underpin these measures.

Slide 6

We will look at each MSDS table in turn; some data items are derived from other items sent by providers, this is identified by the source field shown below.

We will start by looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE.

Slide 7

We want to be able to uniquely identify each baby and mother. We do this through data submitted to the table MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital derives a field called PERSON ID (MOTHER) to identify the mother, and one called PERSON ID (BABY) to identify the baby. These fields are primarily generated from the NHS number(s) submitted for mother and baby.

We also use the field PERSON BIRTH DATE (BABY) – to identify the baby’s date of birth, in order to determine whether it took place in the reporting period; and the field BABY FIRST FEED BREAST MILK INDICATION CODE – to find out if the baby’s first feed was recorded. Both of these fields are submitted to the table MSD401 Baby Demographics by the provider.

Slide 8

We also need the HES average monthly delivery count. This will be derived from counts of deliveries published in the annual report ‘NHS Maternity Statistics, England, 2018-19’. The annual counts can be found in the ‘ALL episodes’ column of the supporting data file ‘HES Provider Level Analysis: MPDP Flat file (18-19)’ and details of the monthly average derivation will be discussed later in this presentation.

Slide 9

We will start by looking at CQIMDQ08. This data quality threshold measure assesses the percentage of babies whose first feed type was recorded.

Slide 10

For this DQ measure, we will begin by looking at the denominator for CQIMDQ08 – the number of babies who were born in the current 1 month reporting period. Let’s see how we do this.

Slide 11

We use PERSON ID (BABY) to count unique babies in the time frame.

We confirm that the babies are born in a given month, by using the PERSON BIRTH DATE (BABY) and REPORTING PERIOD START DATE, ensuring that both dates are in the current reporting period.

This is our denominator, the number of babies who are born in a given month.

Slide 12

We will now look at how the numerator is calculated for CQIMDQ08. We will be looking for all the babies who were born in the current 1 month reporting period, and who were recorded with a valid first feed type. Let’s see which data items we use to do this.

Slide 13

We use the cohort of babies we found for the denominator, then retain only those who were recorded with a valid first feed type at birth. Again, we use PERSON ID (BABY) to count unique babies in the time frame.

To find babies with a valid first feed type – we look for anyone in the denominator who has a BABY FIRST FEED BREAST MILK INDICATION CODE of '01', '02’ or '03’. This includes babies whose first feed was maternal or donor breast milk and babies whose first feed did not involve breast milk, but not babies whose first feed type is unknown.

This is our numerator, the number of babies who are born in a given month, and who are recorded with a valid first feed at birth.

Slide 14

The final step is to take the numerator and the denominator and create the CQIMDQ08 measure as a percentage. We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 15

Next, we will look at CQIMDQ09. This data quality threshold measure assesses the number of births recorded in MSDS, as a proportion of those recorded in HES.

HES is another source of data on women who give birth. Hospital Episode Statistics – or HES – contains records of all hospital admissions that take place at NHS hospitals in England (as well as outpatient and A&E visits). Women who give birth at an NHS hospital in England will have their delivery recorded in HES. HES does not record deliveries where the baby was born elsewhere – at home, at a private hospital, or in Wales.

Slide 16

We will start by looking at the denominator for CQIMDQ09, the number of deliveries in HES (using published 2018-19 data) pro-rata-ed over the number of days in a given month. Let’s see how we do this.

Slide 17

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics, 2018-19, which can be found at the link to that publication.

Then, we count the number of days in the reporting period, that is, in a given month. If the reporting period is July 2021, then this is the total number of days in July 2021.

We calculate the denominator as the number of days in the reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19.

This is our denominator, the pro-rata-ed number of deliveries in HES.

Slide 18

Next, let’s look at how the numerator for CQIMDQ09 is calculated. We will be looking for all women who gave birth in the current 1 month reporting period. Let’s see which data items we use to do this.

Slide 19

We use PERSON ID (MOTHER) to count unique women in the time frame.

We confirm that the births take place in a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record, ensuring that both dates are in the current reporting period.

This is our numerator, the number of women who gave birth in a given month.

Slide 20

The final step is to take the numerator and the denominator and create the CQIMDQ09 measure as a percentage. We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 21

A provider must pass the data quality thresholds for both DQ measures, CQIMDQ08 and CQIMDQ09, with at least 70% in order for the CQIM measure they support, to be included in the published data.

If the denominator of CQIMDQ09 is 0 and the numerator is greater than 0, then the provider will pass that DQ measure.

We will now look at how these measures affect what is shown in the published data.

Slide 22

We will use data published for July 2021 to show how the data quality thresholds work. (The link to this data is available from the transcript below.) We will look at 3 organisations that illustrate the possible outcomes for the DQ measures.

First, we will look at R0A – Manchester University NHS Foundation Trust - we can see that they have passed both data quality thresholds. They have passed CQIMDQ08, with 97.8% of babies having their first feed recorded. They also passed CQIMDQ09 at 115.4%, indicating more deliveries were recorded in MSDS in July 2021 - than was reported to HES, on average, in 2018-19.

Next, we will look at R1H – Barts Health NHS Trust. Here we see that the provider has passed just one of the data quality thresholds. R1H passed CQIMDQ09 with deliveries, recorded in MSDS, equivalent to 82.7% of the monthly average of those reported to HES in 2018-19; while they did not pass CQIMDQ08, as only 46.0% of babies born in the month, have had their first feed recorded.

Finally, we will look at RLT – George Eliot Hospital NHS Trust. We can see that they have failed to pass either data quality threshold. They have failed to pass CQIMDQ08, as only 50.0% of babies born in the month had their first feed recorded. They also failed to pass CQIMDQ09, with deliveries recorded in MSDS being just 50.0% of the average monthly deliveries recorded in HES in 2018-19.

Now, we will look at what this means for the measure CQIMBreastfeeding that these DQ measures are supporting.

Slide 23

As R0A passed both CQIMDQ08 and CQIMDQ09, their data for CQIMBreastfeeding is published, with a rate of 68.4%, and the provider will contribute to national and sub national breakdowns.

As R1H passed only one of the data quality thresholds their data is not published in the final measure of CQIMBreastfeeding and does not feature in the national or sub national figures. And as RLT did not pass either of the data quality thresholds their data is not published in the final measure of CQIMBreastfeeding and does not feature in the national or sub national figures either. Instead, for both R1H and RLT, the denominator and numerator for CQIMBreastfeeding are shown as ‘0’ and the rate as ‘Low DQ’.

Slide 24

Let’s look at the overall picture.

For July 2021, 124 providers submitted data to MSDS.

Looking at CQIMDQ08, 105 providers passed this threshold, with an average pass rate of 92%. Looking at CQIMDQ09, 114 providers passed this threshold, with an average pass rate of 97%. 21 providers did not reach the 70% threshold for both DQ measures, as either they did not supply enough information on babies’ first feed or recorded too few delivery records in comparison to HES, or both.

Note that the data quality thresholds of 70% could alter in future as providers start to improve data quality. Providers should aim to pass both data quality thresholds with a completion rate higher than 70%.

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal; the link to access this can be found in the transcript.

Slide 25

This brings us to the end of the video.

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, the proportion of babies receiving a first feed of breast milk, are built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, for babies who were fully or partially breastfed between 6 and 8 weeks old. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS. 

Slide 2

Today we will look at a measure that finds, for any given month, the proportion of 9 week old babies, who had been fully or partially breastfed at any time between the age of 6 and 8 weeks.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look at the number of babies who were 9 weeks (63 days) old during the current 1 month reporting period, where their birth did not involve a stillbirth or a termination; and they were recorded with one consistent date of birth, and no more than one consistent date of death.

For the numerator, we keep only those babies from the denominator who were fully or partially breastfed at any time between the age of 6 and 9 weeks (42 to 63 days). This allows an additional week of leeway in recording, and any baby with a breastfeeding record between 8 and 9 weeks will be included.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the MSDS technical output specification version 2.0.25, available from the first link here

This measure also uses data items collected in CSDS, the Community Services Dataset. For the data items recorded in CSDS, we will refer to the CSDS technical output specification version 1.5.25, available from the second link here.

There is also published metadata to explain how the measures are built. Links to the latest MSDS metadata are given alongside each monthly data release, available by following the third link.

Slide 5

This video explains how the measure is constructed and the data items used. There are 12 CQIM ‘rate’ measures. This video will look only at CQIMBreastfeeding6to8Weeks, the proportion of 9 week old babies who had been fully or partially breastfed at any time between the age of 6 and 8 weeks.

The metadata that details the build for CQIMBreastfeeding6to8Weeks describes the measure, the numerator, and the denominator.

The timeframe for CQIMBreastfeeding6to8Weeks is a given month only. For example, if the given month was January 2022, then the measure would be counting babies who were 9 weeks old in January 2022. These babies would have been born in October or November 2021.

Now, we will look at all the fields in MSDS that underpin this measure.

Slide 6

We will start with looking at the table MSD401 Baby Demographics.

From the data submitted by all providers, NHS England creates a field called PERSON ID (BABY) to identify the baby. This field is primarily generated from the NHS Number submitted for the baby. We will use this field to be able to uniquely identify the baby.

We also use MSD401 Baby Demographics to get the following pieces of information:

• The PERSON BIRTH DATE (BABY) to ensure the baby was 9 weeks old in the relevant time frame.
• The PERSON DEATH DATE (BABY) and AGE AT DEATH (BABY) to exclude any baby from the measure where death occurred before the baby was 9 weeks old.
• The PERSON BIRTH DATE (BABY) and PERSON DEATH DATE (BABY) are submitted by providers while the AGE AT DEATH is a derived field, the baby’s age at death in hours, calculated from the date and time of the baby’s birth and death. Information about deaths more than 28 days after the birth of the baby is not recorded in MSD401.

Slide 7

• The PREGNANCY OUTCOME to ensure that a baby is not included in the measure if it is known to be a stillbirth ‘02’, ‘03’, ‘04’ or a termination ‘05’.
• AGE IN YEARS AT BABYS BIRTH (MOTHER) to provide a breakdown of CQIMBreastfeeding6to8Weeks by maternal age.
• ORGANISATION IDENTIFIER (CODE OF PROVIDER) to identify the organisation that submitted the data.
• REPORTING PERIOD START DATE and REPORTING PERIOD END DATE to establish the reporting month for the measure.

The AGE IN YEARS AT BABYS BIRTH (MOTHER) is calculated from PERSON BIRTH DATE (BABY) and PERSON BIRTH DATE (MOTHER), both of which are fields submitted by the provider. The other data items are submitted by the provider – either directly to MSD401, or to the MSD000 Header table.

Slide 8

CQIMBreastfeeding6to8Weeks is reported in a number of breakdowns in addition to the standard maternity groupings. We use the MSD001 Mother’s Demographics table to support these additional breakdowns. ETHNIC CATEGORY (MOTHER) is used to break the measure down by maternal ethnicity. The fields LOCAL AUTHORITY DISTRICT/UNITARY AUTHORITY (MOTHER) and the LOWER SUPER OUTPUT AREA (RESIDENCE) OF MOTHER 2011 are used to identify the local authority of the mother, and to derive the decile of deprivation of the mother at the time of the baby’s birth, respectively. These fields are both derived from the POSTCODE OF USUAL ADDRESS (MOTHER), which is submitted by the provider.

Slide 9

We use the MSD002 GP Practice Registration table to identify the GP practice of the mother at the time of the baby’s birth. The fields GENERAL MEDICAL PRACTICE CODE (PATIENT REGISTRATION (MOTHER)), START DATE (GMP PATIENT REGISTRATION) and END DATE (GMP PATIENT REGISTRATION) are used to identify the primary care network of the mother, which is used as another additional breakdown for this measure.

Slide 10

Now, we will look at all the fields in CSDS which are used to calculate this measure.

We use the CYP001 Master Patient Index table to find records of babies in CSDS which do not have records in MSDS, as well as to provide additional demographic information about the babies’ dates of birth and death which may not have been recorded in MSD401.

In order to uniquely identify any baby in CYP001, we use PERSON ID, a derived field which depends on data items submitted by the provider, describing the person’s demographic information.

We also use the PERSON BIRTH DATE and AGE OF PATIENT AT REPORTING PERIOD START (DAYS) to identify babies who were 9 weeks old in the current 1 month reporting period. Information about the baby’s death is taken from the PERSON DEATH DATE, and AGE AT DEATH (DAYS) field. The latter of which is a derived field, dependent on the PERSON DEATH DATE and PERSON BIRTH DATE fields submitted by the provider.

Slide 11

To identify the organisation that submitted the CSDS record for the baby – which may not be the same organisation that submitted a record of the baby’s birth to MSDS – we use the field ORGANISATION IDENTIFIER (CODE OF PROVIDER). To restrict CSDS records to those submitted by the current reporting period, we use the RECORD START DATE. Both of these fields are submitted by the provider to CYP000 Header.

Slide 12

We use the CSDS table CYP201 Care Contact to find when the record of the baby’s breastfeeding status took place. We do this by looking at the CARE CONTACT DATE and comparing to the baby’s birth date to ensure that records are only included for contacts taking place involving babies aged between 6 and 9 weeks old.

Slide 13

Next, we use the table CYP202 Care Activity to find which care contacts were recorded with information about the baby’s breastfeeding status. The DERIVED BREASTFEEDING STATUS (MASTER) is derived from the CODED FINDING (CODED CLINICAL ENTRY) field, where values from that field are SNOMED or Read codes which are associated with breastfeeding, and from BREASTFEEDING STATUS where breastfeeding information is recorded in the table CYP610 Breastfeeding Status.

Slide 14

Finally, breastfeeding information is taken from the CSDS table, CYP610 Breastfeeding Status, using the field BREASTFEEDING STATUS, where providers can directly submit information about the current level of breastfeeding as exclusive, partial or no breastfeeding.

Slide 15

We will start by looking at the denominator, the number of babies who were alive at 63 days old (9 weeks), with no stillbirth or termination, and consistent recording of dates of birth and death, in the current 1 month reporting period.

Let’s see which data items we use to do this.

Slide 16

In table MSD401, we use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. Where babies have no MSDS record, but are only recorded in CYP001, we use the derived field PERSON ID instead.

Firstly, we keep only babies who were born 63 days before the current 1 month reporting period. For babies recorded in MSD401, this is where PERSON BIRTH DATE (BABY) + 63 is in a given month, and the REPORTING PERIOD START DATE is no later than the given month. For babies who are recorded only in CYP001, this is where PERSON BIRTH DATE (or RECORD START DATE – AGE OF PATIENT AT REPORTING PERIOD START DATE) + 63 is in a given month, and the RECORD START DATE is no later than the given month.

Now we have compiled a list of identifiers for babies who may be eligible for the denominator, we check the PREGNANCY OUTCOME, and remove babies who have been recorded with a PREGNANCY OUTCOME of ‘02’, ‘03’, ‘04’ or ‘05’ to ensure that no baby in the denominator was born as a stillbirth or termination. If a baby has no recorded PREGNANCY OUTCOME because there is no MSD401 record, then we keep that baby.

Next, we look at all records in MSDS and CSDS recorded no later than the current 1 month reporting period - using REPORTING PERIOD START DATE (MSD401) or RECORD START DATE (CYP001) – to confirm that the baby has been recorded with one consistent date of birth across the records in both datasets. We check PERSON BIRTH DATE (BABY), PERSON BIRTH DATE and RECORD START DATE – AGE OF PATIENT AT REPORTING PERIOD START DATE to do this, and remove any baby where inconsistent dates of birth are recorded.

Then, we look at how dates of death are recorded in MSDS and CSDS, considering all records submitted no later than the current 1 month reporting period - using REPORTING PERIOD START DATE (MSD401) or RECORD START DATE (CYP001). Here, we keep all babies who have zero or one date of death recorded across the 2 datasets, using information recorded in PERSON DEATH DATE (BABY), PERSON DEATH DATE, AGE AT DEATH (DAYS).

Finally, we keep only babies who either have no date of death recorded, or who have a date of death more than 63 days after the baby was born. We check records from MSDS and CSDS to do this – looking at the field AGE AT DEATH (DAYS) directly and calculating the baby’s age where dates of birth and death are recorded as the following -

• PERSON DEATH DATE – PERSON BIRTH DATE
• PERSON DEATH DATE – (RECORD START DATE – AGE OF PATIENT AT REPORTING PERIOD START (DAYS))
• PERSON DEATH DATE (BABY) – PERSON BIRTH DATE (BABY)

- keeping babies where the age is either null or greater than 63 (recorded in days) and looking directly at AGE AT DEATH (BABY), keeping babies where the age is either null or greater than 1,512 (recorded in hours).

This is our denominator, the number of the number of babies who were alive at 63 days old (9 weeks), with no stillbirth or termination, and consistent recording of dates of birth and death, in a given month.

Slide 17

Next, let’s look at how the numerator is calculated. We will be looking for the number of all babies who were alive at 63 days old, with no stillbirth or termination, and consistent recording of dates of birth and death, in the current 1 month reporting period, where the baby was fully or partially breastfed between 42 and 63 days old (6 to 9 weeks). Let’s see which data items we use to do this.

Slide 18

We use the cohort of babies we found for the denominator, then retain only those who have a care contact which occurred between 42 and 63 days after the date of birth, at which their breastfeeding status was recorded as exclusive (full) or partial breastfeeding.

Again, we use PERSON ID (BABY) and PERSON ID to count unique babies in the time frame.

This time we take the babies found for the denominator, and limit to include only those with a CARE CONTACT DATE taking place between PERSON BIRTH DATE (BABY) + 42 and PERSON BIRTH DATE (BABY) + 63. For babies with no MSD401 record, we use PERSON BIRTH DATE or RECORD START DATE – AGE OF PATIENT AT REPORTING PERIOD START (DAYS) as the birth date instead to find care contacts between the age of 42 and 63 days.

We keep only babies with a DERIVED BREASTFEEDING STATUS (MASTER) of ‘B1’ or ‘B2’, or those with a BREASTFEEDING STATUS of ‘01’ or ‘02’, which show the baby was exclusively or partially breastfed at the time of the care contact.

This is our numerator, the number of babies who were alive at 63 days old (9 weeks) in a given month, with no stillbirth or termination, consistent recording of dates of birth and death, and where the baby was fully or partially breastfed between 42 and 63 days old (6 to 9 weeks).

Slide 19

The final step is to create the measure as a percentage

Slide 20

We simply divide the numerator by the denominator and multiply by 100 to calculate the percentage rate.

Slide 21

CQIMBreastfeeding6to8Weeks is not reliant on any data quality measures. Currently, providers do not need to meet the pass rate for any data quality measure in order for their data for this measure to be considered of good enough quality to be published.

This may change as further assessments are made of the quality of data submitted for this measure.

