Using patient surveys to improve care in general practice

Improving patient experience of NHS care is a high government priority and for the last few years the NHS has carried out a systematic programme to assess patients’ experience of primary care, surveying 5.6 million patients annually in one of the largest programmes of it’s type anywhere in the world. The results of the GP Patient Survey (GPPS) describe all 8500 general practices in England in terms of the quality of care their patients’ experience, and results for individual practices are published on the internet[1]. In addition as part of the GP contract, GPs have an incentive from 2011 to start a Patient Reference Group, carry out surveys, and discuss the results with the Patient Reference Group.

These major policy initiatives will only be of value if the results can be used to improve care, and there is no evidence that simply carrying out a survey (even a very large one) will do this. This programme therefore sets out a series of projects which will each enable primary care professionals as well as practices and PCTs to use the results of the GP Patient Survey to improve patient care. The research will focus particularly on two aspects of patient experience: communication in consultations and being able to see a doctor of the patient’s choice (as a proxy for continuity of care). The reason for focusing on these is that they are highly valued by patients, there has been little focus on them in recent years, and there is evidence that they may have declined since the introduction of the new GP contract in 2004.

At the end of each of the main stages of the programme, we will produce practical guides to practices and GPs on how they can use surveys to improve their care.

The aims of the programme are:

1. To understand how general practices respond to low patient survey scores, focusing on doctor patient communication and continuity of care and identifying a range of approaches that could be used to improve patients’ experience of care.

2. To estimate the extent to which aggregation of scores to practice level in the national survey masks differences between individual doctors. To understand how GPs respond to feedback on their own individual scores.

3. To identify behaviours that could potentially be modified in order to improve performance among GPs with poor patient-rated communication in patient surveys; to determine the extent to which patients’ responses to surveys relate to assessments of the same consultations by external raters.

4. To understand better patients’ responses to questions on communication and seeing a doctor of their choice.

5. To understand the reasons why ethnic minority groups, especially South Asians, give lower scores on patient surveys compared to the white British population. 

6. To carry out an exploratory randomised controlled trial of an intervention to improve patient experience, using tools developed in    earlier parts of the programme.

7. To investigate how the results of the GP Patient Survey can be used to improve patients’ experience of out of hours care.

Methods

Selection of practices

In two geographical areas (East Anglia and North London, and Bristol and the South West), we will select practices which are nationally representative in terms of size and deprivation. Within these constraints, we will then select 15 practices which score in the lowest quartile of GP Patient Scores on communication items on the questionnaire, and 5 practices which score in each of the top and middle quartiles.

The aims of the programme will then be delivered in a series of linked projects.

Project 1. Working with practice teams to understand responses to GP Patient Survey scores and responses to individual GP scores.

We will meet with all the practices and review the results of the GPPS survey scores, leading a structured discussion on the practice’s response to their own scores, their understanding of their meaning, and the practice’s intention to change (if appropriate). We will also explore how they would respond to having feedback on individual GPs (rather than on the practice as a whole, as at present). We will present them with hypothetical scenarios of one GP with poor scores on communication or continuity of care and ask how they would respond to this as a practice. We will ask about possible interventions which might help to improve the performance of low scoring GPs.

Project 2. Re-survey of selected practices using a modified GP Patient Survey to get scores for individual GPs

We will conduct a survey of 50 consecutive patients consulting GPs in our selected 25 practices. There are three aims to this part of the study:

1      To provide scores for individual doctors. This will allow us to estimate the extent to which combining scores at practice level in the national survey masks differences between individual doctors within practices.

2      We will send some patients a second questionnaire four weeks after the first one. This is in order to establish the test-retest reliability of the survey.

3      We will interview individual GPs to understand how they respond to feedback of their own scores.

Project 3. Analysis of tape recordings of consultations.

We need to be sure that survey responses do identify aspects of behaviour that could be modified to improve the performance of doctors with poor scores. So, for example, if poor ratings of ‘listening to the patient’ were always related to not giving the patient enough time, the correct intervention might be to provide longer consultations, rather than trying to listen more carefully in very short ones.

