Life Expectancy, Loss of Life Expectancy, and Lifetime Costs in Bipolar Disorder: A Nationwide Population-Based Study

Bipolar disorder (BD) significantly affects life expectancy (LE), results in substantial loss of LE, and contributes to high medical costs, with these impacts varying by age at onset and gender. Previous studies have often overlooked the significance of age at the onset when estimating LE in individuals with BD. This study aimed to address this limitation and assess the impacts of BD on LE, loss of LE, and medical costs for BD patients categorized by age and gender in Taiwan using a new semiparametric extrapolation method over an 11-year duration.

A rolling-over algorithm estimated the survival function, with lifetime risk extrapolated. LE and loss of LE were calculated by comparing BD patients to matched non-BD referents by sex, age, and diagnosis year. Lifetime medical costs were determined by multiplying average monthly expenses by survival rates. Data from Taiwan's National Health Insurance (2009-2019) identified BD patients aged 5-84 with ≥ 2 outpatient or ≥ 1 inpatient BD diagnosis. The semiparametric survival extrapolation method was validated by comparing it with the Kaplan-Meier analysis.

The results indicate that following a BD diagnosis, patients have an LE of 26.79 years, reflecting a loss of 15.08 years compared to matched referents. On average, patients with BD incurred annual medical expenses of around $2516, with costs rising with age for both sexes. The mean estimated lifetime cost for the study population was about $55,015. The extrapolation method demonstrated high accuracy, with a less than 5% relative bias.

Semiparametric extrapolation is an effective method for estimating LE, loss of LE, and lifetime costs in BD. Future work could refine semiparametric extrapolation and assess factors influencing LE loss and lifetime costs in BD.