Slide 22

A few key points to take away are:

• We take the latest data that is available for the current 1 month reporting period and any earlier months, to find babies who were 63 days old (9 weeks), during the reporting period.
• A baby can be included in this measure if there is a record which indicates they were 63 days old during the reporting period; this may be from a record of the birth in MSDS (MSD401) or if the baby has a record in CSDS (CYP001).
• Babies are considered to be breastfed between 6 and 8 weeks if there is a record in CSDS of exclusive or partial breastfeeding at any point between 42 and 63 days old. If a baby has both a record of ‘No breastfeeding’ and a record of exclusive or partial breastfeeding between 42 and 63 days old then it will be included in the numerator for the measure. An additional week’s leeway is allowed, meaning that babies with a relevant breastfeeding status recorded between 8 and 9 weeks will also be included.
• A baby will be excluded from this measure if
  • There are multiple dates of birth, or dates of death, recorded in MSDS (MSD401) and CSDS (CYP001).
  • There is an MSDS (MSD401) record of a stillbirth or termination for the baby
  • There is a date of death less than 63 days after the date of birth, recorded in MSDS (MSD401) or CSDS (CYP001).
• All relevant records are included in CQIMBreastfeeding6to8Weeks, as currently there are no data quality metrics supporting this CQIM rate.
• In addition to the standard MSDS sub national breakdowns, CQIMBreastfeeding6to8Weeks includes breakdowns for maternal age, ethnicity, and deprivation, as well as local authority and primary care network.

Slide 23

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for babies who were fully or partially breastfed between 6 and 8 weeks old is built from MSDS and CSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, for proportion of women who had a postpartum haemorrhage of 1,500ml or more. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in the three months prior to the reporting period, the proportion of women who had a postpartum haemorrhage of 1,500ml or more.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for the number of women with a submitted birth record, in the three months prior to the reporting period.

For the numerator, it is very similar to the denominator. we will look for the number of women with a submitted birth record, in the three months prior to the reporting period and then also look for only women with a recorded postpartum haemorrhage of 1,500ml or more.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All these data items can be found in the technical output specification version 2.0.25 available from the top link here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release by following the bottom link here.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at CQIMPPH, the proportion of women who had a postpartum haemorrhage of 1,500ml or more.

The metadata that details the build for CQIMPPH describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure

The timeframe for CQIMPPH is three months prior to the reporting period. For example, if the given month was January 2021, then the measure would be counting babies in December 2020, November 2020 and October 2020.

Slide 6

We want to be able to uniquely identify each woman, and we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER). This is primarily generated from the NHS Number.

We also use MSD401 Baby Demographics to get the following pieces of information:

• The PERSON BIRTH DATE (BABY) to ensure the birth is in the time frame
• PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 7

We use the table MSD302 Care Activity Labour and Delivery to get OBSERVATION VALUE and MASTER SNOMED CT OBSERVATION CODE, a derived field based on CODED OBSERVATION (CLINICAL TERMINOLOGY). These are used to find only women who had an observed postpartum haemorrhage of 1500ml or more.

Slide 8

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

Slide 9

We will start by looking at the denominator, the number of women with a submitted birth record in the three months prior to the reporting period. Let’s see which data items we use to do this.

Slide 10

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.

We next consider only the women with a PREGNANCY OUTCOME of 01, 02, 03 or 04.

We then confirm the records are in the three months prior to the reporting period using the Person Birth Date (Baby) and the reporting period start date for each record.

This is the denominator, the number of women with a submitted birth record, in the three months prior to the reporting period.

Slide 11

Next let’s look at how the numerator is calculated. As in the denominator, we will look for the number of women with a submitted birth record in the three months prior to the reporting period, however we will only include women with a recorded postpartum haemorrhage of 1,500ml or more. Let’s see which data items we use to do this.

Slide 12

We use the cohort of women we found for the denominator, then retain only the women who had any of the postpartum haemorrhage SNOMED CT codes listed in the field MASTER SNOMED CT OBSERVATION CODE and an OBSERVATION VALUE of greater than or equal to 1500.

This is the numerator, the number of women with a submitted birth record in the three months prior to the reporting period, who have a postpartum haemorrhage of 1500ml or more.

Slide 13

A few key points to take away are

• We take the latest data that is available for three months prior to the reporting period.
• We retain only the women who had a registered birth recorded using PREGNANCY OUTCOME
• The numerator counts how many of those women had a postpartum haemorrhage of 1,500ml or more that has been recorded, using the relevant SNOMED CT codes, and an OBSERVATION VALUE greater than or equal to 1500.

Slide 14

The final step is to create the measure as a rate.

Slide 15

We simply divide the numerator by the denominator and multiply by 1000 to get the rate per thousand.

Slide 16

CQIMPPH is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ10 has a pass rate of greater than or equal to 70%. CQIMDQ13 passes if the measure is equal to 1.

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 17

The description of the CQIM DQ measures can be found on this slide. This information is also in the metadata file.

Slide 18

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for the proportion of women who had a postpartum haemorrhage of 1,500ml or more is built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for women who had a postpartum haemorrhage of 1,500ml or more. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, referred to as MSDS.

Slide 2

Today we will look at the four data quality measures that support the CQIM measure, CQIMPPH, women who had a postpartum haemorrhage of 1,500ml or more.

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.

This video will look at the DQ measures supporting CQIMPPH and show some examples of how these are currently displayed in the published data.

Slide 4

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built. This is available with each monthly data release, available by following the second link.

Many of the data items which appear in these DQ measures also exist in CQIMPPH, the measure to which they are applied. A video describing the MSDS measure, CQIMPPH, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5

There are four data quality measures that support CQIMPPH, the percentage of women who had a postpartum haemorrhage of 1,500ml or more. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.

The first DQ measure is CQIMDQ10. It calculates women giving birth, as a percentage of HES average monthly deliveries, in the previous 3 month reporting period.

The second is CQIMDQ11. It calculates the percentage of women with a recorded postpartum haemorrhage of 500ml or more, in the previous 3 month reporting period.

Slide 6

The third DQ measure is CQIMDQ12. It calculates the percentage of women with a recorded postpartum haemorrhage of 1,500ml or more, in the previous 3 month reporting period.

The fourth is CQIMDQ13. It calculates whether at least 1 postpartum haemorrhage is recorded, in the previous 6 month reporting period.

The time frame for CQIMDQ10, CQIMDQ11, and CQIMDQ12 is the 3 months prior to a given month; and for CQIMDQ13 it is the 6 months prior to a given month. For example, if the given month was January 2021, then CQIMDQ10, CQIMDQ11, and CQIMDQ12 would be counting women in October 2020 to December 2020; while CQIMDQ13 would be counting women in July 2020 to December 2020.

Now, we will look at all the fields in MSDS that underpin these measures.

Slide 7

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

Slide 8

We want to be able to uniquely identify each mother, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER). This is primarily generated from the NHS Number.

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers:

• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame
• PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 9

We use the MSD302 Care Activity Labour and Delivery table to identify women who had a postpartum haemorrhage or a delivery that involved an event considered to be a maternal critical incident.

To do this we use the MASTER SNOMED CT OBSERVATION CODE to find instances of postpartum haemorrhage, the OBSERVATION VALUE to identify where the postpartum haemorrhage was at least 500 or 1,500 ml, and the CLINICAL INTERVENTION DATE (MOTHER) to ensure that any postpartum haemorrhage took place in the reporting period, the 3 months prior to a given month.

To find maternal critical incidents we look in the MASTER SNOMED CT FINDING CODE and the MASTER SNOMED CT PROCEDURE CODE to identify where the delivery involved an event such as an ICU admission, undiagnosed breech, or any other event that would be considered an MCI.

We will see how these items are used in the next slides.

Slide 10

We will start by looking at the denominator for CQIMDQ10, the number of deliveries in HES (using published 2018-19 data) pro rata-ed over the number of days in the 3 months prior to a given month. Let’s see how we do this.

Slide 11

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link here.

Then, we count the number of days in the reporting period, that is in the 3 months prior to a given month. If the reporting period is January 2021, then this is the total number of days in October, November and December 2020.

We calculate the denominator as the number of days in the reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19.

This is the denominator, the pro-rata-ed number of deliveries in HES.

Slide 12

Next, let’s look at how the numerator for CQIMDQ10 is calculated. We will be looking for all women who gave birth in the 3 months prior to a given month, where there was a live or stillbirth. Let’s see which data items we use to do this.

Slide 13

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next look for a PREGNANCY OUTCOME of 01, 02, 03 or 04 to include only live and stillbirths.

We confirm the records are in the 3 months prior to a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our numerator, the number of women who have a live or stillbirth in the 3 months prior to a given month.

Slide 14

The final step is to create the CQIMDQ10 measure as a percentage

Slide 15

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 16

We will start by looking at the denominator for CQIMDQ11, the number of women who gave birth in the previous 3 month reporting period, where the outcome was a live or stillbirth. Let’s see how we do this.

Slide 17

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next look for a PREGNANCY OUTCOME of 01, 02, 03 or 04 to include only live and stillbirths.

We confirm the records are in the 3 months prior to a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of women who have a live or stillbirth in the 3 months prior to a given month.

Slide 18

Next, let’s look at how the numerator for CQIMDQ11 is calculated. We will be looking for all women who gave birth in the previous 3 month reporting period, where the outcome was a live or stillbirth, and the woman either experienced a critical incident during delivery or a postpartum haemorrhage of 500ml or more. Let’s see which data items we use to do this.

Slide 19

We use the cohort of women we found for the denominator, then retain only those recorded with a maternal critical incident or a postpartum haemorrhage of 500ml or more. Again, we use PERSON ID (MOTHER) to count unique women in the time frame.

To find women with a postpartum haemorrhage we look for anyone in the denominator who has a relevant SNOMED CT code recorded in MASTER SNOMED CT OBSERVATION CODE, together with an OBSERVATION VALUE of at least 500, and a CLINICAL INTERVENTION DATE (MOTHER) indicating that this happened in the 3 months prior to a given month.

To find women who had a maternal critical incident, we look for anyone in the denominator who has one of the listed SNOMED CT codes recorded in either MASTER SNOMED CT PROCEDURE CODE or MASTER SNOMED CT FINDING CODE.

This is our numerator, the number of women who have a live or stillbirth in the 3 months prior to a given month, which involved an MCI or a postpartum haemorrhage of 500ml or more.

Slide 20

The final step is to create the CQIMDQ11 measure as a percentage

Slide 21

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 22

We will start by looking at the denominator for CQIMDQ12, the number of women who gave birth in the previous 3 month reporting period, where the outcome was a live or stillbirth. This is the same denominator that is used for CQIMDQ11.

Slide 23

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next look for a PREGNANCY OUTCOME of 01, 02, 03 or 04 to include only live and stillbirths.

We confirm the records are in the 3 months prior to a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of women who have a live or stillbirth in the 3 months prior to a given month.

Slide 24

Next, let’s look at how the numerator for CQIMDQ12 is calculated. We will be looking for all women who gave birth in the previous 3 month reporting period, where the outcome was a live or stillbirth, and the woman either experienced a critical incident during delivery or a postpartum haemorrhage of 1,500ml or more. This is very similar to the numerator that is used for CQIMDQ11.

Slide 25

We use the cohort of women we found for the denominator, then retain only those recorded with a maternal critical incident or a postpartum haemorrhage of 1,500ml or more. Again, we use PERSON ID (MOTHER) to count unique women in the time frame.

To find women with a postpartum haemorrhage we look for anyone in the denominator who has a relevant SNOMED CT code recorded in MASTER SNOMED CT OBSERVATION CODE, together with an OBSERVATION VALUE of at least 1,500, and a CLINICAL INTERVENTION DATE (MOTHER) indicating that this happened in the 3 months prior to a given month.

To find women who had a maternal critical incident, we look for anyone in the denominator who has one of the listed SNOMED CT codes recorded in either MASTER SNOMED CT PROCEDURE CODE or MASTER SNOMED CT FINDING CODE.

This is our numerator, the number of women who have a live or stillbirth in the 3 months prior to a given month, which involved an MCI or a postpartum haemorrhage of 1,500ml or more.

Slide 26

The final step is to create the CQIMDQ12 measure as a percentage

Slide 27

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 28

We will start by looking at the denominator for CQIMDQ13. The denominator for this DQ measure is 1.

Slide 29

Next, let’s look at how the numerator for CQIMDQ13 is calculated. We will be looking for all women who had been recorded with a postpartum haemorrhage in the previous 6 month reporting period. If there are women with that recorded, then the numerator will be 1; if there are none, then the numerator will be 0.

Slide 30

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We look for women with a SNOMED CT code representing postpartum haemorrhage recorded in MASTER SNOMED CT OBSERVATION CODE.

We confirm the records are in the 6 months prior to a given month, by using the CLINICAL INTERVENTION DATE (MOTHER) and the REPORTING PERIOD START DATE for each record.

If women have been recorded with a postpartum haemorrhage in the 6 months prior, then we set the numerator to one; otherwise we set the numerator to zero.

This is our numerator, representing whether women have been recorded with a postpartum haemorrhage in the 6 months prior to a given month.

Slide 31

The final step is to create the CQIMDQ13 measure as a percentage

Slide 32

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 33

Now, we look at the pass thresholds for the four DQ measures supporting CQIMPPH. A provider must pass all the data quality measures to be included in the published data. The provider’s rate for CQIMDQ10 must be 70% or more; the rate for CQIMDQ11 must be 60% or less; the rate for CQIMDQ12 must be 20% or less; and the rate for CQIMDQ13 must be 100%.

We will see how the values reached by the DQ measures affect what is shown in the published data.

Slide 34

Let’s look at the data published for October 2021.

RWJ – Stockport NHS Foundation Trust – has passed all four of the DQ measures associated with CQIMPPH. RWJ reached 117.1% for CQIMDQ10, 29.8% for CQIMDQ11, 3.9% for CQIMDQ12 and 100.0% for CQIMDQ13.

This means that CQIMPPH figures are published for RWJ, and this provider will contribute to the national and sub-national figures included in the October 2021 published data.

Slide 35

Now, we will look at another trust using the October 2021 data.

RLT – George Eliot Hospital NHS Trust – has passed 3 of the DQ measures associated with CQIMPPH. RLT reached 28.0% for CQIMDQ11, 1.3% for CQIMDQ12 and 100.0% for CQIMDQ13. It failed the DQ measure CQIMDQ10, with only 66.4%.

This means that CQIMPPH figures are not published for RLT. Instead, it is shown as ‘DNS’ or ‘Did Not Submit’ for CQIMPPH and will not contribute to the national and sub-national figures included in the October 2021 published data.

Slide 36

Let’s look at how applying these data quality measures affects the larger picture. 119 providers submitted data for October 2021 births. As the CQIM DQ measures span a time frame longer than just one month, 123 providers were considered for CQIMDQ10, 11, 12 and 13.

In October 2021:

111 providers passed CQIMDQ10, recording at least 70% for this DQ measure. 12 providers failed the measure or did not submit any data for this measure at all.

For both CQIMDQ11 and CQIMDQ12, 119 providers passed, while 4 providers did not submit any data for these measures.

82 providers passed CQIMDQ13, recording at least one postpartum haemorrhage in the 6 months prior to October 2021. 41 providers failed the measure, as they did not record any postpartum haemorrhages over this time frame.

Note that the pass rates of these measures could alter in the future.

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.

Slide 37

This brings us to the end of the video.

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, the percentage of women who had a postpartum haemorrhage of 1,500ml or more, are built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, for the proportion of babies born preterm. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month, the proportion of women who had a preterm birth.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for the number of live babies born whose gestational length was between 154 days and 315 days (22+0 and 45+0 weeks), in a given month.

For the numerator, it is very similar to the denominator. We will look for the number of live babies born whose gestational length was between 154 days and 258 days (22+0 and 36+6 weeks), in a given month.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All these data items can be found in the technical output specification version 2.0.25 available from the top link here.

There is also published meta data to explain how the measures are built. This is available by following the bottom link here.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at CQIMPreterm, the proportion of women who had a preterm birth.

The metadata that details the build for CQIMPreterm describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure

The timeframe for CQIMPreterm is a given month. For example, January 2021.

Slide 6

We want to be able to uniquely identify each woman, and we do this through data submitted to MSD401 baby demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER). This is primarily generated from the NHS Number.

We also use MSD401 baby demographics to get the following pieces of information:

• The PERSON BIRTH DATE (BABY) to ensure the birth is in the time frame
• GESTATION LENGTH (AT BIRTH) to ensure the gestation meets the conditions of the measure. This is measured in days.
• PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 7

We use the table MSD301 Labour Delivery to get BIRTHS PER LABOUR AND DELIVERY, a derived field. This is used to identify singleton births, with 1 meaning singleton.

Slide 8

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

Slide 9

We will start by looking at the denominator, the number of mothers in a given month, who gave birth to live singleton babies with a gestational length at birth between 154 and 315 days (22+0 and 45+0 weeks).

Let’s see which data items we use to do this.

Slide 10

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.

We next consider only the women with a pregnancy outcome of 01, meaning live birth.

Next, we consider only babies with a GESTATION LENGTH (AT BIRTH) that is between 154 and 315 days.

We then confirm the records are in a given month using the Person Birth Date (Baby) and the reporting period start date for each record.

The number of mothers in a given month, who gave birth to live singleton babies with a gestational length at birth between 154 and 315 days.

Slide 11

Next let’s look at how the numerator is calculated, the Number of mothers in a given month, who gave birth to live singleton babies with a gestational length at birth of 154 and 258 days (22+0 and 36+6 weeks)

Let’s see which data items we use to do this.

Slide 12

We use the cohort of babies we found for the denominator, then retain only those with a gestational length at birth of between 154 and 258 days. And that’s the numerator.

Slide 13

A few key points to take away are

• We take the latest data that is available for a given month
• The denominator counts the number of women who:
  • Had singleton live births
  • With a gestational length at birth between 154 days and 315 days

• The numerator counts how many of those had a gestational length at birth between 154 days and 258 days

Slide 14

The final step is to create the measure as a rate

Slide 15

We simply divide the numerator by the denominator and multiply by 1000 to get the rate per thousand.

Slide 16

CQIMPreterm is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ09 has a pass rate of greater than or equal to 70%.

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 17

The description of the CQIM DQ measures can be found on this slide, this information is also in the metadata file.

Slide 18

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for CQIMPreterm, the proportion of women who had a preterm birth is built.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for the proportion of babies born preterm. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, referred to as MSDS.

Slide 2

Today we will look at the 3 data quality measures that support the CQIM measure, CQIMPreterm, the proportion of babies born preterm.

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.

This video will look at the DQ measures supporting CQIMPreterm and show some examples of how these are currently displayed in the published data.

Slide 4

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built. This is available with each monthly data release, available by following the second link.

Many of the data items which appear in these DQ measures also exist in CQIMPreterm, the measure to which they are applied. A video describing the MSDS measure, CQIMPreterm, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5

There are 3 data quality measures that support CQIMPreterm, the proportion of babies born preterm. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.

The first DQ measure is CQIMDQ09. It calculates women giving birth, as a percentage of HES average monthly deliveries, in the current 1 month reporting period​.

