Clinician Reactions to Patients With Eating Disorders: A Review of the Literature

Extensive research indicates that the general public holds more negative attitudes toward some mental illnesses than others (1), and additional research suggests psychiatrists sometimes hold similar attitudes (2). Clinicians, including psychiatrists, may have negative reactions toward particular patient groups, and these reactions may have an impact on the delivery of adequate psychiatric and medical services (3,4).

It is frequently reported that clinicians across professional disciplines experience strong negative reactions in regard to patients with eating disorders (5–7). Research suggests clinicians often decline to treat individuals with eating disorders (8,9), and negative clinician attitudes about patients with these disorders may affect treatment alliances and outcome (10,11). Some researchers suggest that clinician negative reactions are related to the nature of symptoms of eating disorders (6), but others speculate that these reactions are due to pervasive stigma (12).

To date, empirical studies of clinician reactions to patients with eating disorders have not been aggregated; therefore, it is unclear whether clinicians' attitudes toward patients with eating disorders are demonstrably more negative than their attitudes toward other patient groups, which emotions typically characterize these negative reactions, and whether negative reactions are meaningfully associated with particular patient or clinician factors.

This review aimed to summarize the available data concerning the nature and correlates of all types of clinician reactions to patients with eating disorders. The term “reactions” was defined in the broadest sense possible to include emotions, attitudes, and beliefs that clinicians expressed about patients with eating disorders or that they experienced in the course of their work with eating disorders. We reviewed studies sampling all clinician groups, including but not limited to psychiatrists and psychotherapists, because we wanted to obtain any data that might be relevant to a topic that has been the subject of only limited empirical research and because psychiatric services for eating disorders are frequently provided by a multidisciplinary team that includes physicians, nutritionists, and other health care professionals.

Methods

We conducted a comprehensive computerized search for all studies published through 2010 of clinician reactions to patients with eating disorders in standard electronic databases, such as Medline and PsycINFO, as well as in online search programs for journals about eating disorders, such as the International Journal of Eating Disorders. A search was conducted of every combination of the following search terms: clinician, therapist, staff, physician, doctor, professional, provider, reaction, feeling, attitude, emotion, countertransference, alliance, and relationship. We examined the lists of references from articles identified by the search for other pertinent research.

Results

Study findings

The search located 20 studies published from 1984 through 2010 that sampled clinicians about their reactions to patients with eating disorders. We separated the studies into four groups by differences in methods and results. [A table summarizing the 20 studies listed chronologically is available in an online appendix to this report at ps.psychiatryonline.org.]

Studies of inexperienced clinicians and trainees.

Four studies assessed the attitudes and feelings of clinicians with limited experience treating patients with eating disorders. In all four studies there were indications that trainees had particularly negative reactions in regard to patients with these disorders.

In a study published in 1984, first-year residents in psychiatry, medicine, and pediatrics completed a questionnaire aggregating their level of anger, stress, helplessness, sadness, and anxiety into a “dysphoric affect” metric. All residents reported more dysphoric affect toward patients with anorexia nervosa than patients with obesity or diabetes (13). Another study, conducted in 1992, found that medical and nursing students in a general hospital in Australia considered patients with eating disorders to be significantly more responsible for their illness than individuals with schizophrenia and that they endorsed that belief significantly more than doctors (14). Furthermore, the medical students showed no significant change in their “liking” of patients with anorexia nervosa after an eight-week psychiatry rotation providing training in treatment of eating disorders, but their liking of those with schizophrenia after a comparable training period increased.

A study conducted in 1993 assessed reactions to patients with eating disorders by a sample of therapists from mixed professional disciplines, including 42% from psychiatry and 40% from psychology, with limited experience treating eating disorders (8). Results indicated that 31% of the sample reported preferring not to treat patients with eating disorders (8). Among a list of six feelings about treating patients with eating disorders (anger, anxiety, empathy, frustration, helplessness, and satisfaction), participants most commonly endorsed frustration (87%) and anger (63%) (8).

