Collecting and returning the data

This section covers the final stages of the survey. Again, there are several tasks to consider, including distributing the questionnaires and cover letters to the final sample, sending reminder letters to those who have not responded to the initial posting, and adding carers to the sample if necessary to meet the required response rate. This is also when councils will provide any assistance they deem necessary to help carers complete the survey. Councils will also record and check all data, before submitting the completed data return to NHS England. Councils will also have to consider whether any carer’s responses indicate that their confidentiality should be breached to protect their health or safety.

The fieldwork period, which is the period when the questionnaires are distributed and collected, should begin on 1 October and be completed by the 30 November. A questionnaire, along with a copy of the covering letter and a pre-paid envelope to return the questionnaire, must be sent to each member of the final sample. Three weeks later a reminder letter must be sent to all those who have not responded, i.e. those who have neither returned the questionnaire by the return date given by the council nor contacted the council to indicate that they do not wish to participate in the survey. A second copy of the questionnaire should be included with the reminder letter.

Within this period (remember to give time for respondents to complete and return the questionnaires) councils have some latitude in how they schedule the fieldwork. Councils may choose to issue all questionnaires in one week and all reminders three weeks later. Alternatively, councils may consider developing a timetable in which questionnaires and reminders are sent out in waves, so that the workload is spread out over all the available weeks. An example timetable for issuing questionnaires and reminders in four waves is available in Appendix K.

Note that councils must send reminder letters to all those who have not responded to the issuing of the first questionnaire, even if the overall level of response is already sufficient to meet the +/- 5 per cent margin of error and therefore the confidence interval requirement. This must be done to counter non-response bias. Given that it is likely that the type of people who respond to the initial questionnaire post are different to those who would respond only once prompted by the reminder letter, adherence to this protocol is necessary to ensure consistency of results between CASSRs.

If there are insufficient responses after the reminders have been sent then councils have the option, if time allows, to draw from the reserve sample, or to generate an additional top up sample. It is recommended that councils consider this possibility when planning their fieldwork schedule.

If face-to-face or telephone interviews are administered to those who have requested them in response to the cover or reminder letters, then the following must be considered.

Interview scripts will need to be customised by the council, in the same manner as the cover letters and postal questionnaires. The person conducting the interviews should be trained and experienced. They should be briefed about the aims of the survey. It must be assured that they understand the importance of confidentiality and will ensure participant’s confidentiality. It is expected that councils already have established practices in this regard.

If an interview is held, a more active approach to establishing informed consent is needed because of the increased level of intrusion involved. Before asking any questions, the interviewer must explain the purpose of the survey and stress that the carers can decide to withdraw their consent or stop the interview at any time. The interview scripts will guide the interviewer in this regard. If the interview is face-to-face, the participant must be asked to sign the consent form (see Appendix F).

Using the unique code CASSRs will need to merge the questionnaire and administrative data and transfer them onto the data return workbook. You should read the Instructions worksheet of the data return before commencing this task.

All those who have been sent a questionnaire (the final sample) must be recorded on the data return. A response must be recorded for a questionnaire returned that includes answers to at least one of questions 1 to 27 (mandatory or optional) from the questionnaire documents supplied by NHS England. A non-response must be recorded for

• a questionnaire that is not returned
• a questionnaire returned completely blank
• a questionnaire returned containing only answers to locally-added questions and/or optional question 32 (concerning participants’ willingness to take part in further research)

If a carer has removed their unique number-code by for example tearing off the corner of the questionnaire which contained it, these respondents should still be included in your data return. You should mention the number of respondents doing this in the first box in the Validation_Tables tab of the data return. You should however still be able to answer some of the mandatory fields for these respondents such as whether a translated version was used etc. This will then enable the responses to be included in as much of the summary analysis as possible.

There may be instances in which a respondent has answered a questionnaire in a contradictory manner. If two or more boxes are ticked for a question in which only one box should be ticked, then the response must be recorded as “Unknown” (-9). All other cases, for example if a respondent has chosen conflicting answers for a question that asks them to “tick all that apply” or for two different questions, must be recorded as reported.

The CASSR should record which version of the questionnaire was used, i.e. either the original or the reminder. If the questionnaire was completed on the original version but after the reminder letter was sent out, the respondent should still be coded as replying to the original version. If the carer returns both the original and reminder version of the questionnaire, the CASSR should review which questionnaire was most completed and use this data in the data return. If the completion is the same for both questionnaires, then the reminder survey should be used.

It is at this point that councils may discover that carers have given responses suggesting that their health or safety is at risk. It is at the council’s discretion to decide what should trigger a breach of the confidentiality clause. However, it is acceptable to do so in circumstances where:

The respondent indicates that their health and safety is at risk, in response to question 10, any other survey question or though free text comments.

The responses of the individual indicate that they are experiencing a serious issue with any aspect of their care.

The respondent uses the questionnaire to make a comment about an aspect of their care and the comment shows a clear expectation that this will be followed-up by the council

Whilst the most negative response to question 10 would clearly indicate a threat to safety you might also consider that a combination of answers to other questions might constitute grounds for contacting the survey participant.

Contact must be made with the respondent in the first instance to check whether they are happy for their comments to be passed on to an appropriate member of staff for consideration. Initial contact should not be made by a social worker or the person’s care worker. Once consent has been obtained CASSRs should follow local protocol for dealing with the issues raised.

The data return must be checked prior to submission. You should be sure that all survey data are included; worksheets have not been corrupted by the addition or removal of columns; there are no blank rows in the main data worksheet; and all data correctly represent administrative records and questionnaires as appropriate. Even the deletion of unused columns (pertaining to voluntary fields and optional questions that councils have chosen not to record) will corrupt the data return.

The workbook contains features to help quality assure the data before submission to NHS England. Conditional formatting has been applied to the cells in the Carers_Data worksheet to show where data are missing or may not be correct. The rules for these formats are detailed on the Carers_Data - Formats worksheet. You should also review the Validation_Tables worksheet which displays the results of aggregate checks on data input on the Carers_Data worksheet. If a check yields a query for you to investigate, for example if data is missing or if a result falls outside its expected range, then the appropriate cell is shaded red until an explanation is input.

Councils must take care to avoid deleting or overriding these validations, especially if cutting and pasting data from other workbooks.

Once the data has been sent to NHS England, a validation report will be sent to the council. Further information on this will be sent to council contacts. 

The completed data return must be submitted to NHS England by the first mandated deadline, 29 February 2024. 

The data will be collected by the data collections team using the Strategic Data Collection Service (SDCS), which is a secure data collection system. The team will contact those registered to upload the data with instructions on the submission procedure in due course.

As already noted (see above) NHS England requires that blank copies of final questionnaires, cover and reminder letters are emailed to [email protected] by 30 November at the latest, although preferably as soon as amendments are finalised.

We also welcome comments on the survey process, for example:

• thoughts on how it might be improved
• suggestions of practice to share with others
• other questions that might be collected nationally

Some of your views can be communicated in the data return, especially the final comments box in the Survey Process Information tab.

NHS England will check the validity of data returns which have been completed and submitted by the first mandated deadline. Feedback will be provided to councils who will then be granted the opportunity to resubmit data.

Where council data contain significant data quality issues it is likely that they will be named in the data quality section of the final NHS England report to ensure that anyone using the data is aware of any potential anomalies.

If a council fails to submit by the first mandated deadline then this will also be included in the data quality report to ensure that anyone using the data is aware that it has not been validated.