The second is CQIMDQ22. It calculates the percentage of singleton babies with a valid gestational length at birth between 154 and 315 days (22 - 45 weeks) in the current 1 month reporting period. The third DQ measure is CQIMDQ23. It calculates the percentage of singleton babies with a gestational length at birth between 259 and 315 days (37 - 45 weeks) in the current 1 month reporting period.

The time frame for CQIMDQ09 and CQIMDQ22-23 is the given month only. For example, if the given month was January 2021, then all of these CQIM DQ measures would be counting women and babies in January 2021.

Now, we will look at all the fields in MSDS that underpin these measures.

Slide 6

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE.

Slide 7

We want to be able to uniquely identify each mother and baby, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER) to identify the mother, and one called PERSON ID (BABY) to identify the baby. These fields are primarily generated from the NHS number(s) submitted for mother and baby.

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers:

• The GESTATION LENGTH (AT BIRTH) to identify births at full term, and those babies with a valid gestation.
• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame.

Slide 8

We use the MSD301 Labour and Delivery table to identify births where a singleton baby was born.

To do this we use the field BIRTHS PER LABOUR AND DELIVERY, identifying singleton births where this field equals 1. This is a derived field that depends on the number of births submitted by providers under each LABOUR AND DELIVERY IDENTIFIER.

We will see how these items are used in the next slides.

Slide 9

We will start by looking at the denominator for CQIMDQ09, the number of deliveries in HES (using published 2018-19 data) pro-rata-ed over the number of days in a given month. Let’s see how we do this.

Slide 10

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link to that publication.

Then, we count the number of days in the reporting period, that is, in a given month. If the reporting period is January 2021, then this is the total number of days in January 2021.

We calculate the denominator as the number of days in the reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19.

This is our denominator, the pro-rata-ed number of deliveries in HES.

Slide 11

Next, let’s look at how the numerator for CQIMDQ09 is calculated. We will be looking for all women who gave birth in a given month. Let’s see which data items we use to do this.

Slide 12

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.

We confirm the records are in a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our numerator, the number of women who gave birth in a given month.

Slide 13

The final step is to create the CQIMDQ09 measure as a percentage.

Slide 14

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 15

We will start by looking at the denominator for CQIMDQ22, the number of singleton babies who were born in the current 1 month reporting period. Let’s see how we do this.

Slide 16

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We confirm the records are in a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of singleton babies in a given month.

Slide 17

Next, let’s look at how the numerator for CQIMDQ22 is calculated. We will be looking for all singleton babies born in the current 1 month reporting period, where the baby had a gestation length at birth between 154 and 315 days (between 22 and 45 weeks). Let’s see which data items we use to do this.

Slide 18

We use the cohort of babies we found for the denominator, then retain only those recorded with a valid gestation length at birth – that is, between 154 and 315 days. Again, we use PERSON ID (BABY) to count unique babies in the time frame.

To find babies with a valid gestation length at birth we look for anyone in the denominator who has a GESTATION LENGTH (AT BIRTH) between 154 and 315 days.

This is our numerator, the number of babies who have a singleton birth in a given month, and a gestation length of between 154 and 315 days (between 22 and 45 weeks) at birth.

Slide 19

The final step is to create the CQIMDQ22 measure as a percentage.

Slide 20

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 21

We will start by looking at the denominator for CQIMDQ23, the number of singleton babies who were born in the current 1 month reporting period. This denominator is the same as the denominator used for CQIMDQ22.

Slide 22

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We confirm the records are in a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of singleton babies in a given month.

Slide 23

Next, let’s look at how the numerator for CQIMDQ23 is calculated. We will be looking for all singleton babies born with a gestational length at birth between 259 and 315 days in the current 1 month reporting period.

Slide 24

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We keep only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days.

We confirm the records are in a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our numerator, the number of singleton babies born with a gestational length at birth between 259 and 315 days in a given month.

Slide 25

The final step is to create the CQIMDQ23 measure as a percentage.

Slide 26

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 27

Now, we look at the pass thresholds for the 3 DQ measures supporting CQIMPreterm. A provider must pass all the data quality measures to be included in the published data. The provider’s rates for CQIMDQ09, CQIMDQ22 and CQIMDQ23 must each be 70% or more.

If the denominator of CQIMDQ09 is 0 and the numerator is greater than 0, then the provider will pass that DQ measure.

We will see how the values reached by the DQ measures affect what is shown in the published data.

Slide 28

Let’s look at the data published for October 2021.

RCF – Airedale NHS Foundation Trust – has passed all 3 of the DQ measures associated with CQIMPreterm. RCF reached 91.4% for CQIMDQ09, 100.0% for CQIMDQ22, and 90.9% for CQIMDQ23.

This means that CQIMPreterm figures are published for RCF, and this provider will contribute to the national and sub national figures included in the October 2021 published data.

Slide 29

Now, we will look at another trust using the October 2021 data.

RKE – Whittington Health NHS Trust – has passed 2 of the DQ measures associated with CQIMPreterm. RKE reached 96.9% for CQIMDQ22 and 93.8% for CQIMDQ23. It failed the DQ measure CQIMDQ09 with 52.5%.

This means that CQIMPreterm figures are not published for RKE. Instead, the denominator and numerator are shown as ‘0’ and the Rate as ‘Low DQ’ for CQIMPreterm and will not contribute to the national and sub national figures included in the October 2021 published data.

Slide 30

Let’s look at how applying these data quality measures affects the larger picture. 119 providers submitted data for October 2021 births. As the CQIM DQ measures CQIMDQ09, CQIMDQ22 and CQIMDQ23 include providers that did not submit data, 123 providers were included in these measures.

In October 2021:

112 providers passed CQIMDQ09. 11 providers failed the measure or did not submit any data for this measure. 115 providers passed CQIMDQ22 and CQIMDQ23; each provider that passed CQIMDQ22 also passed CQIMDQ23.

Note that the pass rates of these measures could alter in the future.

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.

Slide 31

This brings us to the end of the video.

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, the proportion of babies born preterm, are built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, for caesarean section delivery rate in Robson group 1 women. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month and the two months prior, the percentage of women in Robson group 1 having a caesarean section.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for all women in Robson Group 1 who gave birth in a given month and the two months prior.

For the numerator, we take the women in the denominator and retain only those where the mother had a caesarean section.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, available by following the bottom link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at the Caesarean section delivery rate for Robson group 1 women.

The metadata that details the build for CQIMRobson01 describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure

The timeframe for CQIMRobson01 is a given month and the two months prior. For example, if the given month was January 2021, then the measure would be counting women in January 2021, December 2020 and November 2020.

Slide 6

We want to be able to uniquely identify each mother, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called Person ID (mother). This is primarily generated from the NHS Number.

We also use MSD401 baby demographics to get the following 4 pieces of information:

• The Person Birth Date (Baby) to ensure the child is born in the relevant time frame
• Presentation of fetus at onset of labour or delivery to ensure the presentation is cephalic, a condition of Robson group 1, with 1 meaning cephalic.
• Gestation Length (at birth) to ensure gestation is between 37 and 45 weeks (so between 259 and 315 days), which is a condition of Robson group 1
• Delivery Method Code to ensure delivery is by caesarean section, with ‘7’ or ‘8’ meaning caesarean section

Slide 7

We use the table MSD101 Pregnancy and Booking Details to get 3 pieces of information:

• Pregnancy Total Previous caesarean sections
• Pregnancy total previous live births
• and pregnancy total previous still births
• These are all used to ensure a woman had no previous births, a condition of Robson group 1.

Slide 8

• We use MSD301 Labour Delivery to get 3 pieces of information:
• LABOUR OR DELIVERY ONSET METHOD CODE to ensure the onset of labour a is non-operative vaginal birth, a condition of Robson group 1, with 1 meaning non-operative vaginal birth
• ONSET OF ESTABLISHED LABOUR DATE
• and PROCEDURE DATE (CAESAREAN SECTION) to ensure the caesarean is after the onset of labour

Slide 9

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

We will see how these items are used in the next slides

Slide 10

We will start by looking at the denominator, all women in Robson Group 1 in a given month and the two months prior. Let’s see which data items we use to do this.

Slide 11

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next consider only the mothers who are in Robson Group 1. The conditions for Robson Group 1 are confirmed using the following fields:

• A mother’s PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS and PREGNANCY TOTAL PREVIOUS STILLBIRTHS should be equal to zero or null, with at least one of them being not null.
• PRESENTATION OF FETUS AT ONSET OF LABOUR OR DELIVERY is ‘01’, indicating it was cephalic
• GESTATION LENGTH (AT BIRTH) is greater than 37 weeks
• LABOUR OR DELIVERY ONSET METHOD CODE is ‘1’, indicating it was a non-operative vaginal birth or spontaneous
• The recorded ONSET OF ESTABLISHED LABOUR DATE is greater than the PROCEDURE DATE (CAESAREAN SECTION) or with the PROCEDURE DATE (CAESAREAN SECTION) being null. This does not include the time of day, only the date. Where the two dates are the same, the mother would not be counted.

We confirm the records are in a given month and two months prior using the Person Birth Date (Baby) and the reporting period start date for each record.

This is our denominator, all the women in Robson Group 1 who gave birth in a given month and two months prior.

Slide 12

Next let’s look at how the numerator is calculated. We will be looking for all women in Robson Group 1 who gave birth in a given month and the two months prior who have had a caesarean section. Let’s see which data items we use to do this.

Slide 13

We use the cohort of women we found for the denominator, then retain only those where DELIVERY METHOD CODE is 7 or 8, indicating a caesarean. Again, we use PERSON ID (MOTHER) to count unique women in the time frame.

Slide 14

A few key points to take away are:

• We take the latest data that is available for a given month and the two months prior to it.
• The denominator counts women who are in Robson Group 1.
• The numerator counts how many of those women had a caesarean section.
• Note, if the PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS and PREGNANCY TOTAL PREVIOUS STILLBIRTHS fields are all nulls and not zeroes, the record will not be counted for this measure.

Slide 15

The final step is to create the measure as a percentage,

Slide 16

we simply divide the numerator by the denominator and multiply by 100 to get a percentage

Slide 17

CQIMRobson01 is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ30 has a pass rate of greater than or equal to 70%

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 18

The description of the CQIM DQ measures can be found on this slide, this information is all in the metadata file.

Slide 19

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for Caesarean section delivery rate in Robson group 1 women is built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, for caesarean section delivery rate in Robson group 2 women. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month and the two months prior, the percentage of women in Robson group 2 having a caesarean section.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for all women in Robson group 2 who gave birth in a given month and the two months prior.

For the numerator, we take the women in the denominator and retain only those where the mother had a caesarean section.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, available by following the bottom link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at the Caesarean section delivery rate in Robson group 2 women.

The metadata that details the build for CQIMRobson02 describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure

The timeframe for CQIMRobson02 is a given month and the two months prior. For example, if the given month was January 2021, then the measure would be counting women in January 2021, December 2020 and November 2020.

Slide 6

We want to be able to uniquely identify each mother, we do this through data submitted to MSD401 baby demographics. From the data submitted by all providers, NHS Digital creates a field called Person ID (mother). This is primarily generated from the NHS Number.

We also use MSD401 baby demographics to get the following 4 pieces of information:

• The Person Birth Date (Baby) to ensure the child is born in the relevant time frame
• Presentation of fetus at onset of labour or delivery to ensure the presentation is cephalic, a condition of Robson group 2, with 1 meaning cephalic.
• Gestation Length (at birth) to ensure gestation is between 37 and 45 weeks (so between 259 and 315 days), which is a condition of Robson group 2.
• Delivery Method Code to ensure delivery is by caesarean section, with ‘7’ or ‘8’ meaning caesarean section

Slide 7

We use the table MSD101 Pregnancy and Booking Details to get 3 pieces of information:

• Pregnancy total previous caesarean sections
• Pregnancy total previous live births
• and pregnancy total previous still births

These are all used to ensure no previous births, a condition of Robson group 2.

Slide 8

We use MSD301 Labour Delivery to get 3 pieces of information:

• LABOUR OR DELIVERY ONSET METHOD CODE to ensure the onset of labour meets the conditions of Robson group 2, with 2 meaning caesarean section and 3,4 and 5 meaning induced labour.
• ONSET OF ESTABLISHED LABOUR DATE
• and PROCEDURE DATE (CAESAREAN SECTION) to ensure the caesarean is after the onset of labour

Slide 9

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

We will see how these items are used in the next slides

Slide 10

We will start by looking at the denominator, all women in Robson group 2 who give birth in a given month and the two months prior. Let’s see which data items we use to do this.

Slide 11

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next consider only the mothers who are in Robson group 2. The conditions for Robson group 2 are confirmed using the following fields:

• A mother’s PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS and PREGNANCY TOTAL PREVIOUS STILLBIRTHS should be equal to zero or null, with at least one of them being not null.
• PRESENTATION OF FETUS AT ONSET OF LABOUR OR DELIVERY is ‘01’, indicating it was cephalic
• GESTATION LENGTH (AT BIRTH) is between 37 weeks and 45 weeks (or between 259 and 315 days)
• And one of the following:

• LABOUR OR DELIVERY ONSET METHOD CODE is any of ‘2’ (caesarean), ‘3’, ‘4’ ,‘5’ (induction)
• Or the recorded ONSET OF ESTABLISHED LABOUR DATE is null and PROCEDURE DATE (CAESAREAN SECTION) is not null.

We confirm the records are in a given month and two months prior using the Person Birth Date (Baby) and the reporting period start date for each record.

This is our denominator, all the women in Robson group 2 who gave birth in a given month and two months prior.

Slide 12

Next let’s look at how the numerator is calculated. We will be looking for all women in Robson group 2 in a given month and the two months prior who have had a caesarean section. Let’s see which data items we use to do this.

Slide 13

We use the cohort of women we found for the denominator, then retain only those where DELIVERY METHOD CODE is 7 or 8, indicating a caesarean. Again, we use PERSON ID (MOTHER) to count unique women in the time frame.

Slide 14

• A few key points to take away are
• We take the latest data that is available for a given month and the two months prior to it.
• The denominator counts women who are in Robson group 2.
• The numerator counts how many of those women had a caesarean section.
• Note, if the PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS and PREGNANCY TOTAL PREVIOUS STILLBIRTHS fields are all nulls and not zeroes, the record will not be counted for this measure.

Slide 15

The final step is to create the measure as a percentage,

Slide 16

we simply divide the numerator by the denominator and multiply by 100 to get a percentage

Slide 17

CQIMRobson02 is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ30 has a pass rate of greater than or equal to 70%

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 18

The description of the CQIM DQ measures can be found on this slide, this information is all in the metadata file.

Slide 19

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for Caesarean section delivery rate in Robson group 2 women is built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, caesarean section delivery rate in Robson Group 5 women. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month and the two months prior, the percentage of women in Robson group 5 having a caesarean section.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for all women in Robson group 5 who gave birth in a given month and the two months prior.

For the numerator, we take the women in the denominator and retain only those where the mother had a caesarean section.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, available by following the bottom link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at the Caesarean section delivery rate in Robson group 5 women.

The metadata that details the build for CQIMRobson05 describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure

The timeframe for CQIMRobson05 is a given month and the two months prior. For example, if the given month was January 2021, then the measure would be counting women in January 2021, December 2020 and November 2020.

Slide 6

We want to be able to uniquely identify each mother, we do this through data submitted to MSD401 baby demographics. From the data submitted by all providers, NHS Digital creates a field called Person ID (mother).  This is primarily generated from the NHS Number.

We also use MSD401 Baby Demographics to get the following 4 pieces of information:

• The Person Birth Date (Baby) to ensure the child is born in the relevant time frame
• Presentation of fetus at onset of labour or delivery to ensure the presentation is cephalic, a condition of Robson group 5, with 1 meaning cephalic.
• Gestation Length (at birth) to ensure gestation is between 37 weeks and 45 weeks (so between 259 and 315 days), which is a condition of Robson group 5.
• Delivery Method Code to ensure delivery is by caesarean section, with ‘7’ or ‘8’ meaning caesarean section

Slide 7

We use the table MSD101 Pregnancy and Booking Details to get Pregnancy Total Previous caesarean sections, used to ensure the mother had a previous caesarean section, a condition of Robson group 5

Slide 8

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

We will see how these items are used in the next slides

Slide 9

We will start by looking at the denominator, all women in Robson group 5 who gave birth in a given month and the two months prior. Let’s see which data items we use to do this.

Slide 10

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next consider only the mothers who are in Robson group 5. The conditions for Robson group 5 are confirmed using the following fields:

• A mother’s PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS
  • Is greater than or equal to 1
  • And less than or equal to 20, in order to remove outliers with an unusually high number of caesarean sections

• PRESENTATION OF FETUS AT ONSET OF LABOUR OR DELIVERY is ‘01’, indicating it was cephalic
• GESTATION LENGTH (AT BIRTH) is between 37 weeks and 45 weeks (between 259 and 315 days).

Next, we confirm the records are in a given month and two months prior using the Person Birth Date (Baby) and the reporting period start date for each record.

This is our denominator, all the women in Robson group 5 who gave birth in a given month and two months prior.

Slide 11

Next let’s look at how the numerator is calculated. We will be looking for all women in Robson group 5 in a given month and the two months prior who have had a caesarean section. Let’s see which data items we use to do this.

Slide 12

We use the cohort of women we found for the denominator, then retain only those where DELIVERY METHOD CODE is 7 or 8, indicating a caesarean. Again, we use PERSON ID (MOTHER) to count unique women in the time frame.

Slide 13

A few key points to take away are

• We take the latest data that is available for a given month and the two months prior to it.
• The denominator counts women who are in Robson group 5.
• The numerator counts how many of those women had a caesarean section.

Slide 14

The final step is to create the measure as a percentage,

Slide 15

 we simply divide the numerator by the denominator and multiply by 100 to get a percentage

Slide 16

CQIMRobson05 is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ30 has a pass rate of greater than or equal to 70%

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 17

The description of the CQIM DQ measures can be found on this slide, this information is all in the metadata file.

Slide 18

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for Caesarean section delivery rate in Robson group 5 women is built from MSDS

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1  

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metrics, also known as CQIMs, for caesarean section delivery rates in Robson group 1, 2, and 5 women. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS. 

Slide 2  

Today we will look at the nine data quality measures that support the three CQIM measures CQIMRobson01, CQIMRobson02, and CQIMRobson05 – respectively, the caesarean section delivery rates in Robson group 1, 2 and 5 women. 

Slide 3  

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly: 

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping 

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.  

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.  

This video will look at the DQ measures supporting CQIMRobson01, CQIMRobson02, and CQIMRobson05 and show some examples of how these are currently displayed in the published data.   

Slide 4  

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built.  Links to the latest metadata are given alongside each monthly data release, and is also available by following the second link.

Many of the data items which appear in these DQ measures also exist in CQIMRobson01, CQIMRobson02, and CQIMRobson05, the measures to which they are applied. Videos describing the MSDS measures, CQIMRobson01, CQIMRobson02, and CQIMRobson05, and how they are built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5  

There are nine data quality measures that support the Robson group CQIMs, the caesarean section delivery rates in Robson group 1, 2 and 5 women. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.  