One study published in 1998 compared primary care nurses in a private hospital who had recently begun working on an eating disorders unit with those beginning work on medical units (15). The nurses were asked to complete a measure assessing negative attitudes toward the patient population with whom they worked, for example, by answering whether they considered the patient to have negative qualities such as being stubborn, demanding, selfish, or lacking in appreciation. At the beginning of the study, the nurses from the eating disorders unit reported more negative impressions of their patients than those who were working on medical units. Furthermore, longitudinal data suggested that unlike nurses working on medical units, nurses on the eating disorders unit reported increasingly negative impressions of their patients over the course of one year.

Studies of specialists and highly experienced psychotherapists.

In contrast to studies of inexperienced clinicians, several studies of highly experienced psychotherapists and eating disorders specialists suggested that these clinicians did not experience high levels of negative feelings about patients with eating disorders; however, when they did experience negative emotions, they were similar to the types of negative emotions reported by trainees.

In a study published in 2005, observers reported uniformly high alliance between patients and experienced therapists who had received extensive training in the efficacious psychotherapy protocols of a clinical trial of treatments for bulimia nervosa (10). A 2009 study examined responses by highly experienced psychiatrists and psychologists treating adolescents in individual therapy with self-selected therapeutic approaches to a 79-item, previously validated measure of positive and negative emotions toward patients. The therapists reported low levels of negative reactions overall (16).

However, factor analysis identified four types of negative reactions—frustration, lack of competence, worry, and boredom. These findings were consistent with another study of feelings about patients with anorexia nervosa, bulimia nervosa, and dysthymia among eating disorders specialists (psychologists and social workers) assessed by an extensive, 34-item measure (5). In this study, the subsample of highly experienced therapists reported feeling less frustration than less experienced therapists, but 87% reported feeling frustration at some time, and feelings of helplessness and hopelessness were also common. Furthermore, clinicians reported feeling less connected and less successful with their patients with eating disorders, particularly those with anorexia nervosa (5).

One study of a group of experienced clinicians at a recent eating disorders conference identified difficulties in working with these patients; consistent themes reported by the therapists included lack of readiness to change, resistance, nature of illness, and managing personal negative affect (17). A 2010 study reported results of interviews with clinicians specializing in the treatment of eating disorders in the United Kingdom about the difficulty of the work. These experienced health care professionals did not find eating disorders patients difficult to work with but rather found it difficult to successfully treat their problems because of limited systemic resources (18).

General surveys of medical professionals.

Five studies that surveyed the attitudes of professional practitioners (primarily but not exclusively medical practitioners and specialists in areas other than eating disorders) indicated that the providers had a lack of confidence or felt a lack of competence in treating eating disorders. A national sample of primary health care physicians—including internists, family practitioners, and pediatricians—published in 1987, found that all types of primary care physicians perceived deficiencies in their own competency to treat eating disorders, substance abuse, and delinquency (19).

A second national survey published in 1990 assessed feelings of competency in a large sample of clinicians, including physicians, nurses, psychologists, and social workers (20). This study found that 54% of physicians identified eating disorders as an area of treatment in which they perceived themselves as deficient, making it the second most commonly reported area of perceived lack of competence. Up to 62% of psychologists and 67% of social workers reported limitations in their treatment of the psychological sequelae of eating disorders.

A study of gynecologists and obstetricians—of interest because irregular menstrual function is a common symptom of eating disorders—in Australia and the United Kingdom, published in 2002, found that 80% reported lacking confidence in diagnosing eating disorders (21). Surveys of family practitioners, one conducted in Ontario and published in 2002 (22) and one conducted in Oregon and published in 2010 (23), indicated that a majority of respondents felt uncomfortable with the treatment and management of patients with eating disorders. In the Oregon study, 78% of the family physicians reported currently having patients with eating disorders and feeling unsure about how to treat them (23). A fear of lack of competence, difficulty treating eating disorders, and a desire for increased training in these disorders were among the themes that emerged from in-depth interviews (23).

Studies of clinician attitudes toward self.

