Liking in the physician–patient relationship

https://doi.org/10.1016/S0738-3991(02)00071-XGet rights and content

Abstract

Patients and physicians in established relationships (261 patients and their 44 physicians) were asked after a medical visit how much they liked each other and how much they felt liked, along with questions concerning patient health, physician and patient satisfaction, and the patient’s affective state. Patients were re-contacted 1 year later and asked about their satisfaction with the same physician and whether they had considered changing physicians during the year. Patients’ and physicians’ ratings indicated mutuality of liking, as well as accuracy of estimating the other’s liking for the self. The physician’s liking for the patient was positively associated with the following variables: better patient health, more positive patient affective state after the visit, more favorable patient ratings of the physician’s behavior, greater patient satisfaction with the visit, and greater physician satisfaction with the visit. The patient’s liking for the physician was positively associated with better self-reported health, a more positive affective state after the visit, more favorable ratings of the physician’s behavior, and greater visit satisfaction. Both the physician’s liking for the patient and the patient’s liking for the physician positively predicted the patient’s satisfaction 1 year later and were associated with a lower likelihood that the patient considered changing physicians during the year. Female physicians reported liking their patients more than male physicians did, and patients’ ratings of how much they felt liked corroborated this difference. Patients also reported liking female physicians more than male physicians. A number of these results remained significant even after controlling for the patient’s overall satisfaction with the medical visit.

Introduction

Typically, investigations of physician–patient communication have focused on the behaviors physicians and patients engage in, e.g. how much information is conveyed by the physician, how many questions are asked by the patient, or how much information is recalled by the patient [1], [2]. Investigators have also studied the skills possessed by physicians and patients, e.g. skills in medical interviewing, skills in nonverbal communication, and skills in effective patienthood [3], [4], [5]. Less often considered are the motives and emotional states that may influence how physicians and patients behave. In the present research, we studied how much physicians and patients in established relationships like each other, whether this liking is mutual, whether perceptions of being liked are accurate, and how liking relates to a variety of other patient and physician variables that may be both antecedent and consequent to liking. Although in a correlational study we cannot be sure of the causal pathways between liking and these other variables, establishing such relations will set the stage for future studies to examine cause-and-effect paths.

Social psychologists have extensively studied liking (interpersonal attraction) in the non-clinical situation [6], [7], [8]. But liking has barely been studied in the context of the physician–patient relationship, although concepts such as satisfaction and rapport are certainly relevant to such an investigation. Like and Zyzanski [9] found positive associations between how much the physician liked the patient and both the physician’s and the patient’s satisfaction with the visit. Hall et al. [10] also found a positive association between how much the physician liked the patient and the patient’s satisfaction, and also found that physicians had elevated liking for healthier and male patients, and that female physicians liked their patients more than male physicians did. The causal relations in these studies are probably complex; for example, liking and satisfaction are likely to have a reciprocal relation, such that more liking leads to more satisfaction and more satisfaction leads to more liking. In this article, we offer speculations about the possible causal relations between liking and other variables.

In the present study, we went beyond the existing studies in the following ways. We measured the physician’s liking for the patient, but we also measured the patient’s liking for the physician and how much each felt liked by the other, neither of which has been measured previously. We included a wider range of potential correlates of liking, including sociodemographic characteristics of both physician and patient, physician and patient satisfaction, the patient’s ratings of specific physician behaviors, and the patient’s health status and post-visit affective state, as rated by both the physician and the patient. Finally, we contacted the patients a year later to ask about their satisfaction with their physician and whether they had considered changing physicians during the intervening year, in order to assess the predictive power of liking.

Patients’ satisfaction with their medical visit is already known to have many correlates relating to physician behavior, as well as to patient and physician background characteristics [2], [11], [12], [13], [14]. Because liking and satisfaction both reflect attitudes toward the physician, an important question is whether they are simply synonymous, or whether liking carries some predictive weight above and beyond the patient’s overall satisfaction level. Therefore, we took overall satisfaction into account when examining the correlates of liking.

Although in some respects the present study is exploratory, it is helpful to articulate the main research questions that were addressed.

We predicted that because physicians’ and patients’ liking of each other is likely to be mutually rewarding and reinforcing [15], their liking for each other would be positively correlated, an association we call ‘actual mutual liking’. Because people tend to like those who they think like them [16], we predicted that each party’s liking for the other would be related to how much each felt liked by the other, which we call ‘assumed mutuality of liking’. Finally, ‘accuracy of perceived liking’ was the degree to which one party’s feeling of being liked (e.g. “doctor likes me”) was related to how much that party was actually liked (e.g. “I like patient”). Although previous research suggested that physicians’ liking for their patients was correlated with patient characteristics [10], it was not known whether varying degrees of physicians’ liking would be evident to their patients; similarly, if patients vary in how much they like their physicians, it was not known whether these variations would be expressed clearly enough for physicians to pick up. So we made no prediction about accuracy of perceived liking.

