This site is intended for health professionals only


Friends and family

23 February 2015

Share this article

What will this test mean for general practice and will it really show how a practice is performing?

On the wall of every waiting room at my local hospital is a small post box labeled ‘Friends and Family’. This has been a fixture of secondary care across England for over a year now. The response to the Friends and Family survey has been a key indicator in how the hospital is assessed for the past year or so and from the start of December 2014 the same method arrived in General Practice.

At its most basic it is a simple enough survey. There is one question set by NHS England – “We would like you to think about your recent experience of our service. How likely are you to recommend our GP Practice to friends and family if they needed similar care or treatment?” with five possible responses from “Very Likely” to “Very Unlikely”.

The results should be reported monthly to NHS England and also displayed in the practice. A second question allowing a free text response should be included but the results of this do not have to reported upwards.

Although the question is the same for all health providers the results will not be considered comparable. Instead they are intended to be used as a spur to the consideration of how services are delivered and how this could be improved in the future.

With such a simple questionnaire NHS England should not be making it too complicated for practices to implement and generally this is the case. There is a lot of guidance but mostly it is about applying the test fairly and consistently both across England and throughout all parts of the NHS. Most of the guidance is common sense and is designed to allow flexibility while minimising “cheating” the system.

The general aim is that as many patients as possible should be encouraged to give their views on the service that they receive. This is not meant to be a survey of a representative sample of patients but, ideally should reflect the views of them all. In practice this is not likely to happen but the general thrust behind the guidance is to ensure that there are no groups who are disadvantaged or prevented from recording their opinions.

Practices are able to include as many additional questions as they like but the Friends and Family question must be asked first, again in order to avoid bias. All responses must be anonymous but it is permitted and even encouraged to ask questions about age, sex, ethnicity and health status in order to ensure that all groups are participating in the survey.

There should be no pressure or suggestion put onto the patient when filling in survey. The guidance goes as far as to suggest that corporate or practice logos should not be used as this may subtly influence patients.

The results of the Friends and Family question will need to be submitted to NHS England and published by the practice. It is expected that this submission will occur through the calculating quality reporting service (CQRS) system although this will be a manual procedure as there is no standard way of collecting the responses.

NHS England will also publish the results of the main question nationally each month alongside the total number of responses from each practice. Comments made in the free text section will not need to be submitted but a practice may choose to publish them itself. If the practice plans to publish comments then patients should be given the opportunity to opt out of having their comments published as part of the questionnaire.

The way that the question is physically asked is not specified and multiple systems can be used simultaneously. Each has its own advantages and disadvantages. The simplest is using pen and paper in the surgery. This does not need much more than a photocopier and posting box to set up and can be implemented quickly, It does, however make it more difficult for patients who have telephoned the surgery to respond and all of the responses will need to be analysed manually.

There are a variety of electronic methods including online forms, text message based systems or installing consoles for data entry in the waiting room. These should collate the results automatically but some patients may lack the devices or confidence to use these. There is much more opportunity for these systems to collect information where patients have called the surgery on the telephone, used an online service or after a home visit.

Setting up an online form can be quick, easy and cheap using general purpose survey systems. Google Forms is completely free and other services such as Survey Monkey are also inexpensive. There are also a number of specialist providers of Friends and Family services which will do even more of the work for you and may include extra features such as submission via text message or electronic console in the waiting room although inevitably these tend to cost a little more.

The second question is compulsory so although use of a token system for patients (such as is used in John Lewis or Waitrose) would seem a sensible and practical solution it is specifically prohibited in the regulations.

Delivering to certain groups of patients may be more difficult. NHS England has published the main question in twenty different languages although practices will be responsible for translating the second and any subsequent questions that they ask. Experience from the postal GP survey suggests very low uptake of translated questionnaires – in this case made up less than a quarter of 1% of all responses. 

By far the most popular language was Polish which accounted for over half of all submissions in languages other than English.

Patients with learning disabilities or dementia should also have the opportunity to give their views to the extent that they are able. In these cases the patient may require direct assistance to understand and record their response and there should be procedures to provide this.

Similarly there is no lower age limit to the Friends and Family test. In General Practice most children attend with their parents and are likely to respond via their parents but responses should not be declined on the grounds of age.

There is a lot going on in general practice at the moment and the best option with the test is undoubtedly to keep things simple, at least to begin with. This has been introduced as a contractual requirement and there is no payment or other resource associated with the test. Nevertheless the response rate and results are likely to be closely monitored and will be publicly reported. It would seem likely that these results will be used by the CQC, possibly as part of its risk register in a similar way to the current patients survey and QOF results.

The flexibility about how the survey is delivered to patients is something that practices can take advantage of. Experimenting with different approaches to see what works and produces a high response rate can be relatively simple. Although the answers given are important it seems that it will be the number of responses that are of most interest to NHS England in the short term as a measure of how well practices are implementing the test. Ultimately the simplest options for practices are also likely to be simple for patients. It should be easy for patients to respond and the less effort that is required the better. There can be a sense of “feedback fatigue” as every restaurant bill, shopping receipt or visit to a website comes with a request for a few minutes of your time. Often these come with incentives of a free dessert or the chance to win a cash prize which most practices would struggle to compete against.

Practices are required to publish the results but this can be used in a positive way to show that patient views are valued. If patients can see that their time is worthwhile then they are more likely to participate. It may even improve the outcomes for the practice.

When only a few patients participate it is more likely that only the dissatisfied will put their views forward. Happy patients may be more difficult to persuade to fill in the forms and increasing response rate may well improve the overall result for the practice.

Ultimately incorporating the Friends and Family test is unavoidable for a practice but it can be made a routine part of what the surgery delivers.

There is even a possibility that it could prove useful.