Direct and indirect costs for adverse drug events identified in medical records across care levels, and their distribution among payers

Jennie Natanaelsson, Katja M. Hakkarainen, Staffan Hägg, Karolina Andersson Sundell, Max Petzold, Clas Rehnberg, Anna K. Jönsson, Hanna Gyllensten

Adverse drug events (ADEs) cause considerable costs in hospitals. However, little is known about costs caused by ADEs outside hospitals, effects on productivity, and how the costs are distributed among payers. To describe the direct and indirect costs caused by ADEs, and their distribution among payers. Furthermore, to describe the distribution of patient out-of-pocket costs and lost productivity caused by ADEs according to socio-economic characteristics. In a random sample of 5025 adults in a Swedish county, prevalence-based costs for ADEs were calculated. Two different methods were used: 1) based on resource use judged to be caused by ADEs, and 2) as costs attributable to ADEs by comparing costs among individuals with ADEs to costs among matched controls. Payers of costs caused by ADEs were identified in medical records among those with ADEs (n = 596), and costs caused to individual patients were described by socio-economic characteristics. Costs for resource use caused by ADEs were €505 per patient with ADEs (95% confidence interval €345-665), of which 38% were indirect costs. Compared to matched controls, the costs attributable to ADEs were €1631, of which €410 were indirect costs. The local health authorities paid 58% of the costs caused by ADEs. Women had higher productivity loss than men (€426 vs. €109, p = 0.018). Out-of-pocket costs displaced a larger proportion of the disposable income among low-income earners than higher income earners (0.7% vs. 0.2%-0.3%). We used two methods to identify costs for ADEs, both identifying indirect costs as an important component of the overall costs for ADEs. Although the largest payers of costs caused by ADEs were the local health authorities responsible for direct costs, employers