We predicted that physicians would like their patients more when the patients were healthier, as found by Hall et al. [10], reported greater satisfaction with the physician, as found by both Like and Zyzanski [9] and Hall et al. [10], and were in a more positive affective state (as self-rated by the patient and as perceived by the physician). These predictions are consistent with the reward theory of attraction, which states that people are more attracted to those who provide them with tangible or psychological rewards [17]. In the present case, we assumed that interacting with a more healthy, satisfied, and emotionally positive patient would be rewarding to the physician, because such patients can be expected to be more pleasant to interact with and to place fewer demands on the physician. We also predicted that physicians’ liking for their patients would be positively correlated with how satisfied the physicians were with the medical visit, because both variables reflect a positive disposition toward the patient.

We predicted that patients’ liking for their physicians would be positively correlated with their post-visit positive affect and satisfaction, as well as with their satisfaction with their physician at the 1-year follow-up; these variables all reflect a positive attitude toward the physician and should, therefore, be positively related. Similarly, we predicted that patients who liked their physicians more would be less likely to consider changing physicians over the ensuing year.

We predicted that patients would like their physicians more when the patient was in better health, following from research that finds greater satisfaction among healthier patients [18], [19], [20], [21]. We predicted also that patients would like their physicians more when they believed the physician had performed better on a variety of desirable behaviors, such as explaining things and behaving in a considerate manner, consistent with research showing that similar physician behaviors, when objectively measured, do predict patient satisfaction [2].

We predicted that female physicians would like their patients more than male physicians would, as in Hall et al. [10]. Although that study found male patients to be better liked by physicians than female patients, the fact that the physician sample was mostly male raises the possibility that this was a bias in favor of a same-gender patient. Therefore, we did not make a prediction regarding physicians’ liking for male versus female patients. But we examined the possibility that liking by the physician would be greatest in same-gender dyads.

In an exploratory manner, we also examined liking in relation to the patient’s age, income, education, ethnicity, and the physician’s age. Such relations have not previously been examined. For patients’ liking of physicians, we predicted that older patients would like their physicians more, consistent with the satisfaction literature [22].

Section snippets

Patient recruitment and sample characteristics

Individuals with diabetes mellitus type II in Kaiser Permanente, Northern California Region (KPNCR), comprised the patient sample for this study (N=261, 50% male). (Unrelated findings based on this same patient and physician sample were reported in an earlier publication [23].) Patients were recruited immediately following their medical visit by a research assistant who consulted clinic appointment data to identify patients with type II diabetes mellitus scheduled for routine medical visits at

Interrelations among liking variables

The correlations among the liking variables shed light on actual mutuality of liking, assumed mutuality of liking, and accuracy of perceived liking in the physician–patient relationship. Actual mutuality of liking was significantly positive (r=0.23, d.f.=256, P<0.0001) (correlation of “I like patient” with “I like physician”), meaning that how much the physician liked the patient was positively related to how much the patient liked the physician. Assumed mutuality of liking was much stronger

Discussion

This study found many correlates of liking between physicians and patients in an ongoing relationship. Intercorrelations among the liking variables suggested that liking is indeed part of the physician–patient relationship, as much for the physician as for the patient. First, we found that how much each liked the other was related to how much each was liked. Thus, there appears to be reciprocity in terms of liking. We can speculate that one of the many possible ramifications of this mutuality

Acknowledgements

Dr. Richard L. Street Jr. served as Action Editor for this manuscript. This research was funded by a grant from the Kaiser Permanente Innovations Program, Oakland, CA. The authors wish to acknowledge Nancy Rieser for her role in collecting the data and Natasha Frost and Karen Planeta for their help with data entry.

References (32)

  • Berscheid E, Walster EH. Interpersonal attraction, 2nd ed. Reading, MA: Addison-Wesley,...
  • Byrne D. The attraction paradigm. New York: Academic Press,...
  • Berscheid E, Reis HT. Attraction and close relationships. In: Gilbert DT, Fiske ST, Lindzey G, editors. The handbook of...
  • J.A. Hall et al.

    Physicians’ liking for their patients: further evidence for the role of affect in medical care

    Health Psychol.

    (1993)
  • Thompson TL. Interpersonal communication and health care. In: Knapp ML, Miller GR, editors. Handbook of interpersonal...
  • S. Williams et al.

    Doctor–patient communication and patient satisfaction: a review

    Fam. Pract.

    (1998)
  • Cited by (0)

    View full text