Purpose: There is a lack of comprehensive cost-of-illness studies in bipolar disorder, in particular studies based on patient-level data. The purpose of this study was to estimate the societal cost of bipolar disorder and to relate costs to disease severity, depressive episodes, hospitalisation and patient functioning.
Methods: Retrospective resource use data in inpatient and outpatient care during 2006-2008, as well as ICD-10 diagnoses and Global Assessment of Functioning (GAF) scores, were obtained from the Northern Stockholm psychiatric clinic with a catchment area including 47% of the adult inhabitants in Stockholm. This dataset was combined with national register data on prescription pharmaceuticals and sick leave to estimate the societal cost of bipolar disorder. The study was conducted from a societal perspective, with indirect costs valued according to the human capital method.
Results: The average annual cost per patient was <euro>28,011 in 2008 (n = 1,846). Indirect costs due to sick leave and early retirement represented 75%, inpatient costs 13%, outpatient costs 8%, pharmaceuticals 2% and community care another 2% of the total cost. Total costs were considerably higher during mood episodes (six times higher than in remission), for hospitalised patients (<euro>55,500 vs. <euro>22,200) and for patients with low GAF scores.
Conclusions: The high cost of bipolar disorder is driven primarily by indirect costs. Costs were strongly associated with mood episodes, hospitalisations and low GAF scores. This suggests that treatment that reduces the risk for relapses and hospitalizations and improve functioning may decrease both the societal cost of bipolar disorder and patient suffering.