Patient Education and Counseling 48 (2002) 217–224
Do patients talk differently to male and female physicians? A meta-analytic review Judith A. Halla,*, Debra L. Roterb a
b
Department of Psychology, Northeastern University, Boston, MA 02115, USA Department of Health Policy and Management, The Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD, USA
Abstract A meta-analytic review was undertaken of seven observational studies which investigated the relation between physician gender and patient communication in medical visits. In five of the studies the physicians were in general practice, internal medicine, or family practice and were seeing general medical patients, and in two of the studies the physicians were in obstetrics–gynecology and were seeing women for obstetrical or gynecological care. Significant findings revealed that, overall, patients spoke more to female physicians than to male physicians, disclosed more biomedical and psychosocial information, and made more positive statements to female physicians. Patients also were rated as more assertive toward female physicians and tended to interrupt them more. Several results were weaker, or even reversed, in the two obstetrics– gynecology studies. Partnership statements were made significantly more often to female than male physicians in general medical visits but not in obstetrical-gynecological visits. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Literature review; Meta-analysis; Physician gender; Physician–patient communication
1. Introduction Speculation that physicians’ communication may be a quality of care pathway has grown over the past decade and has enhanced the centrality of communication to research and training efforts [1–3]. Within the context of gender, disparities in patient referrals for major diagnostic and therapeutic interventions, as well as preventive services, have drawn attention to possible differences in the content of medical recommendations made to male versus female patients [4,5]. More recently, however, quality of care markers have been broadened from the content of recommendations to the dynamics of the therapeutic relationship. In this regard, debate on whether female physicians provide a more intense therapeutic milieu with their patients than male physicians—one that would allow for more open exchange and collaboration, and ultimately more comprehensive diagnosis and treatment—has gained attention [6–8]. Much of our own investigation and interest in this area has been devoted to understanding how male and female physicians communicate, especially when with patients of the same or different gender [9–11]. * Corresponding author. Tel.: þ1-617-373-3790; fax: þ1-617-373-8714. E-mail address: hall1@neu.edu (J.A. Hall).
One might argue that the traditional focus on physician communication fails to appreciate the influence of patients in shaping the doctor-patient relationship. In fact, discussions of gender effects in medical communication have virtually ignored the question of how patients behave toward male versus female physicians. We believe this is an important question, however, because it shifts a largely physiciancentric view of communication to one that better appreciates the reciprocal and dynamic elements of both patient and physician in the medical interchange. In the present article we address how patients behave toward male and female physicians based on the limited amount of published evidence that we were able to locate. In spite of the relatively small database, some clear findings emerge. 1.1. Why expect patients to communicate differently with male versus female physicians? There are good reasons to expect that patients may behave differently toward male versus female physicians. First, there is evidence from non-clinical studies that people treat men and women differently in conversation; for example, people gaze and smile more at women than at men, approach women more closely, and self-disclose more to women
0738-3991/02/$ – see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 7 3 8 - 3 9 9 1 ( 0 2 ) 0 0 1 7 4 - X
218
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
[12–14]. Thus, women seem to be a different kind of stimulus than men are in social interaction and this may apply in the clinical situation as well. Second, to the extent that male and female physicians communicate differently, one would expect reciprocal behavior patterns in patients [15,16]. Behaviors such as gazing, smiling, posture, a variety of speech behaviors, and the emotional tone of one’s communication are typically reciprocated or matched in social interactions [14,17–19]. For example, if your doctor’s voice sounds angry or anxious, yours will tend to sound that way too [17], and if someone smiles at you, you are likely to smile back [13,20]. The reciprocation of affectively toned verbal and nonverbal behaviors is one of the chief mechanisms underlying the operation of interpersonal expectancy effects (self-fulfilling prophecies) in interpersonal interaction [21,22]. Therefore, behavioral differences between male and female physicians could produce corresponding gender differences in patients’ behavior directed back at them. It happens that there are, indeed, behavioral differences between male and female physicians. In a meta-analysis based on studies using objective observations, Roter et al. [11] determined that female physicians conducted longer visits than male physicians and engaged in significantly more active partnership behaviors, positive talk, psychosocial information giving and question asking, and emotionally focused talk. Female physicians also displayed more positive nonverbal behaviors than male physicians. These effects were especially notable in studies of general medicine practice (internal medicine and family practice) and were, interestingly, sometimes reversed in direction in the two available studies of physicians specializing in obstetrics–gynecology. In the main, the differences between the communication styles of male and female physicians correspond well with gender differences in communication that are extensively documented in non-clinical populations. Compared to men, women have been shown to be more emotionally expressive in both words and nonverbal behavior, to engage in more positive and engaged nonverbal behaviors (such as smiling, nodding, and gazing at a partner in conversation), to engage in more self-disclosure, and to be more egalitarian in interpersonal relations [12,14,23,24]. Thus, it appears that the selection and socialization processes impinging on male and female physicians are not strong enough to erase the pervasive effects of gender-role socialization. Because the communication behaviors exhibited more often by female physicians appear to be associated with positive patient effects in the form of satisfaction and clinical outcomes [25,26], it has been speculated that female physicians create a more favorable therapeutic milieu than do male physicians [6]. By examining how patients behave with female versus male physicians, we can help to flesh out our understanding of how those milieus might differ. The reciprocity principle would lead us to expect that patients treat their male and female physicians in much the same way as they are treated
by them, which is to say that patient behavior toward male versus female physicians should parallel how male and female physicians themselves behave.
