Abstract 272: Severity and Determinants of Underreporting of Obesity in a Diverse Population of Hospitalized Patients in an Academic Medical Center and its Impact on Mortality

Abstract

Listen

Translate article

Obesity has become a national epidemic with significant implications in the development of multiple diseases. Prior studies have shown that physicians often fail to diagnose obesity in the outpatient settings, and consequently fail to formulate treatment plans and medical interventions. With the introduction of electronic medical records, making the diagnosis of obesity in an inpatient setting is a straightforward process. Our objective was to determine the presence and extent of underreporting of obesity in hospitalized patients in an urban academic medical center and to identify determinants and consequences of underreporting. We used a proprietary electronic medical record aggregator platform (Clinical Looking Glass) to collect data on all adult admissions during 2007-2010 to Montefiore Medical Center, a tertiary medical center. Presence or absence of diagnosis of obesity in patients’ problem list was identified by examining inpatient problem list in the electronic record. Our study population included 22,857 obese adults (BMI of ≥ 30 kg/m 2 ) of whom 66 % were female, 37% Black, 18.45% White, and 37.9% Hispanic. Overall, only 18.4% of the study population carried the diagnosis of obesity. There was significant variation in rates of correct diagnosis among different subgroups: 20.4% in women vs. 14.4% in men; and 22.4% in Blacks vs. 13.6% in Whites. There was a linear correlation between severity of obesity and correct diagnosis (53.38% of patients with BMI >45 kg/m 2 vs. 7.89% with BMI 30-35 kg/m 2 ) (Fig. 1). In Kaplan-Meier survival analysis, patients with correctly reported obesity demonstrated decreased mortality rates after an average follow up of two years (6.4%, CI 5.6-7.3 vs. 11.4% CI 10.9-12). In conclusion, the correct diagnosis of obesity is severely underreported in hospitalized adult patients, with race, ethnicity, gender and degree of obesity being important determinants of this observation. Correct diagnosis of obesity is associated with decreased mortality over a two-year follow-up.