A distinct subset of three studies assessed clinicians' feelings about their own shape, weight, and eating in response to treating individuals with eating disorders (15,17,24). In a survey published in 1999, clinicians who reported being “moderately to greatly affected by their work” with patients with eating disorders (28% of the sample) retrospectively reported that they had adopted healthier eating and had experienced positive changes in body image (24). In a study published in 1998, investigators found that nurses on an inpatient unit that treated eating disorders patients reported having less distorted attitudes toward eating than nurses in other medical units. However, this finding may have been due to the fact that the average weight of the nurses in the unit that treated eating disorders was in the normal range and was below that of nurses in other medical units (15).

Another retrospective study, published in 2009, yielded additional conflicting results. Fifty percent of a mixed sample of clinicians reported increased negative evaluation of others' body weight and shape and heightened self-criticism after sessions with patients with eating disorders, but 54% reported positive changes in their own eating habits since beginning to treat eating disorders (17).

Predictors of clinician reactions

Given the available data, it is particularly important to identify factors that may have accounted for clinicians' negative reactions.

Patient prognosis and treatment response.

Three studies investigated clinician reactions to prognosis, treatment response, or motivation among patients with eating disorders. A survey of clinicians published in 1995 found that those who perceived patients with eating disorders to have poor prognoses were more likely to refer them to other providers (8). Recent data from outpatient psychotherapists treating eating disorders indicated that the clinician's perception of the patient's improvement in treatment was the most consistent predictor of lower levels of all types of negative emotional reactions (16). Among treatment providers surveyed at a recent national conference on eating disorders, 55% reported eating disorders patients' resistance to change as the hardest aspect of working with them (17). Two studies—including one in 2009—suggested that physicians misperceive eating disorders as more treatment resistant than is indicated by research (21,25,26).

Stigmatizing attributions by clinicians.

The belief that individuals with eating disorders are responsible for their symptoms and therefore are to blame for their illness has been consistently reported in survey studies of the general public (26–29). Because the symptoms of eating disorders can be ego-syntonic, individuals with severe eating disorders are often ambivalent about recovery. Their symptoms may have the appearance of personal choice, although they may result primarily from factors beyond the patients' control (7). Four studies (12,14,26,27) suggested that clinician beliefs about the patients' personal responsibility for their eating disorders may be associated with clinicians' feelings or behaviors.

Data indicating that clinicians commonly hold blaming or stigmatizing attitudes toward patients with eating disorders were collected across a wide range of years, locations, and professional disciplines. When asked in 1992 to consider the origins of psychiatric illness among patients with anorexia nervosa, schizophrenia, and recurrent overdoses, medical professionals considered patients with the eating disorder significantly more responsible for their illness than patients with schizophrenia and about as responsible as patients with recurrent overdoses (14). Furthermore, the same study found that these attitudes did not change over the course of training (14).

In a 1992 report, 31% of gynecologists and obstetricians in the United Kingdom and Australia endorsed the pejorative or blaming attitude that eating disorders were “abnormal behavior in the context of a weak, manipulative, or inadequate personality” (14). In a rare experimental study (12), one-half of a sample of nursing students was presented with a sociocultural explanation for eating disorders and the other half was presented with a biological explanation. The nursing students who were presented with the sociocultural explanation were more likely to endorse the attitude that individuals with eating disorders were to blame for their illness and that they were vain (12).

In a continuation of this line of research, investigators found that when college students were randomly assigned to view one of three videos describing anorexia nervosa as the result of biology, of culture, or of an interaction of the two, those who viewed the video describing the biological etiological factors showed less stigmatizing attitudes (26). Notably, however, one study found that physicians who believed patients with anorexia nervosa were responsible for their illness were also more likely to offer a follow-up appointment (27).

Clinician experience level.

Across diagnostic categories, less experienced clinicians typically showed significantly more negative reactions concerning their patients (30). Because the treatment of eating disorders requires specialized training in empirically supported methods, safety management, and treatment resistance, the roles of training and experience may be particularly important. One study found that inexperienced psychotherapists reported feeling significantly more frustration (encompassing frustrated, fearful, and angry feelings) toward patients with eating disorders (5). Another study found that compared with medical professionals in the same hospital, medical trainees had more negative reactions to patients with eating disorders (14). Although one study found no relationship between experience level and clinician response, members of the study sample were uniformly highly experienced and reported low levels of negative reactions overall (16).

