Government encouragement for GPs to carry out surveys of patient experience has a chequered history in the United Kingdom. Financial incentives for GPs to carry out their own surveys were included in the 2004 GP contract, and in 2006 a parallel survey was sent to 5.5 million people in England so that GP payments could be linked to patients’ ability to obtain appointments with GPs. This nationally administered survey was augmented to include other aspects of care in 2008 Department of Health(2011), although it remained unpopular with GPs, partly because a flawed payment formula meant that there were large random fluctuations in the payments that GPs received (Roland et al., Reference Roland, Elliott, Lyratzopoulos, Barbiere, Parker, Smith, Bower and Campbell2009). The national GP Patient Survey continues to be administered annually in England, although the payment link was removed in 2011. In addition, in 2011, GPs in England were again given an incentive to carry out surveys in their own practices but this time with a requirement to plan and discuss the results with Patient Reference Groups, a new name for Patient Participation Groups (British Medical Association and NHS Employers, 2011).

The United Kingdom is not alone in believing that it is important to measure patient experience, with patient surveys now being widely used in many countries. A common thread shared among policymakers and clinicians is that patient experience is an important outcome of care. Hence, it is important to determine what should be measured and, more importantly, what difference measurement might make? These issues are becoming more prominent in the United Kingdom, as the General Medical Council plans to include patient feedback as part of the assessment required for periodic revalidation of doctors (Campbell et al., Reference Campbell, Richards, Dickens, Greco, Narayanan and Brearley2008; Baker et al., Reference Baker, Smith, Tarrant, McKinley and Taub2011; Campbell et al., Reference Campbell, Roberts, Wright, Hill, Greco, Taylor and Richards2011)

There are important technical issues involved in conducting surveys, including the numbers needed to obtain reliable results (Lyratzopoulos et al., Reference Lyratzopoulos, Elliott, Barbiere, Staetsky, Paddison, Campbell and Roland2011) and the way in which responses vary in different population groups – for example, by age and ethnicity (Mead and Roland, Reference Mead and Roland2009; Lyratzopoulos et al., Reference Lyratzopoulos, Elliott, Barbiere, Henderson, Staetsky, Paddison, Campbell and Roland2012). Surveys are also not the best way to judge technical aspects of care as patients’ judgement of technical competence is heavily influenced by the quality of communication in the consultation (Chapple et al., Reference Chapple, Campbell, Rogers and Roland2002), except when patients are asked about very specific aspects of care – for example, whether an elderly patient received influenza immunisation when required (Coulter, Reference Coulter2006). Other technical aspects of care are better measured using data from medical records, as in the Quality and Outcomes Framework. Despite these issues, surveys are widely accepted as methods of measuring aspects of care important to patients, including the ability to obtain an appointment, the ability to see a doctor or nurse of the patient's choice and a range of aspects of communication within the consultation (Cheraghi-Sohi et al., Reference Cheraghi-Sohi, Bower, Mead, McDonald, Whalley and Roland2006). Well-designed surveys show that patients generally have positive experiences of general practice care, are satisfied with the care they receive and have a high degree of confidence and trust in their GP (Goodwin et al., Reference Goodwin, Dixon, Poole and Raleigh2011). However, there are important aspects of care wherein a substantial number of patients do not experience the care they expect. For example, a quarter of people who wish to see a particular GP are not able to do so consistently (Aboulghate et al., Reference Aboulghate, Abel, Elliott, Parker, Campbell, Lyratzopoulos and Roland2012), and in the lowest quartile of English practices fewer than 60% of patients rate their GP as good at explaining tests and treatments and involving them in decisions about their care (Goodwin et al., Reference Goodwin, Dixon, Poole and Raleigh2011).

Simply carrying out a survey will not change practice. Therefore, it is essential to ascertain what can be done to use surveys to improve care? There is little evidence that feedback on its own or feedback with brief training improves doctors’ consultation skills (Cheraghi-Sohi et al., Reference Cheraghi-Sohi and Bower2008). The financial incentives provided to doctors in the Quality and Outcomes Framework to improve access did not have an immediate effect (Addink et al., Reference Addink, Bankart, Murtagh and Baker2011), although there is some evidence that they may have improved rapid access to care in the longer term (Schoen et al., Reference Schoen, Osborn, How, Doty and Peugh2009; Schoen et al., Reference Schoen, Osborn, Squires, Doty, Pierson and Applebaum2011). One feature of the English National GP Patient Survey is that the results are publicly available; past research suggests that, although patients make little use of information on quality of care, clinicians do take notice when they are openly compared with their peers (Fung et al., Reference Fung, Lim, Mattke, Damberg and Shekelle2008). The English government's latest strategy is to persuade GPs (again through financial incentives) to engage much more actively with patient groups to plan, deliver and act on the results of patient surveys. Patient participation groups have been around for a long time in general practice but they have largely remained the province of a few committed GPs (Nagraj and Gillam, Reference Nagraj and Gillam2011). The new payments attempt to incentivise this activity and could offer a new way of increasing patient involvement in the National Health Service (British Medical Association and NHS Employers, 2011), provided that GPs do not simply pay lip service to establishing patient groups without any real engagement.

None of these new initiatives are guaranteed to turn the conduct of a survey into care that is actually improved for patients, and indeed there is little evidence for what would really make a difference. Given the clear evidence of deficiencies in some aspects of care, especially patients’ ability to see the doctor of their choice, and their involvement in decisions about their care, research is now needed on how survey initiatives in the United Kingdom or elsewhere could be used to improve patient experience in primary care.