Socioeconomic disparities in national physician practice patterns of obesity diagnosis and management

Background

Limited research has explored socioeconomic disparities in physician diagnosis and management of obesity despite clear evidence of the disproportionate impact of excess weight on minority and disadvantaged communities.

Data and Methods

This study uses data from the 1995-6 (N = 30929) and 2005-6 (N = 21220) National Ambulatory Medical Care Survey (NAMCS) – an annual survey of patient visits reported by physicians – to conduct a cross-sectional analysis of patients aged 18 and older. Using the 2005-6 NAMCS, this study examines socioeconomic disparities in national patterns of physician obesity care and the predictors which may influence these practices. Using both datasets, this study evaluates the impact of a height/weight prompt on physician behavior. (The 2005-6 NAMCS is the first time height/weight measurements were included in the survey. In the 1995-6 NAMCS, a check box was used to indicate an obesity diagnosis). The primary outcome variables of interest are patient receipt of obesity diagnosis or management. Obesity management is characterized as counseling (e.g., diet/nutrition, exercise, and weight reduction) and screening (e.g., blood pressure and cholesterol). Chi-squared tests and multivariate logistic regression are used to test study hypotheses. Data are weighted to be representative the U.S. population.

Results

Physicians diagnosed less than a tenth of adults (8.3%) as obese in the 2005-6 NAMCS (via check box), while more than a third of the sample was clinically obese (36.4%) based on measured height and weight. Among patients with an obesity diagnosis, physicians provided weight reduction counseling half of the time (48%). Blood pressure was measured at 74 percent of visits by patients with an obesity diagnosis while and cholesterol testing was administered at 14 percent. The likelihood of an obesity diagnosis varied significantly by patient characteristics. Physicians were more likely to diagnose obesity among females, (p < 0.001), Hispanics (p < 0.001), younger patients (ages 25-44) and patients living the Midwest (p < 0.001). The biggest predictor of obesity management was an obesity diagnosis (p < 0.001). Other significant predictors of obesity management included: race, age, type of visit (preventive vs. illness/injury), geographical region and physician specialty. The addition of height and weight measures to the 2005-6 NAMCS is associated with significantly higher rates of physician management of obesity (p < 0.001) (as compared to the 1995-6 NAMCS).

Conclusions

Physician diagnosis and management of obesity differs significantly by socioeconomic status. A height/weight prompt may encourage physicians to provide a higher frequency of obesity care.