Physicians' attitudes about obesity and their associations with competency and specialty: A cross-sectional study

Despite the high prevalence of obesity[1] and its associated harmful health effects [2] physicians frequently fail to counsel patients about nutrition and weight management [3‐7] and frequently report and demonstrate a lack of training and competence in obesity management [8‐13]. Clinicians who learn good obesity screening and counseling practices in residency are more likely to report that they always discuss diet or exercise with obese patients [12]. Therefore training of physicians has the potential to impact patient outcomes.

Attitudes, or "individual's positive or negative evaluation of performing a behavior [14]" may also contribute to insufficient or ineffective counseling practices. Major theories predicting practice behavior postulate that attitudes are important influences on the decision to engage in a defined behavior [14, 15]. Attitudes can serve to motivate a provider to take action and encourage a physician to feel capable of tackling a particular behavior (e.g., self-efficacy, perceived behavior control [16]). Indeed, higher self efficacy in delivering preventive care services was predictive of preventive counseling in pediatricians [17]. Attitudes also may instill in a physician the belief that the behavior in question is worth doing because it will produce the intended outcomes (e.g., perceived benefits, outcome expectancies [14]).

Thus, physicians' attitudes about obesity may affect their practice. Studies have shown that physicians' attitudes toward obese patients can be negative, e.g., some primary care physicians rate obese patients negatively with respect to their likelihood of following advice, believe that obese patients are less likely to benefit from counseling, and report less desire to help obese patients [18], and therefore may decrease physicians' motivation to counsel such patients. Even health professionals specializing in obesity have implicit anti-fat bias and are more likely to automatically associate 'fat people' with negative stereotypes than 'thin people' [19, 20]. Physicians' attitudes and beliefs about their own general efficacy in treating obesity may also influence their treatment of patients. In one study, physicians rated obesity treatment as less effective than therapies for 9 out of 10 chronic conditions, and only 14% agreed that they were usually successful in helping obese patients lose weight [21]. In another study, 31% of internal medicine residents believed that treating obesity is futile and only 44% felt qualified to treat obese patients [11]. We have found no studies of direct impact of negative physician attitudes on obesity care.

While attitudes are thought to influence physician behavior with regards to obesity counseling, we know very little about how these attitudes are formed or influenced. Social cognitive theory and the theory of planned behavior suggest that competence (skills or prior performance) is an important influence on attitudes [14, 15]. A few studies have tested this hypothesis in obesity with mixed results: one study showed that perceived competency was correlated with comfort level in treating pediatric obesity and the belief that obesity is a treatable condition [22], while another study did not find any link between knowledge (an essential determinant of competence) and attitudes [11].

Models of behavior change also suggest that social and environmental factors play an important part in shaping attitudes. For physicians these include individual physician characteristics (e.g., age, gender) [20, 23, 24], training (e.g., specialty, when trained, exposure to role models)[25]. Practice contexts (e.g., type of patients seen, public vs. private setting, inpatient vs. outpatient) may also influence attitudes. With the exception of physician bias [19, 26], few studies have characterized physicians' attitudes about obesity with regards to gender, specialty, training and practice characteristics. Further, while several studies have described individual physicians' attitudes about obesity [22, 27‐29], none have characterized how these attitudes relate to each other.

Additional studies are needed to elucidate the relative contribution of individual competence, physician, training, and practice variables on attitude formation. This is necessary in order to design more carefully targeted interventions and ultimately to be able to assess the impact of these attitudes on actual physician practice and patient outcomes.

As part of a needs assessment for a faculty development project to improve obesity care, we surveyed physician faculty members about their attitudes, obesity counseling competency (including report of percentage of their obese patients who lose weight), training, and practice characteristics. The first aim of this study was to understand and classify the nature of physician attitudes regarding obesity and examine how they differ by specialty. The second aim of this study was to better understand how competency in obesity counseling, physician characteristics, training, and practice characteristics are related to attitudes. The final aim was to explore the relative association of each of these on physician attitudes.

This is the first study that we know of to explore physicians' attitudes towards obesity with regards to competency, perceived patient outcome, physician, and practice characteristics. Based on the Theory of Planned Behavior and Social Cognitive Theory, we anticipated that perceived competency and skills would relate to physicians' attitudes. Physician Success/Self-efficacy was the attitude factor most associated with competency and reported patient weight loss. Whether this association is causal needs to be determined by prospective studies looking at actual physician performance and true patient outcomes. The direction of the association is unclear in our analysis. We cannot be certain whether physician self efficacy leads to improvements in competency and reported patient outcomes or vice-versa. Further, these associations may be artificial if, for instance, the Physician Success/Self Efficacy attitude factor and self-reported competency items are both measuring the same underlying variables. Indeed, the definition of self efficacy is "the belief that one can successfully execute the behavior required to produce the outcome [15]" which is similar to perceived competency. However, we believe that our competency and self-efficacy scales are, in fact, distinct. The competency items we used measured physicians' self-reported ability to perform specific 5A's related tasks while the Physician Success/Self Efficacy attitude factor measured a more global belief in their qualifications and success in treating obesity.

We found significant associations between the specific 5A's competency categories and particular attitude factors. The "Assess" competency score was negatively associated with Physician Discomfort/Bias and positively associated with Physician Success/Self Efficacy suggesting that physicians with lower competency assessing risk in obese patients may have more overall discomfort treating obese patients and lower self efficacy. The "Agree" competency score was positively associated with physician self efficacy indicating that physicians who are proficient in mutual goal-setting with patients may possess greater self-efficacy. Finally, the "Assist" competency score was positively associated with Positive Outcome Expectancy indicating that the ability to address barriers to weight loss and motivational interviewing may contribute to higher expectations that obese patients will be successful at losing weight. Overall, this pattern of results suggests that particular skills may be associated with the development and maintenance of particular attitudes. Since the 5A's competencies represent teachable skills, training may improve physician attitudes ultimately leading to better care for obese individuals.

