Received from the Division of General Internal Medicine (MCB, DEF), Department of Medicine, The Johns Hopkins School of Medicine; Department of Health Policy and Management (MCB, DR, SL, DEF), The Johns Hopkins School of Hygiene and Public Health, Baltimore, Md; Department of Medicine (WL), St. Michael's Hospital, University of Toronto, Toronto, Canada; Regenstrief Institute (RF), Indiana University School of Medicine; and Roudebush Veteran's Administration Medical Center (RF), Indianapolis, Ind.
What Do Physicians Tell Patients About Themselves?
A Qualitative Analysis of Physician Self-disclosure
Version of Record online: 20 AUG 2004
Journal of General Internal Medicine
Volume 19, Issue 9, pages 911–916, September 2004
How to Cite
Beach, M. C., Roter, D., Larson, S., Levinson, W., Ford, D. E. and Frankel, R. (2004), What Do Physicians Tell Patients About Themselves?. Journal of General Internal Medicine, 19: 911–916. doi: 10.1111/j.1525-1497.2004.30604.x
- Issue online: 20 AUG 2004
- Version of Record online: 20 AUG 2004
- patient-physician communication;
- physician self-disclosure;
- patient-physician relationship;
- qualitative analysis
OBJECTIVE: Physician self-disclosure (PSD) has been alternatively described as a boundary violation or a means to foster trust and rapport with patients. We analyzed a series of physician self-disclosure statements to inform the current controversy.
DESIGN: Qualitative analysis of all PSD statements identified using the Roter Interaction Analysis System (RIAS) during 1,265 audiotaped office visits.
SETTING AND PARTICIPANTS: One hundred twenty-four physicians and 1,265 of their patients.
MAIN RESULTS: Some form of PSD occurred in 195/1,265 (15.4%) of routine office visits. In some visits, disclosure occurred more than once; thus, there were 242 PSD statements available for analysis. PSD statements fell into the following categories: reassurance (n= 71), counseling (n= 60), rapport building (n= 55), casual (n= 31), intimate (n= 14), and extended narratives (n= 11). Reassurance disclosures indicated the physician had the same experience as the patient (“I've used quite a bit of that medicine myself”). Counseling disclosures seemed intended to guide action (“I just got my flu shot”). Rapport-building disclosures were either humorous anecdotes or statements of empathy (“I know I'd be nervous, too”). Casual disclosures were short statements that had little obvious connection to the patient's condition (“I wish I could sleep sitting up”). Intimate disclosures refer to private revelations (“I cried a lot with my divorce, too”) and extended narratives were extremely long and had no relation to the patient's condition.
CONCLUSIONS: Physician self-disclosure encompasses complex and varied communication behaviors. Self-disclosing statements that are self-preoccupied or intimate are rare. When debating whether physicians ought to reveal their personal experiences to patients, it is important for researchers to be more specific about the types of statements physicians should or should not make.
Physician self-disclosure, defined broadly as any statement made to a patient that describes the physician's personal experience,1 is a controversial communication behavior. Physicians’ personal revelations to patients have been alternatively described as a boundary transgression2 or as a way of fostering trust and rapport in the patient-physician relationship.3–7 In terms of self-disclosure's potential for boundary violation, a special report on boundaries in the patient-physician relationship describes physician self-disclosure as a common starting point down the slippery slope toward a sexual relationship with a patient. The authors report “even if revealing personal issues to a patient does not lead progressively to more extreme boundary violations, self-disclosure is in itself a boundary problem because it is a misuse of the patient to satisfy ones’ own needs for comfort or sympathy.”2
In contrast, others have described physician self-disclosure as an opportunity for fostering intimacy, trust, and reciprocity between doctors and patients.3 One study reports that viewers of a health counseling video in which a physician discloses his/her own positive health behaviors considered the physician to be more credible and more motivating regarding diet and exercise than in a similar video in which the physician does not disclose.6 Another study found that patients who had been disclosed to by their therapist reported liking their therapist more and had less symptom distress.7 Finally, an experimental study focusing on the conditions under which subjects would sue their physicians found a video vignette in which a physician had made a mistake but was self-disclosing (empathic) resulted in low desire to sue, whereas in a companion vignette in which there was a mistake and low self-disclosure there was high desire to sue. The authors concluded from this study that high self-disclosure is protective, whereas low self-disclosure increases risk of medical malpractice.8
Although opinion about self-disclosure is divided, the debate is largely uninformed by any body of evidence that describes the nature of what physicians might tell a patient about themselves. We conducted a qualitative analysis of a large sample of practicing physicians’ self-disclosures and characterized different types of expressions to inform the current controversy.