2. Methods 2.1. Search procedure and criteria for study inclusion To be included in the review, a study had to meet the following criteria: (1) involve physicians, physicians in training (interns or residents), or medical students; (2) measure communication using neutral observers (including simulated patients as observers), audiotape, or videotape; (3) test for an association between physician gender and at least one patient communication variable; (4) deal with nonpsychiatric medical visits; and, (5) be published in an English-language book or journal. Studies of both actual and simulated patients were included. Studies were identified through the following search methods: on-line database searches (MEDLINE 1967– 2000, AIDSLINE, PsycINFO, and BIOETHICS) using the keywords ‘‘doctor–patient interaction; patient–patient interaction; physician–patient interaction; doctor–patient relationship’’. These keywords were combined with other keywords: female; gender effects; female physicians; female doctors; effect of sex of doctor. In addition, a hand search was conducted of our own reprint files and the reference sections of review articles and other publications. One of the articles in the dataset was recently published by one of the present authors (A3), with the relevant effects being described simply as ‘‘nonsignificant’’; but since the database was available the actual test statistics were run and included in the present review. Study A2’s results were reported in two separate publications (both shown in the Appendix A). Studies included in the meta-analysis are listed in Appendix A. 2.2. Description of studies All of the studies were conducted in an outpatient setting. The seven studies in the database had the following characteristics: Study A1: 11 internists (6 male, 5 female), 60 general medical patients (both male and female, but gender breakdown not reported). Study A2: 50 internists (25 male, 25 female), 100 general medical patients (50 male, 50 female). Study A3: 21 obstetricians (11 male, 10 female), 82 obstetrical patients. Study A4: 127 internists and family practitioners (101 male, 26 female), 537 general medical patients (228 male, 309 female). Study A5: 16 general practitioners (8 male, 8 female), 405 female patients in general medical visits.
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
Study A6: 21 gynecologists (13 male, 8 female), 303 obstetrical or gynecological patients. Study A7: 21 family practitioners (17 male, 4 female), 20 patients in general medical visits (9 male, 11 female). 2.3. Analytic approach 2.3.1. Behavior groupings Patient behaviors were conceptually grouped according to principles developed in earlier meta-analyses of communication in medical visits [11,26]. Categories summarized in the present study were total amount of talk, biomedical information giving, psychosocial information giving, question asking, partnership building, social conversation, positive talk, negative talk, emotional talk, interruptions, orientations, and global ratings of affect. Five of the seven studies used the Roter Interaction Analysis System [27] as the instrument for coding verbal behavior (A2–A6). 2.3.2. Quantification of results To quantify the direction and magnitude of the behavior differences, we used the index of effect size called Cohen’s d, which is defined as the difference between two means (in this case, the mean behavior directed toward male physicians subtracted from the mean behavior directed toward female physicians) divided by the pooled within-group standard deviation [28–30]. A positive d means that patients engaged in more of the behavior with female physicians than they did with male physicians, and a negative d means the reverse. In the studies to be summarized, Cohen’s d was never reported directly but was rather calculated by the present authors from the published information using standard formulas (e.g. means and standard deviations, Pearson correlations, t-test, or F-test) [30]. Average ds were calculated on an unweighted basis across those studies that
219
permitted calculation of d, as well as on a weighted basis according to the sample size of patients. In addition to Cohen’s d, a standard normal deviate (Z), the statistic associated with a P-value (for example, the Z associated with P ¼ 0:05, two-tail test, is 1.96) was derived for each result, and these were used to calculate a combined P-value [30]. If an author reported a result as ‘‘nonsignificant’’ and gave no other useful data for calculation of Z, a Z of zero was used in the calculation of the combined probability (sum of the study Z’s divided by the square root of the number of studies included in that analysis). In this way, our review captures information often embedded in null results which is generally lost, and provides us with a commonly understood probability metric to compare results across variables of interest.