Attitudes and competency varied by specialty. Psychiatrists reported being less competent performing many of the 5A's-related skills than the other specialties. This may be due to a lack of experience screening and counseling for obesity. Indeed, the psychiatrists indicated that only 19% of their patients were obese even though the prevalence of obesity in the psychiatric population is known to be higher than the general population [32, 33] which indicates that they may not be sufficiently recognizing obesity in their patients. However, psychiatrists also have less negative outcome expectancy than the other specialties which could facilitate receptiveness to acquire obesity counseling skills. We believe that psychiatrists should be trained to counsel patients to lose weight since many psychotropic medications cause weight gain, stimulate appetite, and increase cardiovascular risk [34, 35]. Further, obesity can be thought of as a neurobehavioral disorder [36] that psychiatrists' training should enable them to treat. Being a pediatrician was associated with having a higher positive outcome expectancy despite the fact that pediatricians reported a lower percentage of patients losing weight in their practice. Further examination of the relationship between specialty and attitudes is needed.

We also explored how physicians' gender is associated with obesity attitudes. Females were more likely to have higher positive outcome expectancies than males in bivariate analysis. However, when we controlled for other physician and practice characteristics, gender was no longer associated with any of the attitude factors. Teachman et al. also did not find gender differences in either implicit or explicit bias in physicians [24]. However, Schwartz et al. showed that being a male physician was associated with lower levels of bias [26]. Future studies should further explore the potential impact of gender on physician attitudes.

In addition to gender, outcome expectancies seem to be associated with the number of years elapsed since residency/postgraduate training. Physicians with greater time lapse since training (and thus presumably more experience) had higher Negative Outcome Expectancy and lower Positive Outcome Expectancy. We have at least two possible explanations for this. First, physicians may become disillusioned with years of attempted treatment of obese patients in the face of modest weight loss or even weight gain. Second, obesity management training may be different for recent graduates who practice in an era in which it is widely accepted that even modest weight loss should be considered a successful outcome [37]. Thus, they may view success differently than their more senior peers, which may be reflected in more positive outcome expectancies. However, both of these hypotheses are speculative and these results need to be replicated. Indeed, Ferrante et al. found that older physicians were actually less likely to agree that treatment of obesity is ineffective than younger physicians [38].

Unfortunately, our models only account for 4–33% of the variance in physicians' attitudes about obesity suggesting that other unmeasured factors are more important influences on physician attitudes. For example, physicians' health behaviors potentially play a significant role in shaping their attitudes about obesity. Several studies show that personal health promotion behavior in physicians is a consistent and powerful predictor of attitudes and/or professional practice towards obesity-related behavior change counseling [39‐43]. Physicians who try to exercise more and maintain a healthy diet are significantly more likely to discuss exercise and weight with their patients and report greater counseling self-efficacy [44]. Thus, a limitation of our study is that we did not include physicians' own behaviors in the survey.

There are several other limitations as well. We relied on physician self-report for competency and patient weight loss items rather than direct measurement. Several studies have suggested that physicians may not be reliable evaluators of their competency [45]. However, our measure may be more sensitive than other measures of competency as it evaluates 15 behaviorally-specific skills. We are in the process of evaluating this measure against actual physician performance. Further, it is unclear how accurately physicians were able to estimate how many of their obese patients lose weight. All the physicians were from the faculty at New York University School of Medicine which may limit generalizability. However, the physicians surveyed practice in a wide array of settings–in both the public and private sector, hospital and clinic settings–and represent 3 different medical specialties. Our findings are also limited in that attitudes items measured do not represent the full spectrum of attitude measures in the literature.

Our study suggests avenues for future research. This work will help us to design future studies to assess the impact of physician attitudes on performance and patient care. Observed structured clinical exams using standardized patients (OSCEs) are often used to measure physician performance, and we plan to measure how attitudes relate to OSCE performance. Further, we propose doing chart reviews and patient exit interviews to help determine how physician attitudes relate to real rather than perceived outcomes in overweight and obese patients such as eating and exercise behaviors and weight loss.

Melanie Jay, MD, MS is a general internist and Clinical Assistant Professor at New York University School of Medicine. She is a member of Research on Medical Education Outcomes Team (ROMEO). Colleen Gillespie, PhD is a community psychologist and Assistant Professor of Medicine. She is a member of ROMEO. Sheira Schlair, MD, MSc is a general internist and clinical instructor at New York University School of Medicine. She is a member of ROMEO. Tavindar Ark has a Masters degree in psychology and is a Research Associate at New York University School of Medicine. She is a member of ROMEO. Scott Sherman MD, MPH is a general internist and geriatrician. He is an Associate Professor of Medicine at New York University School of Medicine. Michelle McMacken MD, MPH is a general internist and a Clinical Assistant Professor at New York University School of Medicine. She is director of the Weight Management Program at Bellevue Hospital. Mary Jo Messito, MD is a pediatrician and an Assistant Clinical Professor at New York University School of Medicine. Sondra Zabar, MD is an associate professor of medicine at New York University School of Medicine and the program director of the NYU primary care residency program. She is a member of ROMEO. Adina Kalet, MD, MPH is a general internist and Professor of Medicine at New York University School of Medicine. She has done extensive research in medical education and is the leader of ROMEO.