The first DQ measure is CQIMDQ30. It calculates women giving birth in MSDS, as a percentage of HES average monthly deliveries, in the current 3 month reporting period​.  

The second is CQIMDQ31. It calculates the percentage of babies with a valid gestational age at birth between 141 and 315 days (20+1 – 45+0 weeks) in the current 3 month reporting period.  

Slide 6  

The third DQ measure is CQIMDQ32. It calculates the percentage of babies with a gestational age at birth between 259 and 315 days (37 - 45 weeks) in the current 3 month reporting period.  

The fourth is CQIMDQ33. It calculates the percentage of babies born with a valid delivery method, in the current 3 month reporting period. CQIMDQ34 is the fifth DQ measure for the CQIM Robson metrics. It calculates the percentage of babies born vaginally in the current 3 month reporting period.  

The timeframe for CQIMDQ30-34 and CQIMDQ36-39 is the given month and the 2 months prior. For example, if the given month was January 2021, then CQIMDQ30-34 and CQIMDQ36-39 would be counting women and babies in November 2020 to January 2021. 

Slide 7  

The sixth DQ measure is CQIMDQ36. It calculates the percentage of women with valid data recorded about previous caesarean sections, live and stillbirths, who gave birth in the current 3 month reporting period. The seventh is CQIMDQ37. It calculates the percentage of women with no previous caesarean sections, live and stillbirths, who gave birth in the current 3 month reporting period.  

The eighth DQ measure is CQIMDQ38. It calculates the percentage of babies with fetus presentation recorded, at births in the current 3 month reporting period. The ninth is CQIMDQ39. It calculates the percentage of women where the labour onset method was recorded, that gave birth in the current 3 month reporting period.  

Now, we will look at all the fields in MSDS that underpin these measures. 

Slide 8  

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE 

Slide 9  

We want to be able to uniquely identify each mother and baby, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER) to identify the mother, and one called PERSON ID (BABY) to identify the baby. These fields are primarily generated from the NHS Number(s) submitted for mother and baby. 

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers: 

• The DELIVERY METHOD CODE to identify how the baby is born, specifically if the delivery was vaginal. 

• The GESTATION LENGTH (AT BIRTH) to identify births at full term, and those babies with a valid gestation.  
• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame. 
• The PRESENTATION OF FETUS AT ONSET OF LABOUR OR DELIVERY to identify births where the fetus presentation was known at the start of labour or delivery. 

Slide 10  

We use the MSD301 Labour and Delivery table to identify births where a valid labour onset method has been recorded - in the field LABOUR OR DELIVERY ONSET METHOD CODE. This field is submitted by the provider. 

Slide 11  

We use the MSD101 Pregnancy and Booking Details table to identify women with a booking appointment, and to find information about their birth history.  

We use these fields: 

• The PERSON ID (MOTHER) identifies the mother who attended the booking appointment. This field is primarily derived from the NHS Number for the mother which is submitted by the provider. 
• The PREGNANCY TOTAL PREVIOUS LIVE BIRTHS, PREGNANCY TOTAL PREVIOUS STILLBIRTHS and PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS are fields that note the number of caesarean sections, live and stillbirths which have resulted from previous pregnancies. This information, submitted by the provider, will be used to find pregnancies with no history of previous births.   

We will see how these items are used in the next slides. 

Slide 12  

We will start by looking at the denominator for CQIMDQ30, the number of deliveries in HES (using published 2018-19 data) pro-rata-ed over the number of days in a given month and the 2 months prior. Let’s see how we do this. 

Slide 13  

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link here.

Then, we count the number of days in the reporting period, that is in a given month and the 2 months prior. If the reporting period is January 2021, then this is the total number of days in November 2020, December 2020 and January 2021. 

We calculate the denominator as the number of days in the reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19. 

This is the denominator, the pro-rata-ed number of deliveries in HES. 

Slide 14  

Next, let’s look at how the numerator for CQIMDQ30 is calculated. We will be looking for all women who gave birth in a given month or the 2 months prior. Let’s see which data items we use to do this. 

Slide 15  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of women who gave birth in a given month or the 2 months prior. 

Slide 16  

The final step is to create the CQIMDQ30 measure as a percentage 

Slide 17

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 18  

We will start by looking at the denominator for CQIMDQ31, the number of babies who were born in the current 3 month reporting period. Let’s see how we do this. 

Slide 19  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born in a given month or the 2 months prior. 

Slide 20  

Next, let’s look at how the numerator for CQIMDQ31 is calculated. We will be looking for all babies born in the current 3 month reporting period, where the baby had a gestational age at birth between 141 and 315 days (between 20+1 and 45+0 weeks). Let’s see which data items we use to do this. 

Slide 21  

We use the cohort of babies we found for the denominator, then retain only those recorded with a gestational age at birth between 141 and 315 days. Again, we use PERSON ID (BABY) to count unique babies in the time frame. 

To find babies with a valid gestation at birth we look for anyone in the denominator who has a GESTATION LENGTH (AT BIRTH) between 141 and 315 days. 

This is our numerator, the number of babies who were born in a given month or the 2 months prior, and had a gestational age of between 141 and 315 days (between 20+1 and 45+0 weeks) at birth. 

Slide 22   

The final step is to create the CQIMDQ31 measure as a percentage 

Slide 23

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 24  

We will start by looking at the denominator for CQIMDQ32, the number of babies born with a gestational age at birth between 141 and 315 days in the current 3 month reporting period. This denominator is the same as the numerator used for CQIMDQ31.  

Slide 25  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.  

We keep only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 141 and 315 days. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born with a gestational age at birth between 141 and 315 days, in a given month or the 2 months prior. 

Slide 26  

Next, let’s look at how the numerator for CQIMDQ32 is calculated. We will be looking for babies born with a gestational age at birth between 259 and 315 days in the current 3 month reporting period.  

Slide 27  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. To ensure births are at full term, we keep only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born with a gestational age at birth between 259 and 315 days, in a given month or the 2 months prior. 

Slide 28  

The final step is to create the CQIMDQ32 measure as a percentage. 

Slide 29

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 30  

We will start by looking at the denominator for CQIMDQ33, the number of babies who were born in the current 3 month reporting period.  

Slide 31  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born in a given month or the 2 months prior. 

Slide 32  

Next, let’s look at how the numerator for CQIMDQ33 is calculated. We will be looking for babies born in the current 3 month reporting period where the method of delivery was recorded with a valid code. Let’s see which data items we use to do this. 

Slide 33  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. 

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’, ’2’, ’3’, ‘4’, ‘5’, ‘6’, ‘7’, ‘8’, or ’9’ indicating a validly recorded method of delivery. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born with a valid delivery method recorded in a given month or the 2 months prior. 

Slide 34  

The final step is to create the CQIMDQ33 measure as a percentage 

Slide 35

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 36  

We will start by looking at the denominator for CQIMDQ34, the number of babies who were born in the current 3 month reporting period, with a valid delivery method. This denominator is the same as the numerator used for CQIMDQ33.  

Slide 37  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.  

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’, ’2’, ’3’, ’4’, ’5’, ’6’, ’7’, ’8’ or ‘9’. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born with a valid delivery method in a given month or the 2 months prior. 

Slide 38  

Next, let’s look at how the numerator for CQIMDQ34 is calculated. We will be looking for babies born vaginally (excluding breech) in the current 3 month reporting period.  

Slide 39  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’, ‘2’, ‘3’ or ’4’ to identify a vaginal delivery (excluding breech). 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of babies born with a vaginal delivery method (excluding breech) in a given month or the 2 months prior. 

Slide 40  

The final step is to create the CQIMDQ34 measure as a percentage. 

Slide 41

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 42  

We will start by looking at the denominator for CQIMDQ36, the number of women who gave birth in the current 3 month reporting period. Let’s see how we do this. 

Slide 43  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women who gave birth in a given month or the 2 months prior. 

Slide 44  

Next, let’s look at how the numerator for CQIMDQ36 is calculated. We will be looking for women who gave birth in the current 3 month reporting period, where valid information had been recorded about the woman’s history of caesarean sections, live and stillbirths. Let’s see which data items we use to do this. 

Slide 45  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We look for women who have PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS recorded between 0 and 20, or who have no value recorded in this field; similarly, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS must be recorded with a value between 0 and 20 or have no value recorded in this field; and finally, PREGNANCY TOTAL PREVIOUS STILLBIRTHS must be recorded with a value between 0 and 20 or have no value recorded in this field. 

We then keep only women who have a value recorded in at least one of PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS, and PREGNANCY TOTAL PREVIOUS STILLBIRTHS, removing any women who have no values recorded in these three fields. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women who gave birth in a given month or the 2 months prior, where their history of caesarean sections, live and stillbirths was recorded.  

Slide 46  

The final step is to create the CQIMDQ36 measure as a percentage. 

Slide 47

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 48  

We will start by looking at the denominator for CQIMDQ37, the number of women who gave birth in the current 3 month reporting period, where their history of caesarean sections, live and stillbirths was recorded. This denominator is the same as the numerator for CQIMDQ36. 

Slide 49  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We look for women who have PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS recorded between 0 and 20, or who have no value recorded in this field; similarly, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS must be recorded with a value between 0 and 20 or have no value recorded in this field; and finally, PREGNANCY TOTAL PREVIOUS STILLBIRTHS must be recorded with a value between 0 and 20 or have no value recorded in this field. 

We then keep only women who have a value recorded in at least one of PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS, and PREGNANCY TOTAL PREVIOUS STILLBIRTHS, removing any women who have no values recorded in these three fields. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women who gave birth in a given month or the 2 months prior, where their history of caesarean sections, live and stillbirths was recorded. 

Slide 50  

Next, let’s look at how the numerator for CQIMDQ37 is calculated. We will be looking for women who gave birth in the current 3 month reporting period, where their history of caesarean sections, live and stillbirths showed that they had had zero previous births. 

Slide 51  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We look for women who have 0 PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS recorded, or who have no value recorded in this field; women must also have 0 PREGNANCY TOTAL PREVIOUS LIVE BIRTHS recorded, or have no value recorded in this field; and they must also have either 0 PREGNANCY TOTAL PREVIOUS STILLBIRTHS recorded, or have no value recorded in this field. 

Of those women, we then remove anyone who has no values recorded in any of the fields PREGNANCY TOTAL PREVIOUS CAESAREAN SECTIONS, PREGNANCY TOTAL PREVIOUS LIVE BIRTHS, and PREGNANCY TOTAL PREVIOUS STILLBIRTHS. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of women who gave birth in a given month or the 2 months prior, with zero previous births recorded in their history of caesarean sections, live and stillbirths. 

Slide 52  

The final step is to create the CQIMDQ37 measure as a percentage. 

Slide 53

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 54  

We will start by looking at the denominator for CQIMDQ38, the number of babies who were born in the current 3 month reporting period.  

Slide 55  

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of babies born in a given month or the 2 months prior. 

Slide 56  

Next, let’s look at how the numerator for CQIMDQ38 is calculated. We will be looking for all babies born in the current 3 month reporting period, where the presentation of the fetus at the onset of labour or delivery had been recorded. Let’s see which data items we use to do this. 

Slide 57  

We use the cohort of babies we found for the denominator, then retain only those recorded with valid presentation of fetus at labour or delivery onset. We continue to use PERSON ID (BABY) to count unique babies in the time frame. We look for any baby in the denominator who has been recorded with a PRESENTATION OF FETUS AT ONSET OF LABOUR OR DELIVERY of ‘01’, ‘02’, ‘03’, ‘04’ or ‘XX’. 

This is our numerator, the number of babies who were born in a given month or the 2 months prior whose presentation at the onset of labour or delivery was recorded. 

Slide 58   

The final step is to create the CQIMDQ38 measure as a percentage. 

Slide 59

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 60  

We will start by looking at the denominator for CQIMDQ39, the number of women who gave birth in the current 3 month reporting period. Let’s see how we do this. 

Slide 61  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women who gave birth in a given month or the 2 months prior. 

Slide 62  

Next, let’s look at how the numerator for CQIMDQ39 is calculated. We will be looking for women who gave birth in the current 3 month reporting period, where valid information had been recorded about the labour onset method for the delivery. Let’s see which data items we use to do this. 

Slide 63  

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We keep women who have a LABOUR OR DELIVERY ONSET METHOD CODE recorded with one of ‘1’, ‘2’, ‘3’, ‘4’ or ‘5’ to identify women with deliveries where the method of labour onset was known.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women who gave birth in a given month or the 2 months prior, where their method of labour onset was known and recorded.  

Slide 64  

The final step is to create the CQIMDQ39 measure as a percentage. 

Slide 65

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 66

Now, we look at the pass thresholds for the 9 DQ measures supporting CQIMRobson01, CQIMRobson02, and CQIMRobson05. A provider must pass all the data quality measures to be included in the published data. The provider’s rates for CQIMDQ30-33 must be 70% or more.  

If the denominator of CQIMDQ30 is 0 and the numerator is greater than 0, then the provider will pass the DQ measure. 

Slide 67

The provider rate for CQIMDQ34 must be 40% or more and the rate for CQIMDQ36 must be 70% or more. 

Slide 68

The provider rate for CQIMDQ37 must be between 20% and 70% inclusive, and the rates for CQIMDQ38-39 must both be 70% or more. 

We will see how the values reached by the DQ measures affect what is shown in the published data. 

Slide 69

Let’s look at the data published for October 2021. 

RDU – Frimley Health NHS Foundation Trust – has passed all 9 of the DQ measures associated with CQIMRobson01, CQIMRobson02, and CQIMRobson05. RDU reached 103.2% for CQIMDQ30, 100.0% for CQIMDQ31, 92.8% for CQIMDQ32, 100.0% for CQIMDQ33, 63.2% for CQIMDQ34, 92.8% for CQIMDQ36, 39.3% for CQIMDQ37, 100.0% for CQIMDQ38 and 99.4% for CQIMDQ39. 

This means that CQIMRobson01, CQIMRobson02, and CQIMRobson05 figures are published for RDU, and this provider will contribute to the national and sub national figures included in the October 2021 published data. 

Slide 70

Now, we will look at another trust using the October 2021 data. 

R1H – Barts Health NHS Trust – has passed 5 of the DQ measures associated with CQIMRobson01, CQIMRobson02, and CQIMRobson05. R1H reached 86.9% for CQIMDQ30, 93.0% for CQIMDQ32, 100.0% for CQIMDQ33, 75.3% for CQIMDQ34, and 94.8% for CQIMDQ38. It failed the DQ measures CQIMDQ31, with 63.6%; CQIMDQ36, with 60.8%; CQIMDQ37, with 14.8%; and CQIMDQ39 with 0.0%. 

This means that CQIMRobson01, CQIMRobson02, and CQIMRobson05 figures are not published for R1H. Instead, the denominator and numerator are shown as ‘0’ and the rate as ‘Low DQ’ for each metric and the provider will not contribute to the national and sub national figures included in the October 2021 published data. 

Slide 71

Let’s look at how applying these data quality measures affects the larger picture. 119 providers submitted data for October 2021 births. As the CQIM DQ measures span a time frame longer than just one month, 123 providers were considered for the DQ measures CQIMDQ30-34 and CQIMDQ36-39. 

In October 2021: 

115 providers passed CQIMDQ30, and the same number of providers passed CQIMDQ34. 8 providers failed each of these measures or did not submit any data. 113 providers passed CQIMDQ31, and 117 providers passed CQIMDQ33.  

CQIMDQ32 and CQIMDQ38 were each passed by 118 providers. 5 providers did not submit any data to CQIMDQ32. 

Only 104 providers passed CQIMDQ39, and 106 providers passed CQIMDQ37. 11 providers submitted data with too high or too low a rate to pass CQIMDQ37, and 6 providers did not submit any data for this measure. 

Note that the pass rates of these measures could alter in the future. 

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.  

Slide 72

This brings us to the end of the video. 

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rates, the caesarean section delivery rates in Robson group 1, 2 and 5 women, are built from MSDS 

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, of women who were current smokers at the time of their booking appointment. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that finds, for any given month, the proportion of women with a booking appointment who were current smokers at the time of that appointment.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look at the number of women with a booking appointment in a given month that had a known smoking status recorded within three days of that appointment.

For the numerator, we keep only those women from the denominator whose recorded smoking status identifies them as a current smoker.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25, which can be found at the first link here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, and can be found at the second link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at Women who were current smokers at the time of their booking appointment ‘CQIMSmokingBooking’.

The metadata that details the build for CQIMSmokingBooking describes the measure, the numerator, and the denominator.

The timeframe for CQIMSmokingBooking is a given month. For example, January 2021, which would count women with a booking appointment in that month.

Slide 6

We will start by looking at all the fields in MSDS that underpin this measure.

We use MSD000 Header to establish the reporting period and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE.

Slide 7

In order to identify the woman, and the date of the booking appointment, we use the table MSD101 Pregnancy and Booking Details. The woman is identified by PERSON ID (MOTHER) which is derived from demographic information, including NHS Number, that is submitted by the provider to MSDS in table MSD001 Mother's Demographics. The booking appointment date is taken from APPOINTMENT DATE (FORMAL ANTENATAL BOOKING).

Slide 8

In order to determine when a smoking status was recorded for a woman, and whether it took place at the time of the booking appointment, we use the table MSD201 Care Contact (Pregnancy) and look for the CARE CONTACT DATE.

Slide 9

We use the table MSD202 Care Activity (Pregnancy) to determine whether information has been recorded which can determine the woman’s smoking status. This information may be submitted directly by the provider as codes describing findings or observations made about the woman’s smoking status in CODED FINDING (CODED CLINICAL ENTRY) or CODED OBSERVATION (CLINICAL TERMINOLOGY).

Slide 10

Information about the woman’s smoking status may also be taken from the fields CIGARETTES PER DAY and CO MONITORING READING. These fields derive information about the woman’s cigarette consumption and carbon monoxide levels from information submitted by providers about observations that they have made.

We will see how these items are used in the next slides

Slide 11

We will start by looking at the denominator, the number of women with a booking appointment in a given month who have a recorded smoking status.

Let’s see which data items we use to do this.

Slide 12

We use the derived field PERSON ID (MOTHER) to count unique women during the reporting month.

First, we look for women who have a booking appointment in the reporting month by checking whether both the REPORTING PERIOD START DATE and APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) are between the start and end dates of the reporting month.

Then, we compare the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and CARE CONTACT DATE to include only women with care contacts within plus or minus three days of the booking appointment.

Finally, we look at activity recorded at those care contacts to find women with information recorded about their smoking status. That is, either CIGARETTES PER DAY is recorded as zero or greater; or CO MONITORING READING is recorded as zero or greater, or a code from a specific list of codes related to smoking status has been recorded in either CODED FINDING (CODED CLINICAL ENTRY) or CODED OBSERVATION (CLINICAL TERMINOLOGY).

This is our denominator, the number of women with a recorded smoking status at booking appointment.

Slide 13

This is the list of codes which are currently used to determine smoking status if recorded in CODED FINDING (CODED CLINICAL ENTRY) or CODED OBSERVATION (CLINICAL TERMINOLOGY). Information about smoking status can be recorded using SNOMED, ICD, READ or CTV3 codes.

