Intended for healthcare professionals

Feature Data Briefing

Are “friends and family tests” useful: agree, disagree, neither, don’t know?

BMJ 2013; 346 doi: https://doi.org/10.1136/bmj.f2960 (Published 14 May 2013) Cite this as: BMJ 2013;346:f2960
  1. John Appleby, chief economist
  1. 1King’s Fund, London, UK
  1. j.appleby{at}kingsfund.org.uk

The NHS is going to ask patients whether they would recommend their hospital or ward to others, despite public scepticism about the value of such feedback. John Appleby examines responses to similar questions asked to NHS staff

Family and friends tests—which ask consumers whether they would recommend a service or organisation to their families and friends—are an established method of obtaining feedback on performance for many private sector businesses. But how well is this sort of customer feedback suited to public services like the NHS?

The Department of Health not only thinks such feedback is suited to the NHS but has the figures to show that the benefits of the data gathered will outweigh the costs. The impact assessment shows a net monetised benefit of £29.6m (€35m; $46m).1 So, it looks like it’s worth doing—which is good, because it is being done. From this month patients attending emergency departments and patients staying overnight as inpatients will be asked whether they would recommend their hospital or ward to family or friends. The scheme is to be extended later in the year to maternity services.

Whether the public think the family and friends test is worth it is another matter. Research conducted by Ipsos MORI for the Department of Health reported that “the public did not appear convinced that this was a necessary measure—with comments about the cost of implementing, and also that they would tend to rely on GP recommendations, or other data, such as mortality rates, to decide which hospital to go to.”2

The same research also rejected what has now been adopted3 as the best way to score the results of the test—to present a single number to the public and patients. The approach takes the proportion of all patients saying they would be “extremely likely to recommend” and subtracts the proportion who responded “neither likely nor unlikely,” “unlikely,” or “extremely unlikely” to recommend, ignoring those who respond “likely to recommend.” This is similar to the “net promoter score” commonly used by private businesses and is akin to, “If you’re not with us, you’re against us.”4 Comments about this scoring method from patients and professionals canvassed by Ipsos MORI included words such as “misleading,” “unethical,” and “ridiculous.”

Of course, there is nothing to stop NHS organisations also presenting the full range of results from their family and friends tests in the way that the results of similar questions asked in national surveys of NHS staff and, indeed, the “Review your hospital” slot on the NHS Choices website5 have been presented for some years.

The latest NHS staff survey results for 2012 which ask a version of the family and friends question (“If a friend or relative needs treatment, I would be happy with the standard of care provided by this organisation”) show a huge range of responses across NHS organisations. Figure 1, for example, shows the net scores for organisations using the proposed method for patients’ responses; specialist hospitals do particularly well, ambulance trusts less so.

Figure1

Fig 1 2012 English NHS staff survey: “If a friend or relative needs treatment, I would be happy with the standard of care provided by this organisation.” Net promoter score (% of people who strongly agree minus % who neither agree or disagree, disagree, or strongly agree) by type of NHS organisation6

There are also variations by staff group (fig 2). Managers and doctors seem more optimistic about the quality of care of their organisations than ambulance staff and allied health professionals. In fact, both managers and medical and dental staff are significantly more positive than nursing staff (fig 3).

Figure2

Fig 2 2012 English NHS staff survey: “If a friend or relative needs treatment, I would be happy with the standard of care provided by this organisation.” Responses by staff group6

Figure3

Fig 3 2012 English NHS staff survey: “If a friend or relative needs treatment, I would be happy with the standard of care provided by this organisation.” Mean (95% confidence interval) responses by nursing, medical and dental, and central functions/general management staff (data from Picker Institute Europe)

The family and friends question now asked in the NHS staff survey has changed, however. The original question (asked in 2003-06) was: “As a patient, I would be happy to have my care provided by my organisation.” As figure 4 shows, results on this version deteriorated over the years to the point in 2006 where nearly a quarter of staff (out of a survey of over 125 000) said they would not be happy to have their care provided by their own organisation. The new version of the question (following two years when no question was asked) seems to show a more positive picture. Is this a result of the wording change or a sign of the NHS improving? I would have to tick the “don’t know” box.

Figure4

Fig 4 Trends in NHS staff responses to family and friends test questions, 2003-12 (surveys typically cover between 100 000 and 212 000 staff )7

Notes

Cite this as: BMJ 2013;346:f2960

Footnotes

  • Competing interests: I have read and understood the BMJ Group policy on declaration of interests and have no relevant interests to declare.

  • Provenance and peer review: Commissioned; externally peer reviewed.

References

View Abstract