3. Results Altogether, seven studies reported quantitative results on the relation of physician gender to patient communication (A1–A7). An additional study [31] indicated that patient behavior was objectively measured and showed no physician effects; however, it was unclear exactly which patient behaviors were being referred to so this study is not included in the summaries that follow. 3.1. Amount of patient talk As shown in Table 1, the four studies that reported on the total amount of patient talk (A2–A5) found that patients talked more to female than to male physicians, as indicated by both the unweighted and weighted average ds (range ¼ 0:56 to 0.75) and the combined Z (significant at P ¼ 0:001). Studies A2, A4, and A5 had statistically significant results in this direction. Interestingly, the one
Table 1 Physician gender effects on patient communication during medical visits Category
N
Unweighted average d
Weighted average d
Combined Z
P-value
Direction
Amount of talk Biomedical information Psychosocial information Question asking Partnership building Social conversation Positive talk Negative talk Emotional talk
4 6 6 5 5 4 5 5 4
0.22 0.40 0.25 0.00 0.19 0.02 0.22 0.03 0.06
0.34 0.36 0.26 0.10 0.25 0.01 0.20 0.08 0.06
3.28 4.47 3.14 0.60 1.54 0.20 2.66 0.62 0.25
0.001 0.0001 0.005 0.60 0.13 0.85 0.01 0.60 0.80
3/4 4/5 4/5 2/3 3/4 1/3 4/4 2/4 0/2
Global ratings of affect Positive Anxious Assertive
4 4 4
0.10 (3) 0.03 (3) 0.39 (3)
0.20 0.46 0.05
2/3 F > M 1/2 F > M 3/3 F > M
(4) (5) (5) (3) (3) (3) (4) (4) (2)
(4) (5) (5) (3) (3) (3) (4) (4) (2)
0.15 (3) 0.13 (3) 0.32 (3)
(4) (6) (6) (5) (5) (4) (5) (5) (4)
1.30 (4) 0.74 (4) 2.13 (4)
F F F F F F F F F
> > > > > > > > >
M M M M M M M M M
Note: P-values are two-tail. Number of studies on which a statistic is based is given in parentheses. Effect size is positive when patients directed more of the behavior to female physicians than to male physicians, negative when the reverse. ‘‘Direction’’ refers to the number of studies showing patients to direct more of the behavior to female physicians than to male physicians (F > M) out of all studies of known direction (regardless of P-value).
220
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
negative effect (meaning that patients spoke more to male than female physicians) came from the obstetrics–gynecology study (A3). Without this study, the overall effect was predictably larger (unweighted average d ¼ 0:48, weighted average d ¼ 0:41, combined Z ¼ 4:52, P < 0:0001), indicating that patients talked more to female than male physicians. 3.2. Patient information giving Information giving (measured in studies A1–A6) was divided into biomedical information giving and psychosocial information giving. There is clear evidence that patients of female physicians provided more of both kinds of information than patients of male physicians (Table 1), with highly significant combined probabilities. For biomedical information giving, ds ranged from 0.04 to 0.75, and for psychosocial information giving, ds ranged from 0.18 to 0.52. Studies A1, A2, A4, and A5 were each statistically significant for biomedical information, and studies A1, A4, and A5 were each significant for psychosocial information. Interestingly, in parallel with the findings described earlier for physicians’ communication [11], the two studies on obstetricians-gynecologists showed results far from significance for both kinds of information. With these studies removed from the calculation, the effects showing more information being given to female physicians became substantially stronger: for biomedical information, unweighted average d ¼ 0:50, weighted average d ¼ 0:40, combined Z ¼ 5:57, P < 0:0001; for psychosocial information, unweighted average d ¼ 0:36, weighted average d ¼ 0:30, combined Z ¼ 4:26, P < 0:0001.