Several studies indicated that there were potential barriers in training professionals to work with patients with eating disorders. For example, results of general surveys in previous decades found that clinicians who reported feeling a lack of competence to treat eating disorders also expressed relatively low levels of interest in receiving additional training or supervision (19,20). However, a more recent report suggested that although family physicians in Ontario continued to feel deficits in competency to treat eating disorders, they reported interest in receiving additional training (22). Studies investigating the simple relationship between knowledge about eating disorders and reactions to patients with these disorders generally have not found the relationship to be significant (27,31,32), although one study found a significant negative relationship between level of knowledge about eating disorders and the degree to which the clinician held patients responsible for their illnesses (14). However, several studies reported that short-term training experiences did not improve clinician attitudes toward patients with eating disorders (14,31).

Clinician gender.

A majority of available research suggested that compared with female clinicians, male clinicians have more negative reactions to patients with eating disorders. The survey of gynecologists and obstetricians in Australia and the United Kingdom found that male clinicians were more likely to hold pejorative attitudes toward individuals with eating disorders and to misperceive bulimia nervosa as untreatable (21). An earlier study found that male clinicians more frequently than female clinicians declined to treat eating disorders (8).

Survey data from Canada suggested that female family practitioners were more likely than their male counterparts to routinely screen for eating disorders (22). A study of clinicians treating adolescents with eating disorders found that male clinicians were significantly more likely than female clinicians to endorse having aggressive feelings toward patients, although they were also more likely to endorse having warm feelings (16). Eating disorders are more common among women than men, and greater familiarity with their symptoms by women therapists—through personal experience or close relationships—may begin to explain some of the observed gender differences (33).

Clinician discipline.

Three studies have examined differences in attitudes toward patients with eating disorders between groups defined by professional discipline. In the study of medical professionals who were newly working with patients with eating disorders in a hospital setting, nurses reported more blaming attitudes toward patients than did physicians and psychiatric residents on the same unit (14). In one study, psychiatrists endorsed higher levels of anger and frustration toward adolescent patients with eating disorders than did psychologists (16). The early study, in 1984, of medical residents' reactions to patients with anorexia nervosa found that psychiatric and pediatric residents reported more negative feelings toward the patients than did residents in internal medicine (13). Research explaining these observed differences is needed.

Patient personality pathology.

Another factor in the intensity of clinician responses to patients with eating disorders may be the substantial comorbidity with personality disorders, which are also known to be associated with negative clinician reactions (16). Multiple studies demonstrated that axis II, cluster B personality disorders are especially common among patients with bulimia nervosa, and axis II, cluster C personality disorders are common among patients with anorexia nervosa (34). One study found that the patients' personality pathology was a strong predictor of clinician reactions to eating disorders (16). After adjustment for differences in eating disorders diagnosis and treatment response, clinicians treating adolescents with eating disorders reported significantly higher levels of anger, frustration, and feelings of lack of competence—and lower levels of warmth—in response to patients with personality constriction (characterized by difficulty experiencing and expressing emotion) and disregulated personality (characterized by interpersonal difficulties and self-destructive patterns).

Discussion

A review of empirical investigations indicated that negative reactions experienced by clinicians toward patients with eating disorders fell into predictable categories, including feeling frustrated, helpless, incompetent, and worried. Clinicians in the early stages of learning to treat eating disorders held particularly negative attitudes toward this patient group, and surveys consistently indicated that medical professionals did not feel competent to treat their patients with eating disorders.

However, the studies of psychiatric trainees were largely conducted decades ago, when accurate information about eating disorders may have been scarcer and when efficacious treatments had not yet been established. New studies are required to ascertain whether trainees continue to have the same negative reactions in a different psychiatric and cultural context. A particular subset of studies also suggested that clinicians had both positive and negative reactions related to their own weight, shape, and eating habits.

Clinicians' reactions to patients with eating disorders appeared to vary according to the perception of the patient's response to treatment, the degree to which clinicians endorsed stigmatizing beliefs about the patient's responsibility for his or her illness, the amount of experience they had had generally and specifically with eating disorders, their gender and discipline, and the patient's personality pathology.