Slide 14

Next, we will look at the numerator, the number of women with a booking appointment in a given month who have been recorded with a smoking status of SMOKER. This is a subset of the women identified for the denominator – they have a booking appointment in the reporting month, and a known smoking status recorded within three days, but with the additional condition that that smoking status identifies them as a current smoker.

Let’s see which data items we use to do this. 

Slide 15

We use the derived field PERSON ID (MOTHER) to count unique women during the reporting month.

To identify women with a booking appointment in the reporting month who have care contacts within plus or minus three days of their booking appointment, we follow the same process that we used to identify women in the denominator.

We check that both the REPORTING PERIOD START DATE and APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) are between the start and end dates of the reporting month.  And then we compare the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and CARE CONTACT DATE, so women are only included if their care contacts are within three days of their booking.

To identify women for the numerator, we look at the activity recorded at those care contacts to find women where the information recorded about their smoking status indicates that they are a current smoker. That is, either CIGARETTES PER DAY is recorded as greater than zero; or CO MONITORING READING is recorded as greater than or equal to four parts per million, or a code specifically describing them as a smoker has been recorded in CODED FINDING (CODED CLINICAL ENTRY).

This is our numerator, the number of women who are recorded as current smokers at booking appointment.

Slide 16

The final step is to create the measure as a percentage

Slide 17

To do this, we divide the numerator by the denominator. We then multiply the result by one hundred to get a percentage.

Slide 18

 A few key points to take away are

• Women may have more than one smoking status recorded within plus or minus three days of the booking appointment. We will take the smoking status as recorded on the day of the booking appointment over a smoking status recorded on another day. Otherwise, we will take the latest data related to their smoking status.
• The care contact, and information about the woman’s smoking status, can be recorded at a different provider than the provider that submitted information about the booking appointment.
• If no information has been recorded about a woman’s smoking status, or if the provider has only recorded information about the smoking status in another table in MSDS, then that woman will not be included in CQIMSmokingBooking.

Slide 19

CQIMSmokingBooking is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

The pass rate for CQIMDQ03 is 50 percent, and the pass rate for CQIMDQ04 is 70 percent. In order to pass CQIMDQ05 the rate for that measure must be between 0.5 and 50 percent.

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 20

The description of the CQIM DQ measures that support CQIMSmokingBooking can be found on this slide; this information is all in the metadata file.

Slide 21

This brings us to the end of the video.

Thank you for watching this video demonstration on how the CQIM measure, women who were current smokers at the time of their booking appointment, is built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1  

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for the proportion of women who were current smokers at their booking appointment. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS. 

Slide 2  

Today we will look at the 3 data quality measures that support the CQIM measure, CQIMSmokingBooking, the proportion of women who were current smokers at their booking appointment. 

Slide 3  

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly: 

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping 

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.  

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.  

This video will look at the DQ measures supporting CQIMSmokingBooking and show some examples of how these are currently displayed in the published data.  

Slide 4  

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.  

There is also published metadata to explain how the measures are built. Links to the latest metadata are given alongside each monthly data release, available by following the second link. 

Many of the data items which appear in these DQ measures also exist in CQIMSmokingBooking, the measure to which they are applied. A video describing the MSDS measure, CQIMSmokingBooking, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link. 

Slide 5   

There are 3 data quality measures that support CQIMSmokingBooking, the proportion of women who were current smokers at their booking appointment. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.  

The first DQ measure is CQIMDQ03. It calculates women booking in MSDS in the reporting period as a percentage of HES average monthly deliveries.  

The second is CQIMDQ04. It calculates the percentage of women who had a booking appointment in the reporting period whose smoking status was known.  

The third is CQIMDQ05. It calculates the percentage of women who had a booking appointment in the reporting period, whose smoking status was recorded as known, that they were current smokers. 

The timeframe for these 3 CQIM DQ metrics is the current reporting month. For example, if the reporting period was April 2022, then the women who had a booking appointment in April 2022 would be counted. 

Slide 6  

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE. 

Slide 7  

We want to be able to uniquely identify each mother. We do this through the data item,  PERSON ID (MOTHER), taken from the pregnancy booking table, MSD101 Pregnancy and Booking Details. This data item is sourced from the mother’s NHS number, or a combination of other data items if the NHS number is not valid. 

One mandatory submitted field, the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING), is also gained from the pregnancy booking table. This is used to identify the date that the woman booked into maternity services. 

Slide 8  

We want to be able to identify the date at which a care activity took place during the mother’s pregnancy. This information is gathered from the MSD201 Care Contact (Pregnancy) table through the mandatory field, CARE CONTACT DATE. 

Slide 9  

We want to be able to identify the information recorded at a specific care activity. This is collected in the table MSD202 Care Activity (Pregnancy).  

In the Care Activity (Pregnancy) table, we can identify the smoking status of the mother in their pregnancy. This can be identified by any one of the following fields: 

• CIGARETTES PER DAY 

• CO MONITORING READING  

• CODED OBSERVATION (CLINICAL TERMINOLOGY)  

• (through the recording of a SNOMED CT code)  

• CODED FINDING (CODED CLINICAL ENTRY)  

• (through the recording of a SNOMED CT, Read or ICD code).  

The codes that are used to identify smoking status can be found in the notes section for the specified CQIM DQ metrics in the metadata file.  

We will see how these data items are used in the upcoming slides. 

Slide 10  

We will start by looking at the denominator for CQIMDQ03, the number of women booking in MSDS in the reporting period, as a percentage of HES average monthly deliveries (using published 2018-19 data), pro-rata-ed over the number of days in the current reporting month. 

Let’s see how we do this. 

Slide 11  

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link here. 

Then, we count the number of days in the current month’s reporting period.  

We calculate the denominator as the number of days in the current month’s reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19. 

This is the denominator, the pro-rata-ed number of deliveries in HES. 

Slide 12   

Next, let’s look at how the numerator for CQIMDQ03 is calculated. We will be looking for all women who had a booking appointment in the current reporting month. 

Slide 13  

We use PERSON ID (MOTHER) to count unique women in the time frame. We confirm the records are in the current reporting month by using the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of women who had a booking appointment in the current reporting period. 

Slide 14  

The final step is to create the CQIMDQ03 measure as a percentage. 

Slide 15  

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 16  

We will start by looking at the denominator for CQIMDQ04, the number of women who had a booking appointment in the current reporting period. 

Slide 17  

We use PERSON ID (MOTHER) to count unique women in the time frame. We confirm the records are in the current reporting month by using the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women who had a booking appointment in the current reporting period. 

Slide 18  

Next, let’s look at how the numerator for CQIMDQ04 is calculated. We will be looking for the number of women who had a booking appointment in the current reporting month whose smoking status was known. 

Slide 19  

We use the PERSON ID (MOTHER) to count the number of women in the reporting period  

We use the CARE CONTACT DATE from MSD201CareContactPreg to identify whether the care activity is within three days either side of the booking appointment.  

We check that the REPORTING PERIOD START DATE is in the given month, and we also confirm that the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) is in the current reporting month. 

Next, we identify whether the woman has a smoking status recorded in at least one of the following fields:  

• CIGARETTES PER DAY, where the input is greater than or equal to 0. This is a derived field based on the provider’s entries in the MASTER SNOMED CT OBSERVATION CODE and OBSERVATION VALUE fields.  

• A CO MONITORING READING recorded as greater than or equal to 0 parts per million (ppm). This is a derived field based on the entries to the MASTER SNOMED CT OBSERVATION CODE, OBSERVATION VALUE and UCUM UNIT OF MEASUREMENT fields. 

• MASTER SNOMED CT OBSERVATION CODE is a derived field, largely taken from CODED OBSERVATION (CLINICAL TERMINOLOGY) which is submitted to by the provider. 

• A CODED OBSERVATION (CLINICAL TERMINOLOGY) code indicating that the person has stopped smoking. This field looks for a specific SNOMED CT code which can be found in the metadata for this metric. 

• A CODED FINDING (CODED CLINICAL ENTRY) code indicating a smoking status was recorded.  

Slide 20  

This is the list of SNOMED, Read and ICD codes which are currently used to indicate the mother’s smoking status. The code list can also be found in the metadata for CQIMDQ04-05. 

Slide 21  

The final step is to create the CQIMDQ04 measure as a percentage. 

Slide 22  

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 23  

We will start by looking at the denominator for CQIMDQ05, the number of women who had a booking appointment in the current reporting month whose smoking status was known. 

Slide 24  

We use the PERSON ID (MOTHER) to count the number of women in the reporting period  

We use the CARE CONTACT DATE from MSD201CareContactPreg to identify whether the care activity is within 3 days either side of the booking appointment.  

We check that the REPORTING PERIOD START DATE is in the given month, and we also confirm that the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) is in the current reporting month. 

Next, we identify whether the woman has a smoking status recorded in at least one of the following fields:  

• CIGARETTES PER DAY, where the input is greater than or equal to 0. This is a derived field based on the provider’s entries to the MASTER SNOMED CT OBSERVATION CODE and OBSERVATION VALUE fields.  

• A CO MONITORING READING recorded as greater than or equal to 0 parts per million (ppm). This is a derived field based on the entries to the MASTER SNOMED CT OBSERVATION CODE, OBSERVATION VALUE and UCUM UNIT OF MEASUREMENT fields.  

• MASTER SNOMED CT OBSERVATION CODE is a derived field, largely taken from CODED OBSERVATION (CLINICAL TERMINOLOGY) which is submitted to by the provider. 

• A CODED OBSERVATION (CLINICAL TERMINOLOGY) code indicating that the person has stopped smoking. This field looks for a specific SNOMED CT code which can be found in the metadata for this metric. 

• A CODED FINDING (CODED CLINICAL ENTRY) code indicating a smoking status was recorded. Details of the SNOMED CT, Read and ICD codes that are used can be found in the notes section of the metadata file for this metric, CQIMDQ05 listed as #SmokingStatusFinding. 

Slide 25  

Next, let’s look at how the numerator for CQIMDQ05 is calculated. We will be looking for the number of women who had a booking appointment in the current reporting month whose smoking status was recorded as known that they were current smokers. 

Slide 26  

We use the PERSON ID (MOTHER) to count the number of women in the reporting period, and we check that the REPORTING PERIOD START DATE is in the given month.  

We use the CARE CONTACT DATE from MSD201CareContactPreg to identify whether the care activity is within 3 days either side of the booking appointment. We also confirm that the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) is in the current reporting month. 

Next, we identify whether the woman has a smoking status recorded as being a current smoker in at least one of the following fields:  

• CIGARETTES PER DAY, where the input is greater than 0. This is a derived field based on the provider’s entries to the MASTER SNOMED CT OBSERVATION CODE and OBSERVATION VALUE fields.  

• A CO MONITORING READING recorded as greater than or equal to 4 parts per million (ppm). This is a derived field based on the entries to the MASTER SNOMED CT OBSERVATION CODE, OBSERVATION VALUE and UCUM UNIT OF MEASUREMENT fields.  

• MASTER SNOMED CT OBSERVATION CODE is a derived field, largely taken from CODED OBSERVATION (CLINICAL TERMINOLOGY) which is submitted to by the provider. 

• A CODED FINDING (CODED CLINICAL ENTRY) indicating a smoking status of current smoker was recorded. Details of the SNOMED CT, Read, and ICD codes that are used can be found in the notes section of the metadata file for this metric, CQIMDQ05, listed as #SmokingFinding.  

Slide 27  

• This is the list of SNOMED, Read and ICD codes which are currently used to indicate whether the mother was a current smoker. The code list can also be found in the metadata for CQIMDQ05. 

Slide 28  

The final step is to create the CQIMDQ05 measure as a percentage. 

Slide 29  

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 30  

Now, we look at the pass thresholds for the 3 DQ measures supporting CQIMSmokingBooking. A provider must pass all the data quality measures to be included in the published data. The provider’s rate for CQIMDQ03, CQIMDQ04, and CQIMDQ05 must be greater than or equal to 50%, greater than or equal to 70%, and between 0.5% and 50%, respectively. 

We will see how the values reached by the DQ measures affect what is shown in the published data. 

Slide 31  

Let’s look at the data published for March 2022. 

RQM – Chelsea and Westminster Hospital NHS Foundation Trust – has passed all 3 of the DQ measures associated with CQIMSmokingBooking. RQM reached 85.5% for CQIMDQ03, 99.4% for CQIMDQ04 and 5.7% for CQIMDQ05. 

This means that CQIMSmokingBooking figures are published for RQM, and this provider will contribute to the national and sub national figures included in the March 2022 published data. 

Slide 32  

Now, we will look at another trust using the March 2022 data. 

RWW – Warrington and Halton Teaching Hospitals NHS Foundation Trust – has passed 2 of the 3 DQ measures associated with CQIMSmokingBooking. RWW reached 91.1% for CQIMDQ03 and 8.3% for CQIMDQ05. It failed the DQ measure CQIMDQ04 with 58.5%.  

This means that CQIMSmokingBooking figures are not published for RWW. Instead, the denominator and numerator are shown as ‘0’ and the Rate as ‘Low DQ’ for CQIMSmokingBooking and will not contribute to the national and sub national figures included in the March 2022 published data. 

Slide 33  

Let’s look at how applying these data quality measures affects the larger picture. 124 providers submitted data for March 2022.  

In March 2022: 

122 providers passed CQIMDQ03, recording at least 70% for this DQ measure. 2 providers failed the measure or did not submit any data. 86 providers passed CQIMDQ04 and 38 failed or did not submit any data for this measure. 106 providers passed CQIMDQ05 and 18 failed or did not submit any data. 

Note that the pass rates of these measures could alter in the future. 

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.  

Slide 34  

This brings us to the end of the video. 

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, the proportion of women who were current smokers at their booking appointment, are built from MSDS. 

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, of women who were current smokers at the time of their delivery. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that finds, for any given month, the proportion of women with a birth in the previous month, who were current smokers at the time of their delivery.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look at the number of women with a birth in the previous month, where the outcome of that birth was a live or stillborn baby, and the woman had a known smoking status recorded within plus or minus three days of the start of delivery.

For the numerator, we keep only those women from the denominator whose recorded smoking status identifies them as a current smoker.

We will take the delivery start date as the date that established labour was confirmed, or the date of knife-to-skin where the delivery involved a caesarean section.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25, which can be found at the first link here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, and can be found at the second link.

Slide 5

This video explains how the measure is constructed and the data items used. There are 12 CQIM ‘rate’ measures. This video will look only at Women who were current smokers at the time of their delivery

‘CQIMSmokingDelivery’.

The metadata that details the build for CQIMSmokingDelivery describes the measure, the numerator, and the denominator. 

The timeframe for CQIMSmokingDelivery is one month prior to any given month. For example, if a given reporting month is January 2021, then this measure would count women with a delivery in December 2020.

Slide 6

We will start by looking at all the fields in MSDS that underpin this measure.

We use MSD000 Header to establish the reporting period and the organisation who submitted the data, from ORGANISATION IDENTIFIDER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE.

Slide 7

In order to identify the woman, and to find details about the birth of the baby, we use the table MSD401 Baby’s Demographics and Birth Details.

The woman is identified by PERSON ID (MOTHER) which is derived from demographic information, including NHS Number, that is submitted by the provider to MSDS in table MSD001 Mother's Demographics. To identify women who had eligible deliveries, we take the date of birth from PERSON BIRTH DATE (BABY) and use PREGNANCY OUTCOME to determine whether the delivery involved a live or stillbirth.

Slide 8

To determine the start of the woman’s delivery, we use the table MSD301 Labour and Delivery. Here, we look for both ONSET OF ESTABLISHED LABOUR DATE and PROCEDURE DATE (CAESAREAN SECTION) to establish the delivery start date, as the labour and delivery record should have at least one of these fields completed.

Slide 9

We use the table MSD302 Care Activity (Labour and Delivery) to determine whether information has been recorded which can be used to determine the woman’s smoking status. The CODED FINDING (CODED CLINICAL ENTRY) field is used to find codes, submitted by the provider, that directly describe the woman’s smoking status.

Other information – about cigarette consumption, carbon monoxide readings and dates of stopping smoking – is recorded in CODED OBSERVATION (CLINICAL TERMINOLOGY) and this can be used to derive her smoking status, when considered in combination with OBSERVATION VALUE – for example, the number of cigarettes smoked – and the UCUM UNIT OF MEASUREMENT, which identifies CO readings recorded with the appropriate units.

The CLINICAL INTERVENTION DATE (MOTHER) is used to find out when the smoking status was recorded, and from that, whether it took place at the time of the delivery. 

Slide 10

We use the table MSD201 Care Contact (Pregnancy) to determine when a smoking status was recorded for a woman, and from that, whether it took place at the time of their delivery. To do that we look for the CARE CONTACT DATE in this table.

Slide 11

We use the table MSD202 Care Activity (Pregnancy) to determine whether information was recorded, during care contacts of the pregnancy, which can determine the woman’s smoking status. This information may be submitted directly by the provider as codes describing the woman’s smoking status in CODED FINDING (CODED CLINICAL ENTRY).

Or, in CODED OBSERVATION (CLINICAL TERMINOLOGY) information is recorded – the cigarette consumption, carbon monoxide readings and dates of stopping smoking – which can be used to derive the smoking status, when considered in combination with OBSERVATION VALUE – for example, the number of cigarettes smoked – and the UCUM UNIT OF MEASUREMENT, which identifies CO readings recorded with the appropriate units.

Slide 12

We will start by looking at the denominator, the number of women who have a recorded smoking status at delivery, in the previous reporting month.

Let’s see which data items we use to do this.

Slide 13

We use the derived field PERSON ID (MOTHER) to count unique women during the previous reporting month.

First, we look for women who have a birth in the previous reporting month by checking whether both the REPORTING PERIOD START DATE and PERSON BIRTH DATE (BABY) are between the start and end dates of the previous reporting month. We keep only women with births that have a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’ recorded to ensure the baby was a live or stillbirth.

Then, we look for information recorded about the smoking status of these women. This can be recorded as activity during the labour and delivery (in table MSD302) or during care contacts of the pregnancy (in table MSD202). In either of these tables, information may be recorded in

• CODED FINDING (CODED CLINICAL ENTRY) as one of a list of codes related to smoking status;
• Or in CODED OBSERVATION (CLINICAL TERMINOLOGY) as SNOMED code ‘160625004’ recording the date the woman stopped smoking;
• Or in CODED OBSERVATION (CLINICAL TERMINOLOGY) with codes related to cigarette consumption, alongside an OBSERVATION VALUE with a non-negative number;
• Or in CODED OBSERVATION (CLINICAL TERMINOLOGY) with codes related to carbon monoxide measurement, together with a non-negative number in OBSERVATION VALUE and the UCUM UNIT OF MEASUREMENT showing units of ‘COPPM’ or ‘PPM’ using any case or spacing.