We conducted our analysis on data collected in 1993 to compare the routine communication styles of physicians with and without a history of malpractice claims.9 The original study was approved by the Institutional Review Board (IRB) of Legacy Good Samaritan Hospital, Portland, Oregon and by the IRB at Johns Hopkins University School of Hygiene and Public Health.
Physicians were selected for participation in the original study from the databases of two large insurance companies according to their history of malpractice claims. Physicians were eligible for inclusion if they were in active practice in Denver, Colorado, or Portland or Salem, Oregon, and had graduated from medical school at least 13 years prior to the study. Eligible physicians included both primary care physicians (general internists and family physicians) and surgeons (general and orthopedic) because it was hypothesized a priori that the communication patterns of generalists and surgeons might differ with respect to the association between malpractice claims and specific communication behaviors. The methods of recruitment are well described in the original study.9
Overall, 81% of eligible physicians agreed to participate. There was a higher acceptance rate for surgeons (89%) than for primary care physicians (74%). All participating physicians gave informed consent. The final sample of 125 physicians included 59 primary care physicians and 66 surgeons. Physicians in the study were primarily white (94%), male (93%), had been in practice a mean of 18.3 years, and worked a mean of 50.5 hours/week providing patient care.
Patients were recruited as a convenience sample by research assistants stationed in the participating physicians’ waiting rooms. The research assistant attempted to recruit at least 10 patients for each physician (range in final sample 8 to 12 patients/physician). Patients were eligible for inclusion if they were older than 18 years, spoke English, and were not in any acute distress. Written consent was obtained and 80% of eligible patients agreed to participate. The final sample consisted of 1,265 patients who were mostly white (86%) and female (57%). The mean patient age was 52.5 years, 26% had an annual income of less than $20,000, and 27% had completed college.
Identification of Physician Self-disclosure
Each of the audiotaped patient visits (N= 1,265) was coded for content by one of three trained coders using the Roter Interaction Analysis System (RIAS). In this extensively studied and well-validated system,10–13 each statement or complete thought made by either physician or patient is coded into one of 38 mutually exclusive and exhaustive categories. Although the purpose of the original study was not to study physician self-disclosure explicitly, the RIAS routinely codes physician self-disclosure.
Self-disclosure is the act of revealing a personal and nonpublic experience. In this study, physician self-disclosures with medical relevance for the patient are distinguished from social disclosure that are generally characterized as friendly conversation or “chitchat.” The operational definition from the RIAS coding manual is “Statements which describe the physician's personal experience in areas which have medical and/or emotional relevance for the patient.”1 Examples of physician self-disclosure used in the RIAS coding manual are “My wife was diagnosed with breast cancer two years ago and I know how rough it is for everyone,”“I used to smoke until I realized it was killing me,” and “That's what it's all about for me—helping patients get over pain—it's rewarding.”1
Coders were trained in the RIAS by SL using a coding manual with detailed definitions and annotated examples and training tapes. Intercoder reliability was calculated on 121 double-coded tapes. The overall mean reliability coefficient for all physician communication behaviors was 0.80 and the reliability coefficient for physician self-disclosure was 0.94.
Analysis of Physician Self-disclosure
After physician self-disclosure was identified using the RIAS, we transcribed the disclosure within its interactional context. Because the RIAS codes each statement separately, we grouped disclosure statements that occurred together (and were part of the same thought) to form disclosure units. Our task was then to categorize the disclosure units into logical categories, a task well suited to qualitative descriptive methods.14
Two investigators (MCB, DR) read through all the disclosure units and met regularly to compare independent assessments of the transcripts and to develop a consensus of the coding categories. We developed themes through a combined inductive and deductive consensus-building process. Although we knew how physician self-disclosure had been described in the extant literature (and looked for evidence of boundary violations, for example), we remained open to the possibility of finding themes that had not been previously described. We then separately coded all statements and resolved disagreement with discussion and consensus.