The significance of clinicians' reactions to eating disorders is multifaceted. The relationship between clinicians' feelings toward patients and the therapeutic alliance is well documented, as is the association between alliance and outcome (10,11). Thus one implication of this review is that the negative reactions by clinicians to patients with eating disorders may adversely affect treatment outcome. A second area of importance is related to the increasing incidence of eating disorders, particularly among younger populations. The number of available and trained providers will need to increase to treat higher numbers of patients with eating disorders. Understanding the reluctance of providers to treat eating disorders and developing educational, training, and supervision opportunities to counter these reactions are critical tasks.

Furthermore, given the lack of specialized, expert services delivered by psychiatrists in many areas, care for these psychiatric patients may fall disproportionately to medical professionals who may feel incompetent to treat patients with eating disorders. Finally, negative reactions to patients with these disorders in the form of stigma may contribute to the reluctance of patients to seek or continue treatment. Decreasing providers' stigmatizing beliefs and behaviors would likely result in better access to care and more successful referrals and outcomes for patients with eating disorders.

Important areas of investigation remain unstudied. No studies were found examining the possibility that the risk of medical complications of eating disorders might affect the feelings of the clinicians who work with them. At 10% to 15%, the mortality rate for anorexia nervosa is among the highest of any psychiatric illness (35,36). Although life-threatening complications are less frequent in bulimia nervosa and eating disorders not otherwise specified, all behavioral symptoms of eating disorders are associated with health and medical risks (37,38). Several authors have speculated that clinicians' reactions may be mediated by their perception of risk (5,16), and research investigating that relationship is needed.

Clinicians also reported that supervision and consultation are subjectively helpful when they are working with patients with eating disorders (5,17); however, no empirical studies have yet examined the influence of quality, quantity, or type of supervisory activities other than short-term or didactic education about clinician reactions. Some research has suggested that the nature of didactic education about prognosis and etiology may affect clinicians' attitudes and behavior toward patients with eating disorders (12,27), but the type and frequency of training and supervisory activities that would be of optimal use to help clinicians deal with negative reactions to patients with these disorders has not been studied.

Other potential areas for future research include treatment parameters, for example, type of treatment—such as group, individual or family—as well as duration, setting, payment options, and elements of clinicians' personal histories with eating disorders or weight issues. Furthermore, the age of the patient was not examined in the studies reviewed in this article, and to our knowledge no data on this subject exist. Nevertheless, it is likely that younger patients elicit different reactions than older patients.

Conclusions

Overall, this review suggests that the topic of clinician reactions to patients with eating disorders is important, given indications that clinicians have historically held negative attitudes toward this patient group and the relative dearth of treatment providers for individuals with eating disorders. A lack of sufficient services for eating disorders may be attributable to the perceived or actual lack of adequate knowledge among treatment providers as well as to stigmatizing attitudes toward these disorders and to the difficulties involved in treating them.

A large body of research has documented the influence of provider attitudes and difficulties providing treatment on service delivery for other complicated medical and psychiatric disorders (39–43). In a similar vein, clinician attitudes, knowledge, and stigma likely adversely affect both the quality and the availability of services for patients with eating disorders. Research in this area is highly feasible yet extremely limited. Moreover, the existing studies are scattered across a number of related topics, have utilized disparate methods with significant limitations, and were published across three decades during which research into the treatment of eating disorders has changed dramatically. As a result, it is difficult to draw solid conclusions about the current state of clinician attitudes toward eating disorders.

Sensitive and multidimensional measurements are needed to better understand clinicians' responses and to appreciate the relative influence of different variables on the course and outcome of treatment. Future studies might examine the relationship between clinician reactions, service delivery, and patient outcomes; replicate observed predictors, mediators, and moderators of clinician reactions with standardized measures; conduct multivariate analyses that control for associated independent variables; and pursue research, including experimental and intervention studies of training and supervision, to evaluate practices that improve clinician reactions to patients with eating disorders, particularly concerning their perceptions of treatment resistance, stigma, and personality pathology.