ECDS DQ Dashboard current view guidance

The pages on this dashboard offer Site and Provider Views (described below) of data quality information for the Emergency Care Data Set (ECDS) submitted into the Secondary Uses Service (SUS). 

Site views

Site views are based on aggregate activity counts from the perspective of the provider-site, focusing on completeness and validity. The relationship between provider and its site(s) is determined by the record level provider code and site code as submitted into the Secondary Uses Service (SUS).

Provider views 

Provider views are based on aggregate activity counts from the perspective of the provider, focusing on completeness and validity.

Higher level information by Integrated Care Board (ICB) and NHS England Region, such as counts and percentages, have been derived based on the geographical relationship that exists between provider and ICB as defined and published by the Organisation Data Service (ODS).

The data used on this dashboard has been sourced from the SUS. The dashboard is usually updated twice weekly, you can check the ‘Introduction’ front page to see what data is currently contained within the report. When it’s updated, activity for the current activity year from the 1 April up to the extraction date is extracted from SUS. This is then processed overnight and published on the ECDS DQ Dashboard – Current View the following day.

Activity data presented on this dashboard for the previous activity year is frozen and based on data that was extracted from SUS for that year as at March’s post reconciliation inclusion date.

Provider codes on this dashboard are based on values submitted in the CDS data item ORGANISATION IDENTIFIER (CODE OF PROVIDER). Where providers may have submitted a site code in ANANA (Alpha, Numeric) format instead of the required provider code, this has been identified and mapped back to its parent provider during data processing.

Site focus views on this dashboard are based on values submitted in the CDS data item ORGANISATION SITE IDENTIFIER (OF TREATMENT)

Underlying data feeding the dashboard from the provider level (from page 5 Provider Focus) has been suppressed at the activity date level of reporting. Suppressed rows are not incorporated into higher levels of geographical aggregation there-after for example ICB, Region and National. The suppression removes records where the total record count against an activity date, department type (Dept Type) and data item for a given provider is between 1 and 7 inclusive.

For the Site Focus, Site Time-Series and Site by Provider Comparison pages the suppression removes records where the total record count against an activity month, department type (Dept Type) and data item for a given site is between 1 and 7 inclusive. Suppressed values are not incorporated into higher level calculations on these pages for example Provider and National.

This difference in suppression, with one applying at a date level and the other at a month level, can explain why total record counts for the same department type, data item and time-period can be different when viewed at the site level versus the provider level.

Valid data item entries, depending on the data item and in conjunction with the corresponding validation rule, are when national standard values, designated default values and SNOMED CT codes used are listed in the current version of the NHS Data Dictionary and the latest ECDS Enhanced Technical Output Specification (ETOS), or where appropriate dates, times or datetimes have been used.

Where national codes have been excluded from the analysis of invalids, this will be explained in the help text of the Data Item Help page, towards the end of the Power BI report.

Default values published in the NHS Data Dictionary, for use against specific data items, generally have not been classed as invalid. However, in some cases (for example ETHNIC CATEGORY) default values can be classed as invalid where they indicate no attempt has been made to obtain a relevant value for a data item.

Please check the Data Item Help page for further information relating to specific data items and the validation criteria that has been applied to them.

Data slicers at the top of the dashboard pages are synchronised across two or more pages. For example, if you select a provider on the Provider Focus page (5) then the slicer on the Provider Time-Series page (6) will also change to show information for that provider too. The synchronisation works both ways between pages that are slicer synchronised and is applied either to all slicers (Site, Provider, Is UTC, Activity Year (Yr.)  and Month, Dept. Type, Data Item, Grouping and In DQMI slicers), or only slicers where it is applicable to do so. Not all pages are synchronised with each other. The pages are slicer synchronised are detailed below. 

The Injures Subset button, which appears on multiple pages, is also a slicer and is synchronised between pages as shown above.

Tables can be sorted (ascending/descending) as required by clicking on the column headings.

The width of table columns can be altered by clicking on the vertical column border in the heading (left mouse button) and dragging left or right.

Drill-down capability is available on the Focus and Time-Series tables so that more granular counts can be observed. The Provider Focus and Provider Time-Series tables allow drill-down to activity month and activity dates within the activity year. The Site, ICB, Region and National Focus tables allow drill-down to activity month only.

On the Focus pages, drill-down starts at the data item level, then by department type then by activity month and activity date (for Provider pages only). What you see on the drill-down table will be based on what has been selected in the slicers. The slicer selections always drive the data that you see in the tables beneath them, so users should always be aware of their slicer selections when viewing the tables.