Finally, we compare the date the smoking status information was recorded to both the ONSET OF ESTABLISHED LABOUR DATE and the PROCEDURE DATE (CAESAREAN SECTION). We use the CLINICAL INTERVENTION DATE (MOTHER) if the smoking status was recorded in table MSD302, or the CARE CONTACT DATE if the smoking status was recorded in table MSD202. If the smoking status date is within plus or minus three days of either the labour onset date or the caesarean section date, then we retain the record for that woman.

This is our denominator, the number of women with a recorded, known, smoking status at delivery.

Slide 14

This is the list of codes which are currently used to determine smoking status if recorded in CODED FINDING (CODED CLINICAL ENTRY) or CODED OBSERVATION (CLINICAL TERMINOLOGY). Information about smoking status can be recorded using SNOMED, ICD, READ or CTV3 codes.

Slide 15

Next, we will look at the numerator, the number of women with a delivery in the previous month who have been recorded with a current smoking status of SMOKER at time of delivery. This is a subset of the women identified for the denominator – they have a delivery in the previous reporting month, and a known smoking status recorded within three days of that delivery, but there is the additional condition that the smoking status identifies them as a current smoker.

Let’s see which data items we use to do this.

Slide 16

We use the derived field PERSON ID (MOTHER) to count unique women during the previous reporting month.

To identify women with a delivery in the previous reporting month who have a known smoking status of current smoker within plus or minus three days of the start of that delivery, we follow the same process that we used to identify women in the denominator.

First, we look for women who have a birth in the previous reporting month by checking whether both the REPORTING PERIOD START DATE and PERSON BIRTH DATE (BABY) are between the start and end dates of the previous reporting month. We keep only women with births that have a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’ recorded to ensure the baby was a live or stillbirth.

Then, we look for information recorded about the smoking status of these women that identifies them as a current smoker. This may be recorded as activity during the labour and delivery (in table MSD302) or during care contacts of the pregnancy (in table MSD202). In either of these tables, information can be recorded in

• CODED FINDING (CODED CLINICAL ENTRY) with one of the listed codes that indicates the woman is a current smoker;
• Or in CODED OBSERVATION (CLINICAL TERMINOLOGY) with codes related to cigarette consumption, alongside an OBSERVATION VALUE which is greater than 0;
• Or in CODED OBSERVATION (CLINICAL TERMINOLOGY) with codes related to carbon monoxide measurement, together with an OBSERVATION VALUE greater than or equal to 4, and a UCUM UNIT OF MEASUREMENT with units of ‘COPPM’ or ‘PPM’ using any case or spacing.

Finally, we compare the date the smoking status information was recorded to both the ONSET OF ESTABLISHED LABOUR DATE and the PROCEDURE DATE (CAESAREAN SECTION). We use the CLINICAL INTERVENTION DATE (MOTHER) if the smoking status was recorded in table MSD302, and CARE CONTACT DATE if the smoking status was recorded in table MSD202. If the status of ‘current smoker’ was recorded on a date within plus or minus three days of either the labour onset date or the caesarean section date, then we retain the record for that woman.

This is our numerator, the number of women who are recorded as current smokers at delivery.

Slide 17

The final step is to create the measure as a percentage

Slide 18

To do this, we divide the numerator by the denominator. We then multiply the result by 100 to get a percentage.

Slide 19

A few key points to take away are

• Women may have more than one smoking status recorded within plus or minus three days of the start of their delivery. We will take the smoking status as recorded on the day their delivery started over a smoking status recorded on another day. Otherwise, we will take the latest data related to their smoking status.
• The measure CQIMSmokingDelivery reports on women who were current smokers at the time of their delivery for the previous reporting month. For example, in the January 2021 monthly publication CQIMSmokingDelivery reports on deliveries from December 2020.
• This measure collects smoking status information from the table MSD202 Care Activity (Pregnancy) and the table MSD302 Care Activity (Labour and Delivery). If a woman’s smoking status is recorded in another table – such as MSD109 Finding and Observation (Mother) – it has not been included in CQIMSmokingDelivery.

Slide 20

CQIMSmokingDelivery is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

The pass rate is 70 percent for both CQIMDQ02 and CQIMDQ06. In order to pass CQIMDQ07 the rate for that measure must be between 0.5 and 50 percent.

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 21

The description of the CQIM DQ measures that support CQIMSmokingDelivery can be found on this slide; this information is all in the metadata file.

Slide 22

This brings us to the end of the video.

Thank you for watching this video demonstration on how the CQIM measure, women who were current smokers at the time of their delivery, is built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Slide 23

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for women who were current smokers at the time of their delivery. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, referred to as MSDS.

Slide 2

Today we will look at the three data quality measures that support the CQIM measure, CQIMSmokingDelivery, the proportion of women with a birth in the previous month who were current smokers at the time of the delivery.

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.

This video will look at the DQ measures supporting CQIMSmokingDelivery and show some examples of how these are currently displayed in the published data.

Slide 4

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built. Links to the latest metadata are given alongside each monthly data release, and is also available by following the second link.

Many of the data items which appear in these DQ measures also exist in CQIMSmokingDelivery, the measure to which they are applied. A video describing the MSDS measure, CQIMSmokingDelivery, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5

There are three data quality measures that support CQIMSmokingDelivery, women who were current smokers at the time of their delivery. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.

The first DQ measure is CQIMDQ02. It calculates women giving birth in MSDS, as a percentage of the average number of HES deliveries, in the previous 1 month reporting period.

The second is CQIMDQ06. It calculates the percentage of women that gave birth in the previous 1 month reporting period, whose smoking status was known at the time of the delivery.

The third DQ measure is CQIMDQ07. It calculates the percentage of women that gave birth in the previous 1 month reporting period, who were known to be current smokers at the time of the delivery.

The time frame for CQIMDQ02 and CQIMDQ06-07 is a month prior to the given month. For example, if the given month was January 2021, then these DQ measures would all be counting women in December 2020.

Now, we will look at all the fields in MSDS that underpin these measures.

Slide 6

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

Slide 7

We want to be able to uniquely identify each mother, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER) to identify the mother. This field is primarily generated from the NHS Number submitted for the mother.

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers:

• The PERSON BIRTH DATE (BABY) to ensure the birth took place in the relevant time frame.

• The PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 8

We use the MSD301 Labour and Delivery table to find when the delivery associated with the birth took place. To do this we use the fields ONSET OF ESTABLISHED LABOUR DATE and PROCEDURE DATE (CAESAREAN SECTION) to establish the start date of the delivery. Both of these dates are included because a birth should have at least one of these recorded. These fields are submitted by the provider.

Slide 9

We use the MSD302 Care Activity (Labour and Delivery) table to find information recorded during the delivery about the woman’s smoking status.

We use these fields:

• The CLINICAL INTERVENTION DATE (MOTHER) describes when the activity took place. It is used so we can ensure that the smoking status was recorded within 3 days of the delivery of the birth. This field is submitted by the provider.
• We use the CODED FINDING (CODED CLINICAL ENTRY) field to find codes that have been recorded to describe the woman’s smoking status.
• CODED OBSERVATION (CLINICAL TERMINOLOGY) to find where information about the woman’s use of cigarettes, carbon monoxide readings and dates related to stopping smoking have been recorded.
• The OBSERVATION VALUE determines whether the carbon monoxide readings or number of cigarettes smoked per day indicate whether the woman was a smoker or a non-smoker.

• The UCUM UNIT OF MEASUREMENT is used to ensure that any carbon monoxide readings have been recorded in the appropriate units of parts per million.

All of these fields are directly submitted by the provider.

Slide 10

We use the MSD201 Care Contact (Pregnancy) table to find details about the care contacts at which smoking status information was recorded. We use CARE CONTACT DATE to find when the care contact where the smoking status was recorded took place. This field is submitted by the provider.

Slide 11

We use the MSD202 Care Activity (Pregnancy) table to find further information recorded at care contacts during the pregnancy.

We use these same fields as in table MSD302 to find information recorded related to the woman’s smoking status:

• We use the CODED FINDING (CODED CLINICAL ENTRY) field to find codes that have been recorded to describe the woman’s smoking status.

• CODED OBSERVATION (CLINICAL TERMINOLOGY) to find where information about the woman’s use of cigarettes, carbon monoxide readings and dates related to stopping smoking have been recorded.

• The OBSERVATION VALUE determines whether the carbon monoxide readings or number of cigarettes smoked per day indicate whether the woman was a smoker or a non-smoker.

• The UCUM UNIT OF MEASUREMENT is used to ensure that any carbon monoxide readings have been recorded in the appropriate units of parts per million.

All of these fields are directly submitted by the provider.

We will see how these items are used in the next slides.

Slide 12

We will start by looking at the denominator for CQIMDQ02, the number of deliveries in HES (using published 2018-19 data) pro rata-ed over the number of days in the month prior to a given month. Let’s see how we do this.

Slide 13

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link here.

Then, we count the number of days in the reporting period, that is in the month prior to a given month. If the reporting period is January 2021, then this is the total number of days in December 2020.

We calculate the denominator as the number of days in the reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19.

This is the denominator, the pro-rata-ed number of deliveries in HES corresponding to the month prior to a given month.

Slide 14

Next, let’s look at how the numerator for CQIMDQ02 is calculated. We will be looking for all women who gave birth in the month prior to a given month, where the outcome of the pregnancy was a live or stillbirth. Let’s see which data items we use to do this.

Slide 15

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. Then we look for and retain births with a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’, to ensure the birth is a live or stillbirth.

We confirm the records are in the month prior to a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our numerator, the number of women who have a live or stillbirth in the month prior to a given month.

Slide 16

The final step is to create the CQIMDQ02 measure as a percentage

Slide 17

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 18

We will start by looking at the denominator for CQIMDQ06, the number of women who have a live or stillbirth, in the previous 1 month reporting period. Let’s see how we do this.

Slide 19

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. Then we look for and retain births with a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’, to ensure the birth is a live or stillbirth.

We confirm the records are in the month prior to a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of women who have a live or stillbirth, in the month prior to a given month.

Slide 20

Next, let’s look at how the numerator for CQIMDQ06 is calculated. We will be looking for all women who have a live or stillbirth, in the previous 1 month reporting period, and who have a known smoking status at the time of their delivery – recorded within ± 3 days. Let’s see which data items we use to do this.

Slide 21

We use the cohort of women we found for the denominator, who had a live or stillbirth in the previous 1 month reporting period.

As for the denominator, we use PERSON ID (MOTHER) to count unique women in the time frame.

Then we retain only those women with a smoking status that was recorded in either of the tables MSD202 or MSD302. The smoking status may be recorded as

• One of a specified list of SNOMED, Read or ICD codes describing their smoking status, recorded in CODED FINDING (CODED CLINICAL ENTRY)

• Or as a code recorded in CODED OBSERVATION (CODED CLINICAL ENTRY) relating to the number of cigarettes smoked per day, alongside a non-negative number entered in OBSERVATION VALUE.

• Or it could be recorded in CODED OBSERVATION (CODED CLINICAL ENTRY) as a code relating to a carbon monoxide reading, alongside a non-negative number entered in OBSERVATION VALUE and with the appropriate units – ‘COPPM’ or ‘PPM’ – recorded in UCUM UNIT OF MEASUREMENT using any case or spacing.

• Or as the SNOMED code ‘160625004’ entered in CODED OBSERVATION (CODED CLINICAL ENTRY) to indicate a record which contains the date the woman stopped smoking.

We then keep any record where the smoking status was recorded within ± 3 days of the delivery.

Where the smoking status was recorded in MSD302, we look at the CLINICAL OBSERVATION DATE (MOTHER) and retain the mother where that date is within ± 3 days of either the ONSET OF ESTABLISHED LABOUR DATE or the PROCEDURE DATE (CAESAREAN SECTION). If the record of smoking status was found in MSD202, then we make the same comparison to the labour onset and caesarean section dates using CARE CONTACT DATE and retain women where that date is within ± 3 days of the start of the delivery.

This is our numerator, the number of women who had a live or stillbirth in the month prior to a given month, and whose smoking status is known at the time (within ± 3 days) of delivery.

Slide 22

This is the list of SNOMED, Read and ICD codes which are currently used to indicate the mother’s smoking status. This code list can also be found in the metadata for CQIMDQ06-07.

Slide 23

The final step is to create the CQIMDQ06 measure as a percentage.

Slide 24

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 25

We will start by looking at the denominator for CQIMDQ07, the number of women who had a live or stillbirth in the previous 1 month reporting period, and who had a known smoking status recorded within ± 3 days of their delivery. This denominator is the same as the numerator used for CQIMDQ06.

Slide 26

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.

We keep only women who have given birth where the outcome of the pregnancy – PREGNANCY OUTCOME – was recorded as ‘01’, ‘02’, ‘03’ or ‘04’ to indicate a live or stillbirth.

Then we retain only those women with a smoking status that was recorded in either of the tables MSD202 or MSD302. The smoking status may be recorded as

• One of a specified list of SNOMED, Read or ICD codes describing their smoking status, recorded in CODED FINDING (CODED CLINICAL ENTRY)

• Or as a code recorded in CODED OBSERVATION (CODED CLINICAL ENTRY) relating to the number of cigarettes smoked per day, alongside a non-negative number entered in OBSERVATION VALUE.

• Or it could be recorded in CODED OBSERVATION (CODED CLINICAL ENTRY) as a code relating to a carbon monoxide reading, alongside a non-negative number entered in OBSERVATION VALUE and with the appropriate units – ‘COPPM’ or ‘PPM’ – recorded in UCUM UNIT OF MEASUREMENT using any case or spacing.

• Or as the SNOMED code ‘160625004’ entered in CODED OBSERVATION (CODED CLINICAL ENTRY) to indicate a record which contains the date the woman stopped smoking.

We then keep any record where the smoking status was recorded within ± 3 days of the delivery.

Where the smoking status was recorded in MSD302, we look at the CLINICAL OBSERVATION DATE (MOTHER) and retain the mother where that date is within ± 3 days of either the ONSET OF ESTABLISHED LABOUR DATE or the PROCEDURE DATE (CAESAREAN SECTION). If the record of smoking status was found in MSD202, then we make the same comparison to the labour onset and caesarean section dates using CARE CONTACT DATE and retain women where that date is within ± 3 days of the start of the delivery.

We confirm the records are in the month prior to a given month, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of women who had a live or stillbirth in the previous 1 month reporting period, with a known smoking status recorded within ± 3 days of their delivery.

Slide 27

Next, let’s look at how the numerator for CQIMDQ07 is calculated. We will be looking for women who had a live or stillbirth in the previous 1 month reporting period, and who were known to be current smokers at the time of their delivery.

Slide 28

We use the cohort of women we found for the denominator, who had a live or stillbirth in the previous 1 month reporting period, and who had a known smoking status recorded at the time of their delivery.

As for the denominator, we use PERSON ID (MOTHER) to count unique women in the time frame.

Then we retain only those whose recorded smoking status indicates that they were a current smoker. A woman is considered a current smoker if her smoking status has been identified from:

• A SNOMED, Read or ICD code recorded in CODED FINDING (CODED CLINICAL ENTRY) which is listed as a ‘Smoker’ code.

• A code recorded in CODED OBSERVATION (CODED CLINICAL ENTRY) related to the number of cigarettes smoked per day, where the OBSERVATION VALUE associated with that code is greater than 0.

• A code recorded in CODED OBSERVATION (CODED CLINICAL ENTRY) related to a carbon monoxide reading, where the OBSERVATION VALUE is greater than or equal to 4, and the UCUM UNIT OF MEASUREMENT is recorded as ‘COPPM’ or ‘PPM’ with any case or spacing.

These are only kept where the smoking status has been recorded within ± 3 days of the delivery.

This is done by comparing the corresponding CLINICAL OBSERVATION DATE (MOTHER) – for records of smoking status taken from table MSD302 – or CARE CONTACT DATE – for records of smoking status taken from table MSD202 – to each of the ONSET OF ESTABLISHED LABOUR DATE and PROCEDURE DATE (CAESAREAN SECTION) of their delivery.

This is our numerator, the number of women who had a live or stillbirth in the month prior to a given month, and whose smoking status is that of a current smoker at the time (within ± 3 days) of delivery.

Slide 29

The final step is to create the CQIMDQ07 measure as a percentage.

Slide 30

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate.

Slide 31

Now, we look at the pass thresholds for the three DQ measures supporting CQIMSmokingDelivery. A provider must pass all the data quality measures to be included in the published data. The provider’s rates for CQIMDQ02 and CQIMDQ06 must be 70% or more, and the provider rate for CQIMDQ07 must be between 0.5% and 50% inclusive.

If the denominator of CQIMDQ02 is 0 and the numerator is greater than 0, then the provider will pass that DQ measure.

We will see how the values reached by the DQ measures affect what is shown in the published data.

Slide 32

Let’s look at the data published for October 2021.

RXF – Mid Yorkshire Hospitals NHS Trust – has passed all three of the DQ measures associated with CQIMSmokingDelivery. RXF reached 100.0% for CQIMDQ02, 99.0% for CQIMDQ06, and 13.3% for CQIMDQ07.

This means that the CQIMSmokingDelivery figures are published for RXF, and that data from this provider will contribute to the national and sub national figures included in the October 2021 published data.

Slide 33

Now, we will look at another trust using the October 2021 data.

RXC – East Sussex Healthcare NHS Trust – has passed two of the DQ measures associated with CQIMSmokingDelivery. RXC reached 312.5% for CQIMDQ02 and 21.4% for CQIMDQ07. It failed the DQ measure CQIMDQ06 with 28.0%.

This means that the CQIMSmokingDelivery figures are not published for RXC. Instead, the denominator and numerator are shown as ‘0’ and the Rate as ‘Low DQ’ at provider level for the metric and it will not contribute to the national and sub-national figures included in the October 2021 published data.

Slide 34

Let’s look at how applying these data quality measures affects the larger picture. 119 providers submitted data for October 2021 births. As the CQIM DQ measures include providers which did not submit data in this month, 123 providers were considered for the DQ measures CQIMDQ02, CQIMDQ06 and CQIMDQ07.

In October 2021:

115 providers passed CQIMDQ02. 8 providers failed the measure or did not submit any data for this measure. 44 providers passed CQIMDQ06, and 63 providers passed CQIMDQ07.

Note that the pass rates of these measures could alter in the future.

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.

Slide 35

This brings us to the end of the video.

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, CQIMSmokingDelivery, women who were current smokers at the time of the delivery, are built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the Clinical Quality Improvement Metric, also known as CQIM, for women delivering vaginally who had a 3rd or 4th degree tear. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS.

Slide 2

Today we will look at a measure that calculates, in any given month and the two months prior, 3rd and 4th degree tear rate among women delivering vaginally.

Slide 3

We do this by establishing a denominator and numerator. For the denominator we will look for all women who have a vaginal birth (excluding breech) of a singleton baby born between 37 and 45 weeks gestation, in a given month and the two months prior.

For the numerator, we take the cohort of women in the denominator and retain only those who have a 3rd or 4th degree tear.

Slide 4

We will look at the data items in MSDS that contribute to this measure. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published meta data to explain how the measures are built. This is available with each monthly data release, available by following the bottom link.