for patients to address more partnership building behaviors to female than male physicians. The three available ds were mixed in direction ( 0.29, 0.26, and 0.61, studies A3, A4, and A2), with an unweighted average d of 0.19 and a weighted average d of 0.25. However, if the two obstetrics–gynecology studies are omitted (one of which supplied a d), the unweighted average d is 0.44, the weighted average d is 0.31, and the combined Z is 2.78 (P < 0:01). Thus, in general medical practice patients were more promotive of a partnership relationship with female than with male physicians. 3.5. Patient social conversation Four studies measured social conversation, that is, nonmedical chitchat that usually occurs at the beginning and end of the medical visit (A2, A3, A4, and A6). As shown in Table 1, there was no significant difference by physician gender. The three known ds were 0.16, 0.00, and 0.04 (A2–A4), with none coming close to statistical significance. 3.6. Patient positive talk Positive comments by the patient, including statements of agreement, were measured in five studies (A2–A6). As Table 1 shows, the overall gender effect was significant, with an unweighted average d of 0.22 and a weighted average d of 0.20. In this case, the findings were not different in the obstetrics–gynecology studies, and indeed the largest effect was for an obstetrics–gynecology study (A3, d ¼ 0:38). Thus, patients’ utterances were more positive toward female physicians regardless of visit type.
3.3. Patient question asking
3.7. Patient negative talk
Five studies coded patient question asking; for the three studies that separately reported biomedical and psychosocial questions (A2, A5, and A6), these two kinds of questions were combined before the calculations were done to be comparable to the two studies that reported only total patient questions (A3 and A4). As Table 1 shows, there were no overall effects, and individually no study reached statistical significance, though study A4 achieved P < 0:06 (d ¼ 0:18), showing patients to ask female physicians more questions than they asked male physicians. In the aggregate, however, it is evident that patients’ question asking was not related to the physician’s gender.
Patient negative talk, which included disagreements, was measured in five studies (A2–A6). Table 1 shows that there was no significant gender effect (range of ds ¼ 0:18– 0.16). Excluding the obstetrics–gynecology studies for this variable made no appreciable difference.
3.4. Patient partnership building The category of partnership building by patients reflects components of active enlistment including facilitation of physician input through requests for opinion, understanding, paraphrase and interpretations, and verbal attentiveness. Five studies coded this type of variable (A2–A6). As Table 1 shows, there was a non-significant overall tendency
3.8. Patient emotional talk Four studies measured patient emotional talk, which included statements of concern, worry, and personal feelings (A2, A3, A5, and A6). Table 1 shows that there was no evidence of a physician gender effect on patient emotional talk; this was true for both obstetrics–gynecology studies and general medical studies. None of the studies individually approached statistical significance. 3.9. Patient interruptions Two studies (A2, A7, not shown in Table 1) measured verbal interruptions, meaning an apparently motivated intrusion into the other’s speaking turn with the purpose of
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
changing the topic and overtaking the floor. These two studies showed very different results. Study A2 showed no physician gender effect (no d reported) while study A7 showed a large and significant one indicating that patients interrupted female physicians more than male physicians (d ¼ 1:46, P < 0:01). The combined Z over both studies was 1.91 (P < 0:06). Study A7 was, however, very small—only 21 patients. On the other hand, study A2 had 100 patients. It is possible that study A7 is not representative of interruptive patterns in general medical practice. 3.10. Patient orientations Orientations indicate what is expected or what might be coming next. Patients’ use of this category is often an indication of items on an agenda (e.g. ‘‘I have a bunch of questions I’d like to ask you’’) or instructions regarding procedures (e.g. ‘‘Give me a minute to get ready’’). Only the obstetrics–gynecology studies (A3 and A6) reported this behavior (not shown in Table 1). Although in one of the studies there was a significant effect such that more orientations were directed toward female physicians (A6), in the other the effect was nonsignificant and went in the opposite direction. Together the combined Z was 0.62 (P < 0:60). 