Intended for healthcare professionals

Editorials

Quality of care through the patient's eyes

BMJ 1996; 313 doi: https://doi.org/10.1136/bmj.313.7061.832 (Published 05 October 1996) Cite this as: BMJ 1996;313:832
  1. Thomas L Delbanco, Professor of medicine
  1. Beth Israel Deaconess Medical Center, Harvard Medical School, Division of General Medicine and Primary Care, Boston, MA 02215, USA

    Satisfaction surveys are just the start of an emerging science

    As techniques to measure the quality of health care proliferate and improve, health professionals are beginning to understand that patients and their families hold unique vantage points as expert witnesses to care. For some time, hoteliers, bankers, car manufacturers, and politicians have been enthralled by information about the public's view of their services, including anecdotes, insights offered by focus groups, and data gathered from large scale surveys. Slow to turn to new fashions, health professionals are now learning that those they serve can make important suggestions as individuals, generate worthy hypotheses in small groups, and provide aggregate data through surveys that describe what is and is not going well and that can document the impact of efforts to improve.

    Unfortunately, “satisfaction surveys” are fraught with hazards. As Cohen and colleagues show (p 841), surveys may produce widely varying data depending on how questions are framed and when and how they are asked.1 Moreover, uncertainty is compounded when one problematic set of data is compared to another obtained with even slightly differing techniques or from a different population of respondents.

    Even if questions are skilfully framed, administered in a standardised way, and used on similar populations, asking individuals to evaluate their care can be problematic. Consider two patients I may see in a morning. Both are young men starting treatment with an antihypertensive drug that in some patients induces impotence. One is a highly educated, contemplative man who wants his doctor to share decision making and to inform him fully of benefits and risks. The other is hot blooded, less educated, and convinced that doctors know best. In the spirit of informed consent, I explain to both that impotence may ensue. The first man encounters no problems; the second develops impotence soon after. A patient satisfaction survey a week later asks both to evaluate my care. The first, free of adverse consequences, praises my approach. The second vilifies me, convinced (possibly correctly) that I induced impotence through my recitation of possibilities. Thus, since their responses may be confounded by both their expectations and outcomes, a summary score of “50% satisfaction” does not convey a helpful message as I try to draw on patient perceptions to improve care.

    Where the two patients will probably agree is on what I did. Whatever their expectations or outcomes, both are likely to report that I explained the possible side effects of the new drug. Such reports are far more helpful to those who want to improve care. If I ask patients whether they were told what activities they could or could not do after leaving hospital and learn that a quarter report not being told, I have collected specific, actionable, clinically important data that can inform and shape my future behaviour.2

    Although the science of learning from patients may be in its infancy, it brings exciting opportunities to clinicians, researchers, and policy makers. For clinicians, data generated from patient reports serve initially as screening rather than diagnostic tests. By analogy, knowing that a screening test for blood in the stool is positive is the first step; determining the cause requires further inquiry. When the Lothian Health Authority finds that a quarter of its patients feel they were not encouraged to ask questions about their treatment,1 the next step is to search further and find out why. A subsequent series of questions with greater specificity may help in the hunt, just as examining the intestines will help the clinician seeking accurate diagnosis and treatment for blood loss.

    For researchers, there is ample and rewarding room for inquiry. They can begin by sharpening questions to patients and their families, learning when and where best to ask them, and developing databases useful to patients, clinicians, scholars, and policy makers. Should one ask for patients' views as they leave a hospital ward or examining room or is it better to wait several weeks and give them time to reflect? How and to what degree do responses differ when gathered face to face or by telephone, by computer or by hand? Do patient reports about the processes of care correlate with clinical outcomes? Questions abound, and the rush of inquiry augurs well for the evolving science.3 4 5 6 7 8

    Those who pay for care and those who shape and change the delivery system are also drawing on patient generated data with growing eagerness. National surveys of probability samples of patients provide baseline data and benchmarks, pointing analysts toward both deficient and outstanding practices.9 10 11 In the United States consortiums of employers, providers, insurers, and patients are measuring the quality of care in urban and rural settings and are beginning to make major decisions about resource allocation based on their findings.12 Such assessments of quality now invariably include patient generated data.13

    At a time when anger and distrust seem ubiquitous in the health systems of so many countries, asking patients to report on the quality of their care may bring clinicians and those they serve closer together. We may break down barriers if we work hard to learn from patients—and then invite them to collaborate in tackling the problems they uncover.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.