Physicians made at least one self-disclosing statement in 15.4% (195/1,265) of patient visits. Although the number of self-disclosing statements made by the physicians who disclosed ranged between 1 and 69 statements per visit, the median number of self-disclosing statements made (among physicians who disclosed at all) was 2. The majority (96%) of physicians who disclosed during a patient visit made 3 or fewer self-disclosing statements. The 195 self-disclosure visits yielded 242 self-disclosure units. Only 4 of 242 physician self-disclosures were in response to a patient question; the remainder were spontaneous. Themes that categorized these disclosures are described in Table 1. The themes we identified were similar for surgeons and primary care physicians, and so all data were analyzed and reported together.
|Reassurance (short)||A short statement indicating that the physician has personally shared the same experience as the patient||I sometimes do that.|
|Same thing with me.|
|I take it, too.|
|Sure I do. Everyone does.|
|I know my wife had it about a year and a half ago.|
|Reassurance (long)||A statement providing information about the physician's personal experience||It probably was ’cause I've had it happen to me. That's part of scar reaction, and that can take up to 3 to 4 months to disappear. I found that in my knee when I had the same thing happen.|
|Counseling||A statement about the physician's personal experience that could guide patient action||I’ll write the prescription for twice a day, and then you can play with it. When I took this stuff, I was taking one three times a day ’cause I didn't like a lot of Sudafed before bedtime. It keeps me awake. Makes me restless. Three times a day didn't bother me.|
|I've had mine. I think it's worth doing. (flu shot)|
|I just got one. Yeah, and it works pretty well. I think it's a pretty good idea for a knee brace if you’re going to ski.|
|Rapport (humor)||A humorous personal story||Like my dad told me, when I was getting on him about smoking when he had the bypass, he said, “You know, I see a lot of old jazz musicians, but I don't see too many old doctors.”[laughs]|
|Rapport (empathy/legitimation)||An expression of empathy or legitimation of patient experience||They usually give you a little something to help relax you a little bit. ’Cause I know I'd be nervous.|
|Casual||A short statement that has little obvious connection to the patient's condition or state||(after compliment on sutures) You know, I was kind of sickly as a child. And my parents, my twin brother and I, they had us doing embroidery.|
|Intimacy (emotional/physical)||An intimate emotional or physical revelation||I cried a lot with my divorce, too. You know, when our bodies say we need to cry, maybe there's something that needs cleansing.|
|Take a supplement. I, personally, use Metamucil. I started on it when I went to a seminar about soluble fiber and its effects on cholesterol. But its other effect in terms of stooling is just real pleasant, too.|
|Intimate (relationship)||Any indication of a desire for a personal relationship with a patient||I like to think that my patients are my friends. I like to think that if I happen to run into you here in town, I know who you are.|
|Extended narrative||A lengthy description of physician's personal experience that has seemingly little relevance for patient||Crunches. And what I do...I dedicated myself...if I'm going to watch a basketball game or football game, for one quarter I'm going to do a (unintelligible) workout. And sometimes when I've done a workout, it's on an isometric. Going on them. When I drive, and I get to a stoplight, each time I do it, I just reflexively put my hands under the wheel and push my knees up and do an isometric. Not very hard, a little bit. And that strengthens the gut. Push your legs up, push down with your hands. Tightens the gut. One of the exercises I’ll do during the game is this. People watching with me, I just might do this. And you can't even see I'm doing it. But I'm picking them up just enough to tighten the gut. And then I end up doing 1,000 of them. And it makes a difference. I’ll get down on the floor and I’ll do crunches. I’ll do maybe groups of 30, and then maybe try to do 200 by the time you do them. Every little thing you can do to exercise your belly on a routine basis I think makes it better. And, of course, weight loss may or may not make a difference, but it sure makes you feel better. It's worthwhile. (Pt: Wow)|
|I just went through a low-fat diet for 6 months ’cause my cholesterol was high—and I lost 30 lbs. Then I got back on a low-fat…my parents had good numbers, and I thought, “nothing wrong with me.” But then I got tested—270. And I figured three donuts in the morning on the way to the hospital—they weren't a big problem. So I stopped doing that. Took all the fat out of my diet. Quit eating meat. And I went down to 186. So I haven't had it tested again. I’ll be disappointed, probably. But I eat very little fat. I've never watched what I eat. I eat a ton. Lots of salad, lots of fruit. I start the day with oat bran, which is supposed to be cholesterol-free. Fills me. I don't get that empty feeling when you grab donuts. Don't eat meat anymore.|
The most common type of self-disclosure (n= 71) was a statement that indicated the physician has personally had the same experience as the patient, which we refer to as the “Reassurance” category. In this category, we distinguished those disclosures that were very short statements from those that provided more information about the physician's experience. Examples of short “Reassurance” statements are “I sometimes do that,”“My mother had one,” and “I'm having a little back pain myself.” Examples of longer reassuring statements are “Unfortunately, blepharitis is not a curable problem; it's there. I actually have a little blepharitis in this one eye and it flares up from time to time” and “I got bit by a mosquito myself 5 days ago. I still have a little bit of a welt. I'm sensitive to it. So it might take a week for the swelling to go down,” and (in the context of the patient complaining about stress and muscle ache) “Well, in myself and in most other people I know, the tension that we go through, the stresses that we have, do affect the muscles.”
