Abstract / Summary
Obesity, a multifactorial chronic condition shaped by biological, environmental, and social determinants, is often inadequately measured, monitored, diagnosed or managed. Therefore, we aimed to study the prevalence and management of obesity within the Western healthcare cluster of Singapore using a novel AI-driven dashboard called Adipoview. A population-based cross-sectional study was conducted on a cohort of adult patients (≥ 18 years old) captured within the electronic health records (EHR) from 23 August 2023 to 2024. Deidentified patient-level data was derived from Adipoview, a dashboard that uses various data analytics and visualisation methods to display the obesity management landscape in real-time. Sociodemographic (i.e. age, sex, race) and clinical characteristics (i.e. obesity classification, obesity diagnosis, co-morbidities, and obesity treatment) were extracted. Our study analysed 313,369 patients with a mean age of 58.3 (17.1) years, with a female majority (53%). The most common co-morbid chronic diseases were hyperlipidaemia (40.0%), hypertension (36.7%), and diabetes (22.1%), where patients from primary care tended to be older with a higher prevalence of DHLs. Among patients meeting BMI-based criteria for excess adiposity, only 10.5% had a documented diagnosis. The proportion of patients meeting BMI criteria was slightly higher in primary compared to tertiary care (89.8% vs. 86.1%). Our prevalence number was expectedly larger than that derived from the National Population Health Survey (NPHS) (58.9%). Patients aged 18–74 years old with overweight and obesity were 18.1% and 56.9% respectively. Among patients referred to the Centre for Obesity Management and Surgery (COMS), 7,761 (68.9%) were adherent to follow-ups. The high prevalence of excess adiposity paired with its underdiagnosis as a chronic disease limits the potential to curb cardiometabolic implications of obesity. Greater efforts are needed to increase weight management awareness, refine obesity assessment measures, ensure proper diagnosis using clinical decision support aids, review existing weight-loss programme efficacies and promote screening of comorbidities through real-time visualisation tools such as Adipoview.