3.11. Global ratings of patient communication The final category of behavior to be presented consists of global ratings made of patients’ communication by neutral observers (A1–A3 and A5). In all but study A2, observers listened or watched the entire physician–patient interaction and then made global ratings of the patient; in study A2, observers listened to short clips of patients’ speech that had been electronically filtered to obscure the verbal content. Anger-irritation was rated in study A5 and showed that patients directed less anger-irritation toward female than male physicians (d ¼ 0:22, P < 0:05). This study plus three others (A1–A3) also gathered ratings of positive affect (friendly, warm, kind; see Table 1). The unweighted average d was 0.10 and the weighted average d was 0.15 (both based on three studies, range ¼ 0:20 to 0.29) and the combined Z was 1.30 (four studies), P < 0:20. Of the four studies, study A5 showed a significant tendency for patients to display more positive affect to female physicians. Four studies reported global ratings of anxiety (or relaxation, reversed in polarity; A1–A3 and A5). There was no trend (unweighted average d ¼ 0:03, weighted average d ¼ 0:11, both based on three studies; range ¼ 0:22 to 0.14). Study (A5) found that female physicians were spoken to less anxiously than male physicians were. Ratings of assertiveness–dominance (or submissiveness, reversed in polarity) were obtained in studies A1–A3 and A5. The unweighted average d was 0.39 and the weighted average d was 0.32 (both based on three studies, range ¼ 0:11 to 0.95), combined Z ¼ 2:13, P < 0:05 (four studies), showing that patients were more assertive with
221
female than male physicians. This effect was significant (P < 0:001) in Study A1.
4. Discussion Although based on a limited dataset, the present quantitative summary was able to detect several significant trends in the behavior of patients toward their male versus female physicians. Female physicians received more positive statements in all kinds of visits, but they received more talk overall, more biomedical and psychosocial information, and more partnership behaviors to a greater extent (or only) in routine medical visits as opposed to visits to obstetriciansgynecologists. Findings for global ratings of positive and negative affect tended (nonsignificantly, overall) to concur with the result for number of positive statements. As noted earlier, a previous quantitative analysis of physician behavior according to physician gender also found that the same obstetrics–gynecology studies produced results that deviated from studies based on general medical practice [11]. In that review, female obstetrician-gynecologists did not differ from their male counterparts for a number of behaviors, and for some the trend was even reversed. It was speculated that patient preferences for female physicians has put male obstetricians at a competitive disadvantage, leading male physicians to ‘‘try harder’’ with their patients to establish a more patient-centered atmosphere than would otherwise be expected [6]. For the present review of patient behaviors to show similarly deviant results for the same obstetrics–gynecology studies is consistent with the reciprocity mechanism described above, in that the relative lack of a difference in physicians’ behavior is mirrored back in the behavior shown to them by patients. In interpreting the different results for obstetrics–gynecology, however, one might also point to the fact that such studies involve only female patients, and so the results could have less to do with medical specialty than with patient gender. However, a large study on general medical practice in the present dataset that was also based on only female patients (A5) had results that were typically similar to other studies of general medical practice and not to the obstetrics– gynecology studies. We also found evidence, in one study at least, that female physicians were interrupted while speaking more than male physicians were, and global ratings from several studies confirmed that female physicians were treated in a more assertive manner by patients than were male physicians. Behaviors that showed little or no difference were patient questions, social conversation, negative statements, emotional talk, orientations, and global ratings of anxiety. As predicted on the basis of the reciprocity principle, some of these behaviors are the same ones which an earlier meta-analysis [11] found to differentiate the behavior of male versus female physicians, specifically positive talk, psychosocial information giving, and partnership building.
