|Home | About | Journals | Submit | Contact Us | Français|
An unhappy patient suggests poor quality care, but Glyn Elwyn and colleagues point out that using measures of satisfaction to assess health providers is not without problems
Modern health care is recognising, albeit with difficulty, that it is a service industry and has to pay more attention to those who use it. It may have unique features—in that it deals with high stake issues—but in common with other knowledge intensive services, it has to balance the expert skills with the expectations and experiential expertise of users. Service industries have learnt that sustained profitability stems from meaningful customer focus, collaboratively designed products and services, and positive interpersonal exchanges that management science calls “moments of truth.”1 Healthcare organisations are now keen to take patients' perspectives seriously, but it's not as simple as it may sound.
Reviews of patient priorities are consistent.2 Summarised, they state that patients assume technical competence at both professional and organisational levels. Patients admit difficulty in judging whether these assumptions are met, although they further assume that systems are in place to ensure that basic standards are maintained and that all aspects of care are safe. They expect, however, to have good access to care and be respected. Patients also report high expectations about the experience of receiving and, often co-producing, health care—that it should be timely and that their views and preferences will be considered at least equally important as those of health professionals. Patient perceptions of what constitutes high quality care are likely to be fluid and will change as they understand that performance is not uniformly high. Providing patients with a framework to help them understand a broader array of performance and quality measures would help.3
Patients' preferences do not exactly overlap with good quality care; meeting their priorities is necessary but not sufficient. Indeed, some of their wishes, particularly when uninformed or ill informed, can be detrimental. In many circumstances, rapid access to health care is necessary and important—for example, fast treatment is essential in thrombolysis and early identification and treatment of cervical neoplasia improves prognosis. However, given that access will inevitably be limited in a resource bounded system, it is unrealistic to have a service without limits. We therefore need to influence how patients choose to use health services, especially when treatments are ineffective or potentially harmful. It would be better, for example, if some self limiting illnesses—such as some musculoskeletal problems, many viral infections, and situational reactions to stresses—were less medicalised, with less drug treatment. Strategies to reduce use, such as increased education of doctors and patients or requiring patients to contribute to the cost of treatments that have limited clinical value, are being tested.
When it comes to providing care with respect and ensuring dignity—especially for people who are frail, cognitively impaired, or terminally ill—the perception of the patient or the patient's carer is of uncontested importance. However, respect and dignity are not given sufficient attention. Empowering patients—giving them voice and demonstrating responsiveness—is essential to improving these aspects of care.
Although considerable evidence shows consistently low levels of patient involvement in healthcare processes, patients' understanding of involvement may differ from that of researchers and health professionals.4 Patients' conception of what it means to be involved in their care varies widely—from being made to feel welcome, to being able to share their anxieties, to weighing the pros and cons of treatments. So we might meet some patients' perceptions of involvement without giving them information on treatment options or engaging them in decision making. Nevertheless, not informing patients risks misunderstanding and denies them an active role in self management.
Tension exists about who should have responsibility for decisions—is it the patient alone or is it a negotiated preference?5 Quality measurement must be sensitive to the complexity of tailoring actions to patients' preferences.
Lastly, not all patients are capable of understanding the risks and the benefits of clinical choices. Measurements of quality therefore cannot be limited to data on patient experience, although they should be a central element.
Commonwealth Fund surveys of patients in Australia, Canada, Germany, New Zealand, the United Kingdom, and the United States reported problems in many areas, particularly in the communication and coordination of care between professionals and organisations, provision of care to people with long term conditions, and patient safety.6 A survey of 1400 adults in the same countries found wide disparities in access and experience of care, with the US worst.7 A similar report of primary care experiences among adults found shortfalls and variation in the delivery of safe, effective, timely, and patient centred care.8 Surveys by the Picker Institute Europe show the UK has a more paternalistic approach than other countries.9 Patients report comparatively good levels of doctor-patient communication and provider continuity but low scores for choice, involvement, and information.9
Reports of patient experience are increasingly replacing assessments of patient satisfaction.10 Items that focus on specifics—such as length of wait for an appointment or amount of information provided—may not only be less prone to ceiling effects in measurement but also enable respondents to develop clearer ideas about how to improve quality.
Of concern, however, is that insufficient attention has been given to the validity, reliability, and feasibility of the new methods—information on development and testing is lacking for many assessment tools.11 12 There is often confusion about the focus of interest: is it on the individual practitioner, the microsystem (team), or the provider organisation? The aims of surveys are often poorly defined or mix potentially conflicting aims—all these problems lead to a lack of ownership, making it difficult to generate interest in the results and easy to dismiss feedback. Another desirable development would be greater standardisation. A common set of measures to assess the patient experience would aid comparisons, with the prospect that benchmarks might guide planning and commissioning. The US, for instance, has developed the Consumer Assessment of Healthcare Providers and Systems (CAHPS)—standardised surveys that ask consumers and patients to evaluate their experiences with health care.13
Patients will increasingly play a part in assessing the quality of care. Measures of patient perspectives can be used for many purposes, including educational feedback, accreditation, certification, public reporting, pay for performance, purchasing, quality control, and quality improvement.
However, using patients to assess the performance of individual practitioners is qualitatively different from using patient surveys to assess organisational characteristics. The data are by definition personal and more sensitive; in short, the stakes are higher. Thus the tools need to have a sound scientific basis and clear aims. Should the results be used in a formative framework or for summative judgments? Are they designed to help practitioners improve and, if so, how? Are the data regarded as benchmarks and, if so, will practitioners welcome this? There is a scarcity of research on these issues.12 Studies so far suggest that doctors do not always respond positively14 and that feedback, especially of less than average performance, will not necessarily lead to improvement. We need to specify the constructs on which professionals believe that patients are capable of providing valid judgments and design fair and transparent data collection processes to meet the tenets of procedural justice.12
We know enough about contractual frameworks that set goals to predict that paying doctors according to patients' scores will result in behaviour changes to maximise financial rewards. But unless we fully understand what is being measured and, critically, how to help professionals improve, we run the risk of superficial modification and gaming to attain high scores. We also need to be aware that the motivators for good practice go far beyond externally set goals (targets and financial rewards), which, although successful in the short term, often have perverse effects.15 We remain convinced, however, that over the next decade medicine will give patients' perspective more attention. Firstly, we need to respect subjective experiences, both evaluations and reports; secondly, we need to measure them, and then manage them–aiming for improvements.
This is the third article in a series looking at use of performance indicators in the UK and elsewhere.
This series is edited by Azeem Majeed, professor of primary care, Imperial College London (firstname.lastname@example.org) and Helen Lester, professor of primary care, University of Manchester (email@example.com).
We thank Peter Lapsley.
Contributors and sources: GE, SB, JH and MW relied on systematic reviews, recent overviews, and reports to write this article. All authors are actively engaged in studying both the methods and the impact of patient involvement in healthcare contexts, from the micro (dyadic interaction) to the macro (planning and policy) levels.
Competing interests: None declared.
Provenance and peer review: Commissioned; externally peer reviewed.