Slide 5

This video explains how the measure is constructed and the data items used. There are eleven CQIM ‘rate’ measures. This video will look only at women delivering vaginally who had a 3rd or 4th degree tear.

The metadata that details the build for CQIMTears describes the measure, the numerator, and the denominator. First, we will look at all the fields in MSDS that underpin this measure.

The timeframe for CQIMTears is a given month and the two months prior. For example, if the given month was January 2021, then the measure would be counting women in January 2021, December 2020 and November 2020.

Slide 6

We want to be able to uniquely identify each mother; we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER). This is primarily generated from the NHS Number.

We also use MSD401 Baby Demographics to get the following 4 pieces of information:

• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame
• PREGNANCY OUTCOME to ensure the baby is born, with 01 meaning live birth and 02, 03 and 04 meaning still birth.
• GESTATION LENGTH (AT BIRTH) to ensure gestation is between 37 and 45 weeks (so between 259 and 315 days).
• DELIVERY METHOD CODE to include only vaginal births (excluding breech) with 0, 1, 2, 3 or 4 for vaginal birth.

Slide 7

We use the table MSD301 Labour and Delivery to get the BIRTHS PER LABOUR AND DELIVERY in order to identify singleton births, with 1 meaning singleton.

Slide 8

We use MSD302 Care Activity (Labour and Delivery) to identify women who had a 3rd or 4th degree tear. We do this using the field GENITAL TRACT TRAUMATIC LESION which is derived using the data submitted as FINDING or PROCEDURE on the CLINICAL INTERVENTION DATE (MOTHER) to MSD302.

Slide 9

And lastly, we use MSD000 Header to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE

We will see how these items are used in the next slides.

Slide 10

We will start by looking at the denominator, number of women who have a vaginal birth (excluding breech) of a singleton baby born between 37 and 45 weeks gestation, in a given month and the two months prior.

Let’s see which data items we use to do this.

Slide 11

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next consider only singleton births by checking if BIRTHS PER LABOUR AND DELIVERY is equal to 1. This is also a derived field.

Then we look for PREGNANCY OUTCOME of 01, 02, 03 or 04 to include live and still births.

We include only vaginal births, with DELIVERY METHOD CODE recorded as 0, 1, 2, 3 or 4.

Next, we consider only babies with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days (between 37 and 45 weeks).

We confirm the records are in a given month and two months prior using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record.

This is our denominator, the number of women who have a vaginal birth (excluding breech) of a singleton baby born between 37 and 45 weeks in a given month and the two months prior.

Slide 12

Next, let’s look at how the numerator is calculated. We will be looking for all women who have a vaginal birth (excluding breech) of a singleton baby born between 37 and 45 weeks who have a 3rd or 4th degree tear, in a given month and the two months prior.

Let’s see which data items we use to do this.

Slide 13

We use the cohort of women we found for the denominator, then retain only those who have a 3rd or 4th degree tear recorded as SNOMED CT codes listed in the derived field GENITAL TRACT TRAUMATIC LESION. Again, we use PERSON ID (MOTHER) to count unique women in the time frame.

Slide 14

A few key points to take away are

• We take the latest data that is available for a given month and the two months prior to it.

This measure requires the following data items to be recorded

• Have PREGNANCY OUTCOME recorded to identify live and still births
• Have DELIVERY METHOD CODE recorded to identify vaginal births
• Have GESTATION LENGTH (AT BIRTH) recorded for the baby

Women who had a 3rd or 4th degree tear is defined by the list of SNOMED CT codes submitted as a FINDING or PROCEDURE on the CLINICAL INTERVENTION DATE (MOTHER) to MSD302 Care Activity (Labour and Delivery).

Slide 15

The final step is to create the measure as a rate.

Slide 16

We simply divide the numerator by the denominator and multiply by 1000 to get the rate per thousand.

Maternity Services Monthly Statistics - NHS Digital

Slide 17

CQIMTears is reliant on the following DQ measures and providers will need to meet the pass rate (shown in brackets) for each DQ measure in order for their data for the measure to be considered of good enough quality to be published.

This shows the relevant CQIM DQ measures and their associated pass rates. For example, CQIMDQ14 has a pass rate of greater than or equal to 70%.

A separate video looks at how these DQ measures are constructed, and information about how they are built can be found in the metadata.

Slide 18

The description of the CQIM DQ measures can be found on this slide. This information is all in the metadata file.

Slide 19

This brings us to the end of the video.

Thank you for watching this video demonstration on how the measure for women delivering vaginally who had a 3rd or 4th degree tear is built from MSDS.

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for women delivering vaginally who had a 3rd or 4th degree tear. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset, referred to as MSDS.

Slide 2

Today we will look at the seven data quality measures that support the CQIM measure, CQIMTears, women delivering vaginally who had a 3rd or 4th degree tear.

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers must pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national or sub national breakdowns.

This video will look at the DQ measures supporting CQIMTears and show some examples of how these are currently displayed in the published data.

Slide 4

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built. Links to the latest metadata are given alongside each monthly data release, and is also available by following the second link.

Many of the data items which appear in these DQ measures also exist in the measure to which they are applied, CQIMTears. A video describing the MSDS measure, CQIMTears, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5

This video explains how the DQ measures are constructed and the data items used. There are seven data quality measures that support CQIMTears, the rate of women delivering vaginally who had a 3rd or 4th degree tear. We will explain how each of them is constructed, and the data items used. The metadata, accessed by the link here, describes each measure, its numerator, and its denominator.

The first DQ measure is CQIMDQ14. It calculates women giving birth, as a percentage of average monthly deliveries, in the current 3 month reporting period​ taken from the Hospital Episode Statistics data set, also known as HES. 

The second is CQIMDQ15. It calculates the percentage of singleton babies with a valid gestational length at birth between 154 and 315 days (22 - 45 weeks) in the current 3 month reporting period.

The third is CQIMDQ16. It calculates the percentage of singleton babies with a gestational length at birth between 259 and 315 days (37 - 45 weeks) in the current 3 month reporting period.

Slide 6

The fourth DQ measure is CQIMDQ18. It calculates the percentage of singleton babies born vaginally in the current 3 month reporting period.

The fifth data quality measure is CQIMDQ19. It calculates the percentage of singleton babies born by caesarean section in the current 3 month reporting period.

The sixth data quality measure is CQIMDQ20. It calculates the percentage of singleton babies born vaginally with a 3rd or 4th degree tear in the current 3 month reporting period.

The seventh data quality measure is CQIMDQ21. It calculates whether at least one 3rd or 4th degree tear is recorded in the previous 6 month reporting period.

The timeframe for CQIMDQ14-16 and CQIMDQ18-20 is the given month and the 2 months prior; and for CQIMDQ21 it is the 6 months prior to a given month.  For example, if the given month was January 2021, then CQIMDQ14-16 and CQIMDQ18-20 would be counting women and babies in November 2020 to January 2021; while CQIMDQ21 would be counting women in July 2020 to December 2020. 

Now, we will look at all the fields in MSDS that underpin these measures. 

Slide 7

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data using the fields ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE.

Slide 8

We want to be able to uniquely identify each mother and baby, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER) to identify the mother, and one called PERSON ID (BABY) to identify the baby. These fields are primarily generated from the NHS Number(s) submitted for mother and baby. 

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers: 

• The DELIVERY METHOD CODE to identify how the baby is born, specifically if the delivery was vaginal, or involved a caesarean section.

• The GESTATION LENGTH (AT BIRTH) to identify births at full term, and those babies with a valid gestation.
• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame
• PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 9

We use the MSD301 Labour and Delivery table to identify births where a singleton baby was born. 

To do this we use the field BIRTHS PER LABOUR AND DELIVERY, identifying singleton births where this field equals 1. This is a derived field that depends on the number of births submitted by providers under each LABOUR AND DELIVERY IDENTIFIER. 

Slide 10

We use the MSD302 Care Activity Labour and Delivery table to identify women who had a 3rd or 4th degree tear as part of their labour and delivery.

We use these fields: 

• The CLINICAL INTERVENTION DATE (MOTHER) ensures that the 3rd or 4th degree tear takes place during the reporting period, 6 months prior to a given month. 

• The GENITAL TRACT TRAUMATIC LESION is a derived field. It is used to identify where a delivery involved a 3rd or 4th degree tear, and uses SNOMED coded information submitted by providers as a procedure in CODED PROCEDURE AND PROCEDURE STATUS (CODED CLINICAL ENTRY) or as a finding in CODED FINDING (CODED CLINICAL ENTRY). 

Slide 11

We will start by looking at the denominator for CQIMDQ14, the number of deliveries in HES (using published 2018-19 data) pro rata-ed over the number of days in a given month and the 2 months prior. To do this, we first find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link here. .

Then, we count the number of days in the reporting period, that is in any given month and the 2 months prior. If the reporting period is January 2021, then this is the total number of days in November 2020, December 2020 and January 2021. 

We then multiply the number of days in the reporting period by the number of deliveries recorded in HES for 2018-19 and then divide that total by the number of days in 2018-19 to calculate the pro-rata-ed number of deliveries in HES. 

This is the denominator. 

Slide 12

Next, let’s look at the numerator for CQIMDQ14. We will be looking for all women who gave birth in a given month and the 2 months prior, where there was a live or stillbirth. To do this, we use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next look for a PREGNANCY OUTCOME of 01, 02, 03 or 04 to include only live and still births.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of women who have a live or stillbirth in a given month and the 2 months prior. 

Slide 13

The final step is to create the CQIMDQ14 measure as a percentage. We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 14

The next measures is CQIMDQ15, the percentage of singleton babies with a valid gestational length. The denominator for this measure is the number of singleton babies who were born in the current 3 month reporting period.

To calculate this, we use the derived field, PERSON ID (BABY), to count unique babies in the time frame. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is the denominator, the number of singleton babies in a given month and the 2 months prior. 

Slide 15

Next, let’s look at how the numerator for CQIMDQ15 is calculated. We will be looking for all singleton babies born in the current 3 month reporting period, where the baby had a gestation length at birth between 154 and 315 days (between 22 and 45 weeks). To calculate this, we use the cohort of babies we found for the denominator, then retain only those recorded with a gestation length at birth between 154 and 315 days. Again, we use PERSON ID (BABY) to count unique babies in the time frame. 

To find babies with a valid gestation length at birth we look for anyone in the denominator who has a GESTATION LENGTH (AT BIRTH) between 154 and 315 days. 

This is our numerator, the number of babies who have a singleton birth in a given month and the 2 months prior, and a gestation length of between 154 and 315 days (between 22 and 45 weeks) at birth. 

Slide 16  

The final step is to create the CQIMDQ15 measure as a percentage. To do this, we simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 17

The next measures is CQIMDQ16, the percentage of singleton babies born at full term. The denominator is the number of singleton babies who were born in the current 3 month reporting period, with a valid gestational length at birth (this is between 154 and 315 days). This denominator is the same as the numerator used for CQIMDQ15. As before, we use derived field, PERSON ID (BABY), to count unique babies in the time frame. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We select only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 154 and 315 days. 

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of singleton babies born with a gestational length at birth between 154 and 315 days in a given month and the 2 months prior. 

Slide 18

Next, we look at how the numerator for CQIMDQ16 is calculated. We will be looking for singleton babies born at full term in the current 3 month reporting period. Full term is defined as a gestational length at birth between 259 and 315 days. To calculate this, we use PERSON ID (BABY) to count unique babies in the time frame. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We select only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days. 

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of singleton babies born with a gestational length at birth between 259 and 315 days in a given month and the 2 months prior. 

Slide 19

The final step is to create the CQIMDQ16 measure as a percentage. We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 20

The next measures is CQIMDQ18, the percentage of singleton babies born vaginally in the current 3-month reporting period.  The denominator for this measure in the number of singleton babies who were born in the current 3-month reporting period, with a valid delivery method. To calculate this, we use the derived field, PERSON ID (BABY), to count unique babies in the time frame. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We select only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’,’2’,’3’,’4’,’5’,’6’,’7’,’8’ or ‘9’. 

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is the denominator, the number of singleton babies born with a valid delivery method in a given month and the 2 months prior. 

Slide 21

Next, we look at how the numerator for CQIMDQ18 is calculated. We will be looking for singleton babies born in the current 3-month reporting period where the method of delivery was vaginal (excluding breech). As before we use the derived field, PERSON ID (BABY), to count unique babies in the time frame. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We select only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’,’2’,’3’ or ’4’ for a vaginal birth (excluding breech). 

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is the numerator, the number of singleton babies born with a vaginal delivery method in a given month and the 2 months prior. 

Slide 22

The final step is to create the CQIMDQ18 measure as a percentage. To do this we divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 23

The next data quality measure is CQIMDQ19, the percentage of singleton babies born by caesarean section in the current 3 month reporting period. The denominator for this measure is the number of singleton babies who were born in the current 3 month reporting period, with a valid delivery method. This denominator is the same as the denominator used for CQIMDQ18. As before, we use the derived field, PERSON ID (BABY), to count unique babies in the time frame. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.

We select only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’,’2’,’3’,’4’,’5’,’6’,’7’,’8’ or ‘9’. 

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is the denominator, the number of singleton babies born with a valid delivery method in a given month and the 2 months prior. 

Slide 24

Next, we look at how the numerator for CQIMDQ19 is calculated. We will be looking for singleton babies born by caesarean section in the current 3 month reporting period. To calculate this, we use the derived field, PERSON ID (BABY), to count unique babies in the time frame and where the ‘BIRTHS PER LABOUR AND DELIVERY’ field is 1 to include only singleton births.

We select only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘7’ or ’8’ to identify caesarean sections and confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is the numerator, the number of singleton babies born with a caesarean section delivery method in a given month and the 2 months prior. 

Slide 25

The final step is to create the CQIMDQ19 measure as a percentage. To do this, we divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 26

The next measure is CQIMDQ20, the percentage of singleton babies born vaginally with a 3rd or 4th degree tear in the current 3-month reporting period. The denominator for this measure is the number of full-term, live or stillbirth, babies who were born vaginally in the current 3-month reporting period.

Slide 27  

To calculate the denominator, we use the derived field PERSON ID (BABY) to count unique babies in the time frame.

We look for a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’ to include only live and stillbirths.

We select only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days (37 and 45 weeks) to find births at full term. We use DELIVERY METHOD CODE to include vaginal births (not including breech) with values ‘0’, ‘1’, ‘2’, ‘3’ or ‘4’. 

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is the denominator, the number of full term, live or stillbirth, babies who were born vaginally in a given month and the 2 months prior. 

Slide 28

Next, we look at how the numerator for CQIMDQ20 is calculated. We will be looking for full term, live or stillbirth, singleton babies who were born vaginally, in the current 3 month reporting period, where there was a 3rd or 4th degree tear. 

Slide 29

 As with the other measures, to calculate this we use the derived field, PERSON ID (BABY), to count unique babies in the time frame and BIRTHS PER LABOUR AND DELIVERY field to identify when this field equals 1 to include only singleton births.

We then look for a PREGNANCY OUTCOME of ‘01’, ‘02’, ‘03’ or ‘04’ to include only live and stillbirths.

We select only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days (37 and 45 weeks) to find births at full term. We use DELIVERY METHOD CODE to include vaginal births (not including breech) with values ‘0’, ‘1’, ‘2’, ‘3’ or ‘4’. 

Then we select births where the GENITAL TRACT TRAUMATIC LESION has a SNOMED CT code representing a 3rd or 4th degree tear.

We confirm the records are in a given month and the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of full term, live or stillbirth, singleton babies who were born vaginally in a given month and the 2 months prior, where there was a 3rd or 4th degree tear.

Slide 30

The final step is to create the CQIMDQ20 measure as a percentage. To calculate this, we divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 31

The final data quality measure related to the CQIMTears is CQIMDQ21, which shows where at least one 3rd or 4th degree tear is recorded in the previous 6 month reporting period for each provider.

 The denominator for this data quality measure is 1. 

Slide 32

Finally, we look at how the numerator for CQIMDQ21 is calculated. We will be looking for all women who had been recorded with a 3rd or 4th degree tear in the previous 6-month reporting period. If there are women with that recorded, then the numerator will be 1; if there are none, then the numerator will be 0. To calculate this, we use the derived field, PERSON ID (MOTHER), to count unique women in the time frame. We look for women with a SNOMED CT code representing a 3rd or 4th degree tear recorded in GENITAL TRACT TRAUMATIC LESION.

We confirm the records are in the six months prior to a given month, by using the CLINICAL INTERVENTION DATE (MOTHER) and the REPORTING PERIOD START DATE for each record.

If women have been recorded with a 3rd or 4th degree tear in the six months prior, then we set the numerator to one; otherwise, we set the numerator to zero. This is the numerator, representing whether women have been recorded with a 3rd or 4th degree tear in the six months prior to a given month. 

Slide 33

The final step is to create the CQIMDQ21 measure as a percentage. We divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 34

 Now, we look at the pass thresholds for the seven DQ measures supporting CQIMTears. A provider must pass all the data quality measures to be included in the published data. The provider’s rates for CQIMDQ14-16 must be 70% or more.

Please note- if the denominator of CQIMDQ14 is 0 and the numerator is greater than 0, then the provider will pass the DQ measure. 

Slide 35

 The provider pass rate for CQIMDQ18 must be 40% or more; the rate for CQIMDQ19 must be 50% or less; the rate for CQIMDQ20 must be 10% or less; and the rate for CQIMDQ21 must be 100%. 

We will move to explore how these data quality measures and pass rates affect what is shown in the published data. 

Slide 36

 Here is the data published for October 2021. 

RBK – Walsall Healthcare NHS Trust – has passed all seven of the DQ measures associated with CQIMTears. RBK reached 104.4% for CQIMDQ14, 100.0% for CQIMDQ15, 91.7% for CQIMDQ16, 66.2% for CQIMDQ18, 33.3% for CQIMDQ19, 1.6% for CQIMDQ20 and 100.0% for CQIMDQ21. 

This means that CQIMTears figures are published for RBK, and this provider will contribute to the national and sub-national figures included in the October 2021 published data. 

Slide 37

 Now, we will look at another trust using the October 2021 data. 

R1H – Barts Health NHS Trust – has passed five of the DQ measures associated with CQIMTears. R1H reached 86.9% for CQIMDQ14, 93.6% for CQIMDQ16, 75.9% for CQIMDQ18, 23.8% for CQIMDQ19 and 0.0% for CQIMDQ20. It failed the DQ measures CQIMDQ15, with 63.8%, and CQIMDQ21 with 0.0%. 

This means that CQIMTears figures are not published for R1H. Instead, the denominator and numerator are shown as ‘0’ and the Rate per Thousand as ‘Low DQ’ for CQIMTears and will not contribute to the national and sub-national figures included in the October 2021 published data. 

Slide 38

Here we can see how applying these data quality measures affects the larger picture. 119 providers submitted data for October 2021 births. As the CQIM DQ measures span a time frame longer than just one month, 123 providers were considered for the DQ measures CQIMDQ14-16, CQIMDQ18-21. 