Total records (under the Provider Total column) against data items and subsequent levels are true activity totals. Total records for some data items may be different due to the nature of the data quality rule being applied. When looking specifically at data items the highest total will correspond to the total number of records for an organisation or level, and often this will be the same for several data items. There are many filters that are applied when deriving the denominators for ECDS data items and these are what drive the differences in the total number of records. More information relating to the applied filters and the validation criteria can be found on the Data Item Help page.

Unless otherwise stated the denominator in the percentage calculations is always the corresponding Total, i.e., for Provider, Region, or National levels against each data item and as per selections made via the slicers.

The Overall summary row at the bottom of the table is a sum of all the record counts (for total records, missing and invalid) against each data item and is NOT an activity count. This information is used when calculating overall data quality indicators based on the selections made.

To avoid overcomplicating the visuals on the time-series table and charts, the department type and data item levels have been omitted. The percentages shown are based on aggregated underlying data as per the slicer selections.

The Time-Series demonstrates how the completeness and validity of data items has changed during the rolling 13-months to date. A sudden change could indicate a potential issue with that data item, as well as the month in which it occurred. Each month's data is an up-to-date snapshot when the data was extracted from SUS for the current activity year, as outlined in the Data Source section. Data from the previous activity year is taken from the final (Month 13) position taken on March’s post-reconciliation inclusion date for that year. If a provider re-submits data for the previous activity year after the post-reconciliation inclusion date for that year, then the corresponding data on the time-series chart will not be updated.

The Provider Time-Series (page 6) only allows one provider to be selected at a time. The Site Time-Series (page 3) allows all providers and all sites to be selected. On the latter the default position, (for example, All selected across all slicers) represents the national picture for all the data items covered by the dashboard, hence the single line. The line will diverge into selected providers and sites once the ‘Provider(s)’, ‘Site(s)’ or ‘Is UTC’ slicers are applied.

Green indicates that the % Valid is greater than or equal to 95%.

Amber indicates that the % Valid is less than 95% AND greater than or equal to 70%.

Red indicates that the % Valid is less than 70%.

The same logic applies to the % complete where it is displayed on the Comparison and Time-Series pages.

The Comparison pages allow users to compare the overall data quality for selected activity months, department types and data items, between organisations and their corresponding Integrated Care Board (ICB), Region and National levels of geographical aggregation. When two or more department types are selected, the comparator percentages (ICB, Region, National) may not be the same for all providers within a grouping. This is because some providers may not carry out all the selected activity and their comparator percentages are calculated based on the activity they have actually carried out. The same principle applies to the other filters such as activity month. A provider missing activity months or days’ worth of data will have a Region and National comparator to reflect this.

For example, when 01 and 03 department type is selected on the Provider Comparison page, a provider might only have carried out 01 and not 03 types of activity or vice-versa, where-as some may have carried out both (those that carried out neither will not be listed). When this happens the comparators at higher levels of geographical aggregation take this into account and are only reflective of the actual activity being carried out by a provider (for example type 01 activity only), and NOT all the department types selected. The Dept. Type Activity column shows the types used in the percentage calculations on each table row, and this enables a like for like comparison to be made.

All..% valid an..% complete columns on the dashboard tables (and charts) (for example, Provider % Valid, ICB % Valid, Region % Valid, National % Valid) are calculated dynamically, depending on what selections have been made in the selection slicers. The National rate will change based on your slicer selections. So, for example, where a provider has type 01 and 03 activity and both are selected in the slicer, the National % Valid for the data item Procedure (SCT) – First might be 75.3%. But when you change the slicer to only show type 01 activity the National % Valid might change to 78.6%. The calculations are dynamic and potentially all could change when you change the slicer selections at the top. This is an important point to understand when using the dashboard. It means that the comparative rates are directly comparable, it always compares like with like.

Buttons appear throughout the report, mainly at the bottom of each dashboard page below the Validity chart.

Clicking on the information button invokes a pop-up box containing further information that is relevant to the page on which it is located.

To close the box, click on the back button.

Clicking on the reset button will reset the page to its original published state. This means that all slicers will be reset back to the default All state (or first in the list where only a single selection is available), any drill downs or selections will be cleared and any column ordering in tables reset to the first column ascending alphabetically. The Injuries Subset button will also return to its Not Selected state.

The Injuries Subset button can be found on most of the dashboard pages, next to the reset button.

The injuries subset includes data items that are considered key to analysing injury data within the ECDS. The use of the injuries data in national policy development is becoming increasingly important and by highlighting these key fields and displaying their overall data quality we hope to drive improvements which will lead to better informed policies and positive impacts for patients.