222
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
However, patients also provided more biomedical information to female physicians, even though the earlier review did not find male and female physicians to differ in how much biomedical information they provided to their patients. Female physicians were found, however, to ask more questions than their male counterparts [11] and it may be this behavior, the asking of more questions, that stimulates more patient information giving rather than more information being provided by the physician. More patient disclosure of biomedical information may also be fostered by female physicians’ more active efforts to build partnership through inviting the patients’ questions and opinions [11]. Interestingly, though female physicians in the earlier review made more emotionally focused statements than male physicians (in primary care settings), the present review did not find that patients directed more emotional discussion to female physicians. Taken together, the differences that we documented in patient behavior can be considered to reflect a heightened level of comfort, engagement, disclosure, and assertiveness on the part of patients speaking to female physicians. Whether the last result stems from patients having less respect for female physicians [32] cannot be determined from these data. However, we think the more likely interpretation, taking the entire pattern of results into account as well as evidence for how male and female physicians communicate with their patients, is that patients feel more empowered in interactions with female physicians. Considering that greater patient participation, and a more patient-centered behavioral repertoire (such as possessed more by female physicians), have been positively associated with a variety of clinical outcomes [25], it seems likely that the patient behavior effects found in the present review are an indication of a relatively more health-promoting therapeutic milieu produced by female physicians. Such a conclusion can only be speculative at present, however, since no study has directly investigated whether patients of female physicians fare better on clinical measures. Indeed, such a comparison would be fraught with methodological difficulty considering the many variables that could confound the comparison. Furthermore, whether medical care translates into better clinical outcomes depends on much else besides simply whether the physician seems to be doing the ‘‘right’’ things. Patients must also respect the physician’s judgment and must be willing to follow through on the physician’s suggestions and on their own good intentions (regarding, for example, self-care, lifestyle, and medication adherence). Little or nothing is known about how male and female physicians compare on these kinds of outcomes. However, it is evident that, on average, female physicians do not win out in popularity, as indicated by a review of studies that compare the satisfaction of patients seeing male versus female physicians [33]. Some studies show patients to be more satisfied with male physicians, some with female physicians, and some show no difference. It is premature
to offer an explanation for this variation, and we can only speculate on how much patients’ satisfaction depends on things the physicians actually do versus stereotypes and expectations held by the patients, or differences in patient characteristics such as health status or sociodemographics. When patients are less satisfied with female physicians, the reasons could include prejudice and skepticism toward women in an authority role, or, paradoxically, disappointment that female physicians are not even more warm, participatory, and approachable than they already are [33]. The case of obstetrics and gynecology is again interesting and provides further evidence of patient expectations and preferences influencing satisfaction assessments, regardless of actual physician performance. In the meta-analysis referred to earlier [11], male obstetricians conducted longer visits and engaged in more dialogue than female obstetricians. They were more likely to check that they had understood the patient through paraphrase and interpretations, to use orientations to direct the patient through the visit, and to express concern and partnership than female physicians. And, in the present study, it appears that these differences were associated with much attenuated differences in how patients treated the physicians. Nevertheless, the one obstetrics study that monitored patient satisfaction (A3) found male obstetricians to be rated as less satisfying by their patients than their female counterparts, even after the explanatory power of particular communication variables for satisfaction were taken into account. These lowered satisfaction ratings seem to mimic the inconsistent evaluations of female physicians noted above and may suggest that prejudice and skepticism toward male obstetricians diminishes the positive impact of their actual performance. The challenge for a more positive transformation in the everyday practice of medicine includes not only promulgation of medical practice norms that value communication skills and interpersonal sensitivity, but also the generation of gender-neutral social norms regarding patient expectations and judgments of physician conduct. Having more women in medicine will help contribute to societal norms that do not inherently define ‘‘doctor’’ in gender-linked terms, but this will not be sufficient in itself to transform medical practice. Physician training in interpersonal skills, emphasizing those aspects of communication identified in the growing evidence base of medicine, can make the difference in defining quality standards for interpersonal communication for all physicians. Fortunately, there is ample evidence that training in communication skills is effective in changing physician performance (e.g. [1,34–36]). 4.1. Practice implications The social psychologist Robert Rosenthal said, commenting on differences in how psychological experimenters treat male and female research subjects, that ‘‘male and female subjects may, psychologically speaking, simply not be in the same experiment’’ ([37], p. 56). We could say the same when
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
gender differences in behavior are documented in medical visits. The present results, combined with earlier research on male and female physicians’ behavior, suggest that the tone and content of the medical visit can be quite different depending on the gender of the physician. Female physicians, because of patients’ preexisting stereotypes and expectations or actual differences in physicians’ behavior, or both, receive somewhat different communication from patients than male physicians. It is also likely, considering that social influence is usually mutual, that physicians, male and female alike, are in turn influenced by how their patients behave toward them. In addition to the growing evidence for physician gender effects, there is an additional body of evidence showing that patient gender is associated with communication differences as well [26,38]. It has been reported that women ask for more information and talk more when with physicians than men do [39–41]. Female patients also used more emotionally concerned statements, disagreements, and positive statements than male patients [42]. Furthermore, male and female patients are treated differently by physicians. An earlier meta-analysis reported a pattern of exchange in which female patients received significantly more information and more total communication from their physicians than male patients [26]. Individual studies show greater levels of partnership building and positive talk [43] and more emotionally concerned statements (including empathy, concern, reassurance, and legitimation) and disagreements directed to female rather than male patients [42]. The broad interactive pattern suggests that physicians are more engaged with their female patients, both cognitively, providing more information, and affectively. Recognition of these gender-related social psychological factors in the process of care could be beneficial to physicians when interacting with patients. The potentially powerful impact of reciprocation of behavior style and affect between parties in the medical visit is especially important to recognize, as such recognition could help to create positive exchanges and defuse negatively spiraling interaction patterns.