The second most common category was disclosure in the context of physician counseling (n= 60), in which the physicians’ experience could be used to guide patient action. This sort of counseling occurred around lifestyle issues as well as biomedical issues. Examples of the use of self-disclosure in a counseling context are: “How about some aquasox to give you some traction? My kids live in them,” and “I take a combination of both. The crystals I found, for me, are inconvenient. It's hard to carry a bottle of powder around in my pocket”and“As I've gotten a little bit older, I've found I can't do the push-ups like I used to, but I do other things to keep myself in shape.”
Another rather common category of disclosure was rapport building (n= 55). In this category, there were disclosures that were humorous and those that express empathy or legitimate a patient's emotions. Examples of humorous disclosures are “I know. I had a dentist tell me when I had to wear one of those things. I said no, it's alright, I’ll grind them down and get false ones (laugh) before I wear that thing. Did you have TMJ problems before you fell off the horse?” and (in context of discussing difficulty following diet) “Serving size is one piece of bread—serving size for me is probably 4 pieces of bread!” Examples of disclosures that reflect empathy and legitimation are “It's hard, you have to push yourself to do that sort of stuff. Getting up at 4 am doesn't sound good to me,”“I know, it made me nervous too,” and “Well, I'm sorry you have to go back for a meeting: I have a lot of days like that.”
Short statements with seemingly little direct relation to the patient's problem or condition (n= 31) were categorized as “Casual.” Examples of casual disclosures are (after patient reports on exercise regimen) “Okay. I wish I got to do that much. I'm supposed to” and “I tried to smoke when I was young and got so sick on cigars, it cured me real quick” and “I wish I could sleep sitting—I have to get recumbent.”
Intimate disclosures (n= 14) were those that either revealed a personal physical or emotional aspect of the physician's life, or one that suggested the physician would like a personal relationship with the patient. Examples of emotionally intimate disclosures are “I don't know what to do about stress. I'm not that good at it in my own life” and “If it's any consolation to you, we've had the same thing happen with our 17-year-old. And it began when he was 16. He’ll be 18 and, in the last 3 weeks, we have seen, for the first time in a year and a half, a real person in there.” An example of a physically intimate disclosure is “When you wipe, it irritates the hemorrhoids and people have a tendency to bleed. I do the same thing.” There was only one example of a statement that indicated a physician's desire to have a more personal relationship with a patient, “I like to think that my patients are my friends. I like to think that if I happen to run into you here in town, I know who you are.”
Extended narratives, defined as a lengthy description of the physician's personal experience without relevance for the patient, was evident in 11 instances of disclosure. These statements are distinguished from other types of disclosure simply by the length of the statement. The following example (had it been shorter) could have been considered a “casual” disclosure: “You probably would have. Actually, that's what I did. I went into the service the day after high school graduation, and I was a hospital corpsman. And then when I got out...in [my state] you can challenge the state nursing boards with a certain amount of training and so on. So I challenged the boards and worked as what then was a nurse all the way through college and actually through the first part of medical school. And made a very good salary. And I actually had wonderful people working with me who sorta took care of me, made sure I was able to study. And it worked out great. So I always had a job, had a tremendous amount of experience.”
Our study found that there are a variety of categorizable self-disclosure statements made by physicians, the most common types of which are reassurance, counseling, and rapport building. While our study was not designed to resolve the issue of whether or not self-disclosure is a boundary violation, we found evidence that intimate or extended disclosures (which may be thought of as self-preoccupied or a misuse of the patient's time) are rare.