This button has three states as shown below:

The default published state is Not Selected, the filter for data items included in the injuries subset has not been applied (see above left). Moving the mouse icon over the button (in either state) will activate it and it will change colour as shown above (middle). Clicking on the button, from the default state, will render the injuries subset of data items Selected (see above right).

Upon selection a filter will be applied to the dashboard page and pages that are slicer synchronised with it (see Data slicer synchronisation section). Once selected, only the data items included in the injuries subset are visible and selected on the dashboard page. The Data Item slicer and associated grouping and in DQMI slicers will show all selected.

To remove the injuries subset filter, click on the button in its Selected state or click on the reset button to reset the page to its original published state.

If data item related filters have already been applied, via the ‘Data Item’, ‘Grouping’ or ‘In DQMI’ slicers, before the injuries subset filter is applied then only the data items in the injuries subset of the existing selection will appear in the table. For instance, if the Clinical grouping has been chosen from the ‘Grouping’ slicer, then once the injuries subset is selected only the injuries subset of the clinical data items will be visible and applied to the dashboard page (and any other pages where the slicers are synchronised).

There are a couple of other buttons on the dashboard to note.

The help text for each data item on the dashboard explains the data quality rules and any data filters used when assessing data item completeness and validity. This can be queried and viewed on the Data Item Help page located towards the end of the dashboard. 

Dashboard navigation 

The pages of the dashboard can be accessed by either using the navigation arrows < > either side of the page numbering, for example < 3 of 12 >, or for quicker navigation click on the page numbering between the navigation arrows, < 3 of 12 >. This will bring up the page list which will allow you to jump straight to the specific content you need by clicking on it within the pop-up list that appears.

A slicer is an effective way to filter for the specific information you need. Each page has a specific set of slicers at the top of the page.

Slicer variables on each page are directly affected by selections made in other slicers on the same page. In general, it is best to work from left to right when using the slicers to avoid confusion. For instance, a provider that only carried out department types 01 and 03 activity will not have the option for 02 in the Department type slicer on the Provider focus page. Similarly, if a provider has not submitted data for an activity month, then that activity month will not be available in the Activity Month slicer. Users should be aware of what has been chosen previously to avoid confusion when making further slicer selections.

Clicking on a row item on a table or chart object will automatically focus the visuals on the selected items by dimming out the surrounding values. Multiple selections can be made by holding down the CTRL key whilst clicking. To remove the focus simply click the item(s) again.

If you experience issues when viewing or launching the Power BI dashboards, please try using an alternative browser to resolve the problems being experienced. The reports are predominantly tested using Chrome so we would recommend using this browser when viewing the reports.

To re-use the pages and visuals (tables and charts) in your own reports or presentations we recommend first taking a screenshot of the page using the Snipping Tool. This app can be found under Windows Accessories in the app list when you click on the Windows icon on your desktop's taskbar. Unfortunately, the visuals do not always render correctly when printing directly from the browser, for instance conditional formatting may not appear on the tables. Taking a screenshot of the page or visual, once a selection has been made, and pasting the image into Word or PowerPoint, as required, will maintain the integrity of the visual. Please remember to reference the source of visual if it is not obvious as Source: ECDS DQ Dashboard – Current View, NHS Digital.

It is possible to export data from a table within Power BI by completing the following steps. Firstly, select the rows in the data table you want to export by clicking on the row within the data table. To select multiple rows, hold the Ctrl key down and select as many rows as required. Then right click within the selected rows, select Copy from the drop-down menu and then either Copy value or Copy selection. You then simply paste the selection into the required application.

There is only a limited amount of space available on each page which restricts the size of the tables and charts. Focus mode can be used to expand the size of these visuals. If you click on the edge of a visual, you will see the Focus Mode (box and arrow within a box) icon appear in the top right-hand corner of the visual. If you click on this icon the visual will expand to occupy the full screen. This is particularly useful for viewing the charts in more detail or viewing table row data that would otherwise be only viewable using the scroll bars on the dashboard.

If you click (left mouse button) on the edge of a visual, you will see the More Options (. . .) icon appear in the top right-hand corner of the visual. Clicking on the icon (left mouse button) will reveal further options where you can show the summarised data behind the visual, sort the data in the visual or spotlight the visual so it stands out from other objects on the dashboard.

We would welcome any feedback users might have about this dashboard, good or bad. If you have any suggestions on how it could be improved to better meet your requirements, please email [email protected] marked for the attention of the CDS DQ Dashboards team. Thank you.