Appendix A. Studies in the meta-analysis A1: Charon R, Greene MG, Adelman R. Women readers, women doctors: a feminist reader-response theory for medicine. In: More ES, Milligan MA, editors. The empathic practitioner: empathy, gender, and medicine. New Brunswick, NJ: Rutgers University Press; 1994; 205–21. A2: Hall JA, Irish JT, Roter DL, Ehrlich CM, Miller LH. Gender in medical encounters: an analysis of physician and patient communication in a primary care setting. Health Psychol 1994;13:384–92, and Irish JT, Hall JA. Interruptive patterns in medical visits: the effects of role, status and gender. Soc Sci Med 1995;41:874–81.
223
A3: Roter DL, Geller G, Bernhardt BA, Larson SM, Doksum T. Effects of obstetrician gender on communication and patient satisfaction. Obstet Gynecol 1999; 93:635–41. A4: Roter D, Lipkin Jr M, Korsgaard A. Sex differences in patients’ and physicians’ communication during primary care medical visits. Med Care 1991;29:1083–93. A5: Van den Brink-Muinen A, Bensing JM, Kerssens JJ. Gender and communication style in general practice: differences between women’s health care and regular health care. Med Care 1998;36:100–6. A6: Van Dulmen AM. Communication during gynecological out-patient encounters. J Psychosom Obstet Gynaecol 1999;20:119–26. A7: West C. Routine complications: troubles with talk between doctors and patients. Bloomington, IN: Indiana University Press; 1984.
References [1] Lipkin M, Putnam S, Lazare A, editors. The medical interview: clinical care, education, and research. New York: Springer; 1995. [2] Stewart M, Brown BJ, Weston WW, McWhinney I, McWilliam CL, Freeman TR, editors. Patient-centered medicine: transforming the clinical method. Thousand Oaks, CA: Sage; 1995. [3] Roter DL. The enduring and evolving nature of the patient–physician relationship. Pat Educ Couns 2000;39:5–15. [4] Lurie N, Margolis KL, McGovern PG, Mink PJ, Slater JS. Why do patients of female physicians have higher rates of breast and cervical cancer screening? J Gen Int Med 1997;12:34–43. [5] Steingart RM, Packer M, Hamm P, Coglianese ME, Gersh B, Geltman EM, et al. Sex differences in the management of coronary artery disease. New Engl J Med 1991;325:226–30. [6] Roter DL, Hall JA. How physician gender shapes medical care. Mayo Clin Proc 2001;76:673–6. [7] Roter DL, Hall JA. Patient sex and communication with physicians: results of a community-based study. Women’s Health 1995;1:77–95. [8] Weisman CS, Teitelbaum MA. Women and health care communication. Patient Educ Couns 1989;13:183–99. [9] Roter D, Lipkin Jr. M, Korsgaard A. Sex differences in patients’ and physicians’ communication during primary care medical visits. Med Care 1991;29:1083–93. [10] Hall JA, Irish JT, Roter DL, Miller CM, Miller LH. Gender in medical encounters: an analysis of physician and patient communication in a primary care setting. Health Psychol 1994;13:384–92. [11] Roter DL, Hall JA, Aoki Y. Physician gender effects in medical communication: a meta-analytic review. J Am Med Assoc 2002;288:756–64. [12] Dindia K, Allen M. Sex differences in self-disclosure: a metaanalysis. Psychol Bull 1992;112:106–24. [13] Hinsz VB, Tomhave JA. Smile and (half) the world smiles with you, frown and you frown alone. Personality Soc Psychol Bull 1991;17:586–92. [14] Hall JA. Nonverbal sex differences: communication accuracy and expressive style. Baltimore: Johns Hopkins University Press; 1984. [15] Gouldner AW. The norm of reciprocity: a preliminary statement. Am Soc Rev 1960;25:161–78. [16] Roter DL, Hall JA. Doctors talking to patients/patients talking to doctors: improving communication in medical visits. Westport, CT: Auburn House; 1992. [17] Hall JA, Roter DL, Rand CS. Communication of affect between patient and physician. J Health Soc Behav 1981;22:18–30.