In October 2021:

115 providers passed both CQIMDQ14 and CQIMDQ18. 8 providers failed the measure or did not submit any data for this measure at all. 113 providers passed CQIMDQ15, and 117 providers passed CQIMDQ20.

118 providers passed CQIMDQ16 and CQIMDQ19, while 5 did not submit any data for these DQ measures. 

86 providers passed CQIMDQ21, recording at least one 3rd or 4th degree tear in the six months prior to October 2021. 37 providers failed the measure, as they did not record any 3rd or 4th degree tears over this time frame.

Please note that the pass rates of these measures could alter in the future.

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.

Slide 39

This brings us to the end of the video.

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, women delivering vaginally who had a 3rd or 4th degree tear, are built from MSDS. We hope you found this useful.

This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.

Slide 1

Welcome to this short video demonstration on how to understand the data quality measures that support the Clinical Quality Improvement Metric, also known as CQIM, for women who gave birth to a single second baby vaginally, at or after 37 weeks, following a previous caesarean section. This video is part of a series of videos to help you understand measures built from the Maternity Services Dataset referred to as MSDS. 

Slide 2

Today we will look at the eight data quality measures that support the CQIM measure, CQIMVBAC, women who gave birth to a single second baby vaginally, at or after 37 weeks, following a previous caesarean section. 

Slide 3

Let’s look at why data quality thresholds are used. Measures built from the Maternity Services Dataset (MSDS) are published at various breakdowns, most commonly:

Provider, National, NHS England (Region), Local Maternity System, MBRRACE Grouping

To try and reduce the impact of poor data quality at provider level on the national and sub national breakdowns, we sometimes apply one or more data quality thresholds to measures. Providers have to pass these thresholds for their data for the measure they support to be published and contribute to national and sub national breakdowns.

Provider level data for the data quality measures is published alongside a result of pass or fail. If a provider fails the data quality measures, no data will be published for the measure that the data quality measures were supporting and data for that provider will not be used in any national and sub national breakdowns.  

This video will look at the DQ measures supporting CQIMVBAC and show some examples of how these are currently displayed in the published data.  

Slide 4

We will look at the data items in MSDS that contribute to these DQ measures. All of these data items can be found in the technical output specification version 2.0.25 available from the top link seen here.

There is also published metadata to explain how the measures are built. Links to the latest metadata are given alongside each monthly data release, and is also available by following the second link.

Many of the data items which appear in these DQ measures also exist in CQIMVBAC, the measure to which they are applied. A video describing the MSDS measure, CQIMVBAC, and how it is built, can be found at the NHS Digital Maternity Hub, available by following the third link.

Slide 5

There are eight data quality measures that support CQIMVBAC, the percentage of women who gave birth to a single second baby vaginally, at or after 37 weeks, following a previous caesarean section. We will explain how each of them is constructed, and the data items used. The metadata that details the builds for these DQ measures describes each measure, its numerator, and its denominator.

The first DQ measure is CQIMDQ14. It calculates women giving birth in MSDS, as a percentage of the average number of HES deliveries, in the current 3 month reporting period.

The second is CQIMDQ15. It calculates the percentage of singleton babies with a valid gestational age at birth between 154 and 315 days (22 - 45 weeks) in the current 3 month reporting period.  

Slide 6

The third DQ measure is CQIMDQ16. It calculates the percentage of singleton babies with a gestational age at birth between 259 and 315 days (37 - 45 weeks) in the current 3 month reporting period.  

The fourth is CQIMDQ18. It calculates the percentage of singleton babies born vaginally in the current 3 month reporting period. The fifth is CQIMDQ19. It calculates the percentage of singleton babies born by caesarean section in the current 3 month reporting period.  

The timeframe for CQIMDQ14-16, CQIMDQ18-19 and CQIMDQ26-28 is the given month and the 2 months prior.  For example, if the given month was January 2021, then all of these CQIM DQ measures would be counting women and babies in November 2020 to January 2021. 

Slide 7

The sixth DQ measure is CQIMDQ26. It calculates the percentage of singleton babies born with a valid delivery method recorded in the current 3 month reporting period. CQIMDQ27 is the next DQ measure, and it calculates the percentage of women who have been recorded with valid previous live and stillbirth data in the current 3 month reporting period. 

The eighth and last DQ measure supporting CQIMVBAC is CQIMDQ28. It calculates the percentage of women, recorded in the current 3 month reporting period, with no previous births.  

Now, we will look at all the fields in MSDS that underpin these measures. 

Slide 8

We will start with looking at the table MSD000 Header. We use this table to establish the reporting month and the organisation who submitted the data, from ORGANISATION IDENTIFIER (CODE OF PROVIDER), REPORTING PERIOD START DATE and REPORTING PERIOD END DATE. 

Slide 9

We want to be able to uniquely identify each mother and baby, we do this through data submitted to MSD401 Baby Demographics. From the data submitted by all providers, NHS Digital creates a field called PERSON ID (MOTHER) to identify the mother, and one called PERSON ID (BABY) to identify the baby. These fields are primarily generated from the NHS Number(s) submitted for mother and baby. 

We also use MSD401 Baby Demographics to get the following pieces of information that are submitted by providers: 

• The DELIVERY METHOD CODE to identify how the baby is born, specifically if the delivery was vaginal, or involved a caesarean section.
• The GESTATION LENGTH (AT BIRTH) to identify births at full term, and those babies with a valid gestation.
• The PERSON BIRTH DATE (BABY) to ensure the baby is born in the relevant time frame.
• PREGNANCY OUTCOME to ensure the baby is born, with ‘01’ meaning live birth and ‘02’, ‘03’ and ‘04’ meaning stillbirth.

Slide 10

We use the MSD301 Labour and Delivery table to identify births where a singleton baby was born. 

To do this we use the field BIRTHS PER LABOUR AND DELIVERY, identifying singleton births where this field equals 1. This is a derived field that depends on the number of births submitted by providers under each LABOUR AND DELIVERY IDENTIFIER. 

Slide 11

We use the MSD101 Pregnancy and Booking Details table to identify women with a booking appointment, and to find information about their birth history.  

We use these fields: 

• The APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) ensures that the booking appointment for the pregnancy took place during the reporting period, in a given month and the 2 months prior. 

• The PERSON ID (MOTHER) identifies the mother who attended the booking appointment. This field is primarily derived from the NHS Number for the mother which is submitted by the provider. 

• The PREGNANCY TOTAL PREVIOUS LIVE BIRTHS and PREGNANCY TOTAL PREVIOUS STILLBIRTHS designate the number of live and stillbirths which have resulted from previous pregnancies. This information, submitted by the provider, will be used to find pregnancies with no history of previous births.   

We will see how these items are used in the next slides. 

Slide 12

For CQIMDQ14, we will start by looking at the denominator, the number of deliveries in HES (using published 2018-19 data) pro rata-ed over the number of days in a given month and the 2 months prior. Let’s see how we do this. 

Slide 13

First, we find the number of deliveries recorded in HES for 2018-19. These figures were published in the NHS Maternity Statistics 2018-19, which can be found at the link to that publication. 

Then, we count the number of days in the reporting period, that is in a given month and the 2 months prior. If the reporting period is January 2021, then this is the total number of days in November 2020, December 2020 and January 2021. 

We calculate the denominator as the number of days in the reporting period multiplied by the number of deliveries recorded in HES in 2018-19 and then divide by the number of days (365) in 2018-19. 

This is the denominator, the pro-rata-ed number of deliveries in HES. 

Slide 14

Next, let’s look at how the numerator for CQIMDQ14 is calculated. We will be looking for all women who gave birth in a given month or the 2 months prior, where there was a live or stillbirth. Let’s see which data items we use to do this. 

Slide 15

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We next look for a PREGNANCY OUTCOME of 01, 02, 03 or 04 to include only live and stillbirths.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of women who have a live or stillbirth in a given month or the 2 months prior. 

Slide 16

The final step is to create the CQIMDQ14 measure as a percentage 

Slide 17

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 18

For CQIMDQ15, we will start by looking at the denominator, the number of singleton babies who were born in the current 3 month reporting period. Let’s see how we do this. 

Slide 19

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of singleton babies born in a given month or the 2 months prior. 

Slide 20

Next, let’s look at how the numerator for CQIMDQ15 is calculated. We will be looking for all singleton babies born in the current 3 month reporting period, where the baby had a gestational age at birth between 154 and 315 days (between 22 and 45 weeks). Let’s see which data items we use to do this. 

Slide 21

We use the cohort of babies we found for the denominator, then retain only those recorded with a valid gestational age at birth – that is, between 154 and 315 days. Again, we use PERSON ID (BABY) to count unique babies in the time frame. 

To find babies with a valid gestational age at birth we look for anyone in the denominator who has a GESTATION LENGTH (AT BIRTH) between 154 and 315 days. 

This is our numerator, the number of babies who have a singleton birth in a given month or the 2 months prior, and a gestational age of between 154 and 315 days (between 22 and 45 weeks) at birth. 

Slide 22

The final step is to create the CQIMDQ15 measure as a percentage 

Slide 23

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 24

For CQIMDQ16, we will start by looking at the denominator, the number of singleton babies who were born in the current 3 month reporting period, with a gestational age at birth between 154 and 315 days. This denominator is the same as the numerator used for CQIMDQ15.  

Slide 25

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We keep only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 154 and 315 days. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of singleton babies born with a gestational age at birth between 154 and 315 days in a given month or the 2 months prior. 

Slide 26

Next, let’s look at how the numerator for CQIMDQ16 is calculated. We will be looking for all singleton babies born with a gestational age at birth between 259 and 315 days in the current 3 month reporting period.  

Slide 27

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We keep only babies which have been recorded with a GESTATION LENGTH (AT BIRTH) between 259 and 315 days. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of singleton babies born with a gestational age at birth between 259 and 315 days in a given month or the 2 months prior. 

Slide 28

The final step is to create the CQIMDQ16 measure as a percentage. 

Slide 29

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 30

For CQIMDQ18, we will start by looking at the denominator, the number of singleton babies who were born in the current 3 month reporting period, with a valid delivery method. Let’s see which data items we use to do this.  

Slide 31

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’,’2’,’3’,’4’,’5’,’6’,’7’,’8’ or ‘9’. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of singleton babies born with a valid delivery method in a given month or the 2 months prior. 

Slide 32

Next, let’s look at how the numerator for CQIMDQ18 is calculated. We will be looking for all singleton babies born in the current 3 month reporting period where the method of delivery was vaginal (excluding breech).  

Slide 33

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1, to include only singleton births.  

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’,’2’,’3’ or ’4’ for a vaginal birth (excluding breech). 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of singleton babies born with a vaginal delivery method in a given month or the 2 months prior. 

Slide 34

The final step is to create the CQIMDQ18 measure as a percentage 

Slide 35

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 36

For CQIMDQ19, we will start by looking at the denominator, the number of singleton babies who were born in the current 3 month reporting period, with a valid delivery method. This denominator is the same as the denominator used for CQIMDQ18.  

Slide 37

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’,’2’,’3’,’4’,’5’,’6’,’7’,’8’ or ‘9’. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of singleton babies born with a valid delivery method in a given month or the 2 months prior. 

Slide 38

Next, let’s look at how the numerator for CQIMDQ19 is calculated. We will be looking for singleton babies born by caesarean section in the current 3 month reporting period.  

Slide 39

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘7’ or ’8’. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of singleton babies born with a caesarean section delivery method in a given month or the 2 months prior. 

Slide 40

The final step is to create the CQIMDQ19 measure as a percentage 

Slide 41

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 42

For CQIMDQ26, we will start by looking at the denominator, the number of singleton babies who were born in the current 3 month reporting period.  

Slide 43

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We then look for where BIRTHS PER LABOUR AND DELIVERY is 1, to include only singleton births.  

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of singleton babies who were born in a given month or the 2 months prior. 

Slide 44

Next, let’s look at how the numerator for CQIMDQ26 is calculated. We will be looking for all singleton babies who were born in the current 3 month reporting period, recorded with a valid delivery method.  

Slide 45

We use PERSON ID (BABY) to count unique babies in the time frame. This is a derived field. We next look for where BIRTHS PER LABOUR AND DELIVERY is 1 to include only singleton births.  

We keep only babies whose births have been recorded with a DELIVERY METHOD CODE of ‘0’, ‘1’, ’2’, ’3’, ’4’, ’5’, ’6’, ’7’, ’8’ or ‘9’. 

We confirm the records are in a given month or the 2 months prior, by using the PERSON BIRTH DATE (BABY) and the REPORTING PERIOD START DATE for each record. 

This is our numerator, the number of singleton babies born with a valid delivery method in a given month or the 2 months prior. 

Slide 46

The final step is to create the CQIMDQ26 measure as a percentage 

Slide 47

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 48

For CQIMDQ27, we will start by looking at the denominator, the number of women with a booking appointment in the current 3 month reporting period. 

Slide 49

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field.  

We confirm the records are in a given month or the 2 months prior, by using the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and the REPORTING PERIOD START DATE for each record. 

This is our denominator, the number of women with a booking appointment in a given month or the 2 months prior. 

Slide 50

Next, let’s look at how the numerator for CQIMDQ27 is calculated. We will be looking for all women who had been recorded with valid data at their booking appointment about the number of live and stillbirths that resulted from their previous pregnancies, in the current 3 month reporting period.  

Slide 51

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We look for women where both the number of PREGNANCY TOTAL PREVIOUS LIVE BIRTHS, and the number of PREGNANCY TOTAL PREVIOUS STILLBIRTHS is between 0 and 20.  

We confirm the records are in a given month or the 2 months prior, by using the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and the REPORTING PERIOD START DATE for each record.  

This is our numerator, the number of women who had been recorded with valid data at their booking appointment about the number of live and stillbirths that resulted from their previous pregnancies in a given month or the 2 months prior. 

Slide 52

The final step is to create the CQIMDQ27 measure as a percentage 

Slide 53

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 54

For CQIMDQ28, we will start by looking at the denominator, the number of women who had been recorded with valid data at their booking appointment about the number of live and stillbirths that resulted from their previous pregnancies in the current 3 month reporting period. This is the same as the numerator in CQIMDQ27. 

Slide 55

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We look for women where both the number of PREGNANCY TOTAL PREVIOUS LIVE BIRTHS, and the number of PREGNANCY TOTAL PREVIOUS STILLBIRTHS is between 0 and 20.  

We confirm the records are in a given month or the 2 months prior, by using the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and the REPORTING PERIOD START DATE for each record.  

This is our denominator, the number of women who had been recorded with valid data at their booking appointment about the number of live and stillbirths that resulted from their previous pregnancies in a given month or the 2 months prior. 

Slide 56

Next, let’s look at how the numerator for CQIMDQ28 is calculated. We will be looking for all women who had been recorded at their booking appointment with no live or stillbirths that had resulted from their previous pregnancies, in the current 3 month reporting period.  

Slide 57

We use PERSON ID (MOTHER) to count unique women in the time frame. This is a derived field. We look for women where the number of PREGNANCY TOTAL PREVIOUS LIVE BIRTHS is 0, and the number of PREGNANCY TOTAL PREVIOUS STILLBIRTHS is 0, to identify pregnancies with no previous births.  

We confirm the records are in a given month or the 2 months prior, by using the APPOINTMENT DATE (FORMAL ANTENATAL BOOKING) and the REPORTING PERIOD START DATE for each record.  

This is our numerator, the number of women who had been recorded with no previous live or stillbirths at their booking appointment in a given month or the 2 months prior. 

Slide 58

The final step is to create the CQIMDQ28 measure as a percentage 

Slide 59

We simply divide the numerator by the denominator and multiply by 100 to get the percentage rate. 

Slide 60

Now, we look at the pass thresholds for the eight DQ measures supporting CQIMVBAC. A provider must pass all the data quality measures to be included in the published data. The provider’s rates for CQIMDQ14, 15 and 16 must all be 70% or more.  

If the denominator of CQIMDQ14 is 0 and the numerator is greater than 0, then the provider will pass that DQ measure. 

Slide 61

The provider rate for CQIMDQ18 must be 40% or more; the rate for CQIMDQ19 must be 50% or less; the rates for CQIMDQ26 and CQIMDQ27 must be 70% or more; and the rate for CQIMDQ28 must be between 20% and 70%. 

We will see how the values reached by the DQ measures affect what is shown in the published data. 

Slide 62

Let’s look at the data published for October 2021. 

R1F – Isle of Wight NHS Trust – has passed all eight of the DQ measures associated with CQIMVBAC. R1F reached 84.3% for CQIMDQ14, 97.7% for CQIMDQ15, 92.9% for CQIMDQ16, 67.4% for CQIMDQ18, 32.6% for CQIMDQ19, 100.0% for both CQIMDQ26 and CQIMDQ27, and 53.3% for CQIMDQ28. 

This means that CQIMVBAC figures are published for R1F, and this provider will contribute to the national and sub-national figures included in the October 2021 published data. 

Slide 63

Now, we will look at another trust using the October 2021 data. 

R1H – Barts Health NHS Trust – has passed five of the DQ measures associated with CQIMVBAC. R1H reached 86.9% for CQIMDQ14, 93.6% for CQIMDQ16, 75.9% for CQIMDQ18, 23.8% for CQIMDQ19 and 100.0% for CQIMDQ26. It failed three DQ measures, CQIMDQ15 with 63.8%, CQIMDQ27 with 62.2% and CQIMDQ28 with 15.9%. 

This means that CQIMVBAC figures are not published for R1H. Instead, R1H is shown with a denominator and numerator of ‘0’ with the rate as ‘Low DQ’ for CQIMVBAC, and R1H will not contribute to the national and sub national figures included in the October 2021 published data. 

Slide 64

Let’s look at how applying these data quality measures affects the larger picture. 119 providers submitted data for October 2021 births. As the CQIM DQ measures span a time frame longer than just one month, 123 providers were considered for the DQ measures CQIMDQ14-16, 18-19, and 26-28. 

In October 2021: 

115 providers passed CQIMDQ14, and the same number of providers passed CQIMDQ18. 8 providers failed or did not submit any data for each of these measures. 113 providers passed CQIMDQ15. 117 providers passed CQIMDQ26.  

118 providers passed CQIMDQ16, and the same number of providers passed CQIMDQ19, while 5 did not submit any data for these two DQ measures. 

108 providers passed CQIMDQ27. 110 providers passed CQIMDQ28, recording a proportion of pregnancies with no previous live or stillbirths that was between the data quality limits. 

Note that the pass rates of these measures could alter in the future. 

Providers can access a Data Quality Submission Summary Tool designed to help them understand the validation reports provided after a submission on SDCS cloud portal.  

Slide 65

This brings us to the end of the video. 

Thank you for watching this video demonstration on how the DQ measures supporting the CQIM rate, women who gave birth to a single second baby vaginally, at or after 37 weeks, following a previous caesarean section, are built from MSDS. 

We hope you found this useful. This video is part of a suite of videos to help you understand key measures in MSDS. We value your feedback, please email [email protected] if you wish to get in touch.

Thank you.