It is interesting to note that early work by Wasserman et al.15 in a pediatric context demonstrated that among supportive physician behaviors including reassurance support and empathy, only empathy had a positive relationship to medical outcomes. Reassurance, which was used most frequently by physicians, had no effect, while empathy, which was used least, resulted in a reduction of maternal concerns and satisfaction with the visit.15 Our study identified 2 types of reassurance: short statements that simply indicate the physician has experienced the same phenomenon as the patient (which may or may not be useful) and longer statements that elaborate on the physician's experience and have the potential to be more helpful. Our results confirm that reassurance (both types together) is still a mainstay of self-disclosure, although measurable impact from these types of statements has yet to be demonstrated.
In terms of categorizing disclosure statements, reasonable people could disagree about whether a particular statement fell into one or another category, particularly in the case of categories that might be considered boundary violations (i.e., intimacy and self-preoccupation). For example, in terms of intimacy, what degree of personal revelation is too intimate? In several cases, the line between empathy and emotional intimacy was unclear, and led us to wonder about the conditions under which it would be acceptable, or even desirable, for a physician to share the fact that he/she cried during a divorce. How might that information impact a patient? Different patients would likely respond differently; some might find it comforting, others might see the physician as weak16 and not want to bother him or her with their own problems. Similarly, we found only one example of a physician's desire for a less formal relationship with a patient, and it was not outside the bounds of what might be considered reasonable practice.
There were similar difficulties in distinguishing between physical intimacy and reassurances or counseling. Social convention seems to dictate that it is more acceptable to talk about your heart condition than your bowels or genitourinary tract, and so the statements that we categorized as physically intimate (there were only 2) had to do with bowel function. As in the case of emotional intimacy, the rules of social etiquette, which may vary between cultures and between individuals, has dictated what are acceptable topics for disclosure. In the case of potentially stigmatizing material, Candib has suggested that physicians only disclose to patients whom they know well.3
In terms of extended narratives, it is difficult to decide how long a statement is too long, or when a statement crosses the line from chitchat to an undesirable level of self-absorption. Does disclosure satisfy the physician's own needs? Would it be a boundary violation if it did? Some might argue that any statement made by a physician about herself or himself is an indication of self-preoccupation, perhaps satisfying that physician's own needs. However, it seems that what is relevant is not whether the physician gets any subliminal gratification from the disclosure, but whether the disclosure actually harms the patient.
The results of this analysis should be interpreted with several limitations in mind. First, the RIAS codes only those disclosures with emotional/medical relevance to the patient, so that most social disclosures expressed during the nonmedical chatting that often opens or closes a visit would be missed in our analysis. Second, although we coded based on our perception of the intent or effect of the disclosure, understanding the interactional dynamics of disclosure would require a different type of analysis, which is beyond the scope of this paper. What seems humorous or empathic to one person might seem insensitive to another, and we have not yet established reliability of the coding system. In our study, differences in coding were resolved with discussion and consensus. Third, the data are now 10 years old, and it may be possible that self-disclosures have become more or less common over time. Finally, this sample does not include physicians early in their careers, nor does it include psychiatrists who, compared to other physicians, may have a unique context for self-disclosure.
Despite these limitations, we report on a large sample of physicians, which allows us to examine a relatively infrequent event. Our study is a first attempt to describe actual disclosures and to distinguish different types of disclosures from audiotaped encounters. We have found that physician self-disclosure is a complex communication behavior that ought not be considered as a single entity. When debating whether or not physicians ought to tell patients personal information about themselves, it seems reasonable and responsible to be more specific about the types of disclosures one is referring to in helping to guide physicians to optimize their relationships with patients. Fear that physician self-disclosure to patients will lead down a slippery slope to more extreme boundary violations seems unsubstantiated in the visits we studied. In the routine office practice of primary care physicians and surgeons, the overwhelming majority of disclosures do not appear to represent a danger to patients.
This work was supported by grant 99-504 from the Bayer Institute for Health Care Communication and by grant R01 07289 from the Agency for Healthcare Policy Research.
- 1The Roter Interaction Analysis System (RIAS) Coding Manual. Baltimore, Md: School of Hygiene and Public Health, Johns Hopkins University; 1991..
- 3What doctors tell about themselves to patients: implications for intimacy and reciprocity in the relationship. Fam Med. 1987;19: 23–30..
- 12Pediatrician interview style and mothers’ disclosure of psychosocial issues. Pediatrics. 1994;13: 289–295., , .