We will carry out this part of the programme by tape recording consultations that patients subsequently rate using a modified GP Patient Survey. We will do this in the 15 selected practices with low GPPS scores, in order to maximise the likelihood of identifying consultations which patients subsequently rate low on the modified GP Patient Survey. We will invite GPs in these practices to have consultations taped. From this, we will then have a set of questionnaires which range in scores, and a set of taped consultations associated with these questionnaires.

We estimate that we will need to videotape 473 consultations from 52 GPs in order to select a sample of 28 high scoring and 28 low scoring consultations for more detailed analysis, looking at how the responses to the questionnaires relate to observable behaviours on the tapes. Experienced GP raters will be used to score the videotapes using established consultation rating scales.

Patients will be asked for consent at two separate stages: once to see if they agree to have their consultation videotaped (this is commonly done when training young doctors) and a second time after the consultation.

Project 4. Interviews with patients

We will select up to forty patients with a range of scores for more detailed interview about their questionnaire responses. The aim of this is to understand some of the complex issues that lie behind the way in which patients complete questionnaires. The interviews will be assisted by playing the consultation back to the patient to allow them to identify aspects of the consultation which led to particular responses on the questionnaire.

Project 5. Understanding the needs of patients from ethnic minority groups, and their responses to patient questionnaires.

Patients from ethnic groups rate their care substantially less positively than white British patients. The effect is particularly marked for Asian patients. This effect is found in other surveys in both England and in other parts of the world. The reason for this could be because Asians receive worse care, or because they have different expectations, or because they fill in questionnaires in a different way.

Project 5a. In this project, we will interview Asian patients who give a range of scores on the GP patient survey, using a similar approach to project 4. Interviews will be offered in ethnic minority languages.

Project 5b. We will give Asian and white British patients a DVD with a small set of consultations with actors for them to rate. Because of the requirement to show an identical DVD to all patients, the simulated consultations will be conducted in English. We will send out 4400 questionnaire / DVD sets in all to look at difference in responses between Asian and white British patients.

Project 6 – Exploratory trial of a real-time feedback intervention to improve patient experience in general practice (Improve-RTF)

The aim of this two-phase project is to explore, in a small number of general practices, the acceptability and feasibility of a real-time feedback (RTF) intervention that might have potential to inform change and improve patient experience in general practice. Practices currently scoring in the lowest 25%-50% on specific items of the National General Practice Patient Survey (GPPS) will be eligible to participate.

In the feasibility phase (January to June 2014), we will work with two practices in Devon. In the exploratory trial phase (July 2014 to February 2015) we will be working with ten practices from the South West Peninsula and Cambridgeshire.

Practices will collect anonymised RTF from patients who attend the surgery over a 12 week period, using touch-screen kiosks in the waiting area. Survey items will be drawn from the GPPS and the NHS Friends and Family Test, as well as questions chosen by each practice. Practices will receive a summary of their patient feedback every fortnight. Some will receive feedback at the team level; others will receive both team-level and practitioner-level feedback. Some practices will also receive input from an experienced facilitator who will help the practice team to reflect on their RTF results, and to discuss and plan any changes they could make to improve patient experience at the surgery. 

To assess feasibility and acceptability, we will calculate the proportion of consulting patients who use the touch-screen kiosks to leave RTF. We will also directly observe how patients and staff interact with the touch-screens, and seek the views of patients (via brief ‘exit surveys’) and staff (via team focus groups) about this method of collecting patient feedback.

Project 7.  Improving out of hours care

In this part of the programme, we will work with a small number of out of hours centres, interviewing patients to better understand their responses to the patient surveys, and working with the providers to see how they can improve the care they provide.

We will also be carrying out some more detailed large scale surveys to look in more depth at patients’ experience of out of hours care in England.

Lead responsibilities

The programme is jointly run between Cambridge and Peninsula Medical School (based at Exeter University). While all projects are run jointly and generally involve data collection at both sites, Exeter has lead responsibility for projects 1, 2, 6 and 7, while Cambridge has lead responsibility for projects 3, 4 and 5.


Core staff on IMPROVE projects

Professor John Campbell (PI)

Dr Christine Wright (Project 6 Lead)

Dr Suzanne Richards (Project 7 Lead)

Dr Heather Barry (Researcher on Project 7)

Dr Inocencio Maramba (Researcher on Project 2)

Mrs Antoinette Davey (Researcher on Projects 2,3,4 & 6)

Dr Luke Mounce (Statistics Lead)