Bmj Usa

Patient-centered communication

BMJ 2004; 328 doi: (Published 10 June 2004) Cite this as: BMJ 2004;328:E303
  1. Debra Roter, professor (droter{at}
  1. Department of Health Policy and Management Johns Hopkins Bloomberg School of Public Health Baltimore, MD

    More than a string of words

    The study by Wright and colleagues in this issue of BMJ USA (p 303) presents the perspectives and experience of a small sample of patients with breast cancer as they engaged in the care process with their physicians. As the investigators note, their conclusions depart in substantial ways from current knowledge regarding the positive effects of patient-centered communication in oncology visits. In an attempt to understand how so dramatic a departure from the literature may be explained, conceptual weaknesses of the study and its interpretation of findings—rather than limitations of the established literature—have emerged. The investigators' primary conceptual errors were equating isolated segments of the medical dialogue with meaningful representations of communication skill, and further, attributing the source of perceived relationship characteristics to patients' dependency needs rather than to communication.

    The method of inquiry and sources of data for the study were unusual. Women with primary breast cancer, ranging in disease stage from histological diagnosis after surgery to follow-up at two years, were recruited from a hospital clinic. A surgical or an oncological consultation was audio recorded and a home interview was conducted. Using excerpts selected (by the investigators) from the visit transcripts to ground the home interviews, patients were asked to comment on these aspects of communication and their relationship with the doctors involved in their care. To the investigators' surprise, patients did not focus on the communication but instead commented on enduring characteristics such as trust, expertise, liking, and respect. Because the authors could not attribute these perceptions to particular communications, they concluded that they reflected attachment needs associated with vulnerability and dependence rather than the extent to which physicians “communicated well” during medical visits. Based on this assertion, the investigators concluded that patients do not value the forms of communication that are currently thought important and generally stressed in training programs.

    While the study investigators claim to have captured patients' perspectives on the communication process, they fail to recognize their own bias in two ways. First, by selecting the transcript segments for discussion themselves, rather than having patients identify critical communication events, they weakened the validity of the study's claim of examining the communication of the visit from the patients' perspective. Second, and even more disappointing, by selecting isolated dialogue segments and representing them as proxies for communication skill, the investigators reduced communication to a level of simplicity that undermines the richness and complexity that characterizes the medical dialogue.

    Patients construe meaning from what their doctors actually say, but they also derive meaning from the way in which the message is conveyed—that is, the meta-messages communicated through voice tone, facial expression, and body cues. Communication is further understood by the motivation attributed to the messenger for delivery of the message. An illustration of multiple communication levels can be found in a study in which audiotapes of medical visits were electronically filtered to separate the channels that convey vocal characteristics (voice tone analysis) and literal content (transcripts reflecting words used).1 Raters blind to the substantive content of the audiotapes judged the voice tone of more informative physicians as conveying more concern, seriousness and emotional engagement than less informative physicians. Further-more, the informative visits were judged as more satisfying, and the physicians viewed as more conscientious, interested, and dedicated than other physicians. The study concluded that information delivers more than just substantive meaning; it is the gestalt of the literal, conveyed, and interpreted meaning that leads to patient satisfaction evident in the study.

    Within this context, the important role of skillful communication to oncology patients may be viewed. A number of studies have found significantly positive associations between physician informativeness and partnership in decision making and patients' psychological and physical adjustment to a cancer diagnosis, as well as patient satisfaction with care.24 It is unlikely that the positive effects evident in these studies were simply a result of the choice of treatment that was made or particular facts that were conveyed.5 Still other studies have related the impact of oncologists' positive, patient-centered communication behaviors, both verbal and non-verbal, to measures of patient satisfaction.68 Satisfaction measures often include items reflecting trust, confidence, respect, and judgments of competence.9

    Although not specific to cancer visits, the results of a meta-analysis of patient-physician communication studies similarly support the powerful association between communicated informativeness, interpersonal sensitivity, and partnership with patient satisfaction, compliance, and recall of medical information.10 More recent research finds that patient satisfaction is also linked to the degree to which patients like their physicians and are liked by them.11

    It is unlikely that isolated segments of the dialogue in any of these studies would produce positive attributions or outcomes. Rather than urge physicians to abandon patient-centered communication skills, I would encourage a fuller appreciation for the power of these skills to create impressions that far exceed the literal content of the messages.


    • Conflict of interest None declared.

    • Paper p 303