New study: Assumptions, stigmatising language, and sub-standard practice from student doctors

Title: Influence of weight etiology information and trainee characteristics on physician-trainees’ clinical and interpersonal communication

Authors: Cohen, Persky

Source: Patient Education and Counseling, 2019, 102: 1644–1649. Link to paper

Super-quick summary: Student doctors read an article about behavioural or genetic causes of “overweight” or a neutral article about chronic headaches. They then interacted with a virtual “obese” patient who provided no information about her diet or exercise, but did talk about a previous diagnosis of osteoarthritis but said she was unclear what that meant. Although some effects were stronger in the ‘behavioural’ condition, students in all conditions talked about her weight, provided lifestyle advice, made assumptions about her current diet and exercise level, used stigmatising language, and didn’t provide any information to help her understand her osteoarthritis diagnosis.

In more detail:

What they did

The authors recruited 119 US physician trainees in either their 3rd or 4th year of medical school or their first year of residency in internal medicine and family practice. The students were randomly assigned to read one of three articles, either about behavioural or genetic causes of “overweight” or an unrelated subject (chronic headaches). They then interacted with an “obese” digital patient in a virtual reality environment, interacting with her as they would in a normal consultation.

The patient listed her main symptoms as knee pain, shortness of breath, and a hand rash. She also said her knee pain had previously been diagnosed as osteoarthritis but that she wasn’t sure what that meant. She provided no information about her diet or exercise. Students had to provide the patient with counselling as they felt appropriate. The conversations were recorded and then coded for a range of outcomes.

What they found

Nearly all the students in the behavioural and genetic conditions talked about the patient’s weight, although nearly  half in the control condition did also. Around 60% of students in the behavioural condition, nearly half in the genetic condition, and a quarter in the control (headache) condition, made assumptions about her lifestyle (eating behavioural and physical activity) without actually asking for this information. Most of the students in the behavioural and genetic conditions, and nearly half in the control condition provided lifestyle counseling. Over half of the students in the behavioural condition used words previously found to be stigmatising to patients (including fat, obesity, obese, and morbidly obese), but so did around 30% in the genetic condition and 15% in the control condition. Across all three conditions, between 30 and 35% of students provided information about her osteoarthritis diagnosis. Students’ own BMI didn’t make any difference to any of these outcomes, nor did the students’ gender.

What it means

Although the point of the study was to show the additional impact on stigma outcomes of providing information emphasising that “overweight” is due to behavioural “failings”, I think the most interesting outcome of this study was just how badly everyone did. Most talked about her weight, made assumptions, told her how to change her lifestyle, and many used stigmatising language. And over two-thirds failed to provide her with information about her osteoarthritis despite being given a clear indication that this was an information gap for the patient.