224
J.A. Hall, D.L. Roter / Patient Education and Counseling 48 (2002) 217–224
[18] Cappella JN. Mutual influence in expressive behavior: adult–adult and infant–adult dyadic interaction. Psychol Bull 1981;89:101–32. [19] Knapp ML, Hall JA. Nonverbal communication in human interaction. 5th ed. New York: Wadsworth; 2001. [20] Jorgenson DO. Nonverbal assessment of attitudinal affect with the smile-return technique. J Soc Psychol 1978;106:173–9. [21] Harris MJ, Rosenthal R. Mediation of interpersonal expectancy effects: 31 meta-analysis. Psychol Bull 1985;97:363–86. [22] Snyder M, Tanke ED, Berscheid E. Social perception and interpersonal behavior: on the self-fulfilling nature of social stereotypes. J Pers Soc Psychol 1977;35:656–66. [23] Aries E. Men and women in interaction: reconsidering the differences. New York: Oxford University Press; 1996. [24] Eagly AH, Johnson BT. Gender and leadership style: a metaanalysis. Psychol Bull 1990;108:233–56. [25] Stewart MA. Effective physician–patient communication and health outcomes: a review. Can Med Assoc J 1996;152:1423–33. [26] Hall JA, Roter DL, Katz NR. Meta-analysis of correlates of provider behavior in medical encounters. Med Care 1988;26:657–75. [27] Roter DL. Roter Interaction Analysis System: coding manual. Available upon request to the author and through the website RIAS.ORG. Baltimore: Johns Hopkins University; 2001. [28] Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Mahwah, NJ: Erlbaum; 1988. [29] Rosenthal R. Parametric measures of effect size. In: Cooper H, Hedges LV, editors. The handbook of research synthesis. New York: Sage; 1993, p. 231–244. [30] Rosenthal R. Meta-analytic procedures for social research. Newbury Park, CA: Sage; 1991. [31] Joos SK, Hickam DH, Gordon GH, Baker LH. Effects of physician communication intervention on patient care outcomes. J Gen Int Med 1996;11:147–55.
[32] West C. Routine complications: troubles with talk between doctors and patients. Bloomington, IN: Indiana University Press; 1984. [33] Hall JA, Irish JT, Roter DL, Ehrlich CM, Miller LH. Satisfaction, gender, and communication in medical visits. Med Care 1994;32:1216–31. [34] Roter DL, Hall JA, Kern DE, Barker LR, Cole KA, Roca RP. Improving physicians’ interviewing skills and reducing patients’ emotional distress: a randomized clinical trial. Arch Int Med 1995;155:1877–84. [35] Fallowfield L, Jenkins V, Farewell V, Saul J, Duffy A, Eves R. Efficacy of a cancer research UK communication skills training model for oncologists: a randomized controlled trial. Lancet 2002;349:650–6. [36] Smith RC, Marshall-Dorsey AA, Osborn GG, Shebroe V, Lyles JS, Stoffelmayr BE, et al. Evidence-based guidelines for teaching patient-centered interviewing. Patient Educ Couns 2000;39:27–36. [37] Rosenthal R. Experimenter effects in behavioral research. Enlarged ed. New York: Irvington Publishers; 1976. [38] Gabbard-Alley AS. Health communication and gender: a review and critique. Health Commun. 1995;35–54. [39] Pendleton DA, Bochner S. The communication of medical information in general practice consultations as a function of patients’ social class. Soc Sci Med 1980;14A:669–73. [40] Waitzkin H. Information-giving in medical care. J Health Soc Behav 1985;26:81–101. [41] Waitzkin H. Doctor–patient communication. J Am Med Assoc 1984;252:2441–6. [42] Hall JA, Roter DL. Patient gender and communication with physicians: results of a community-based study. Women’s Health 1995;1:77–95. [43] Stewart M. Patient characteristics which are related to the doctor– patient interaction. Fam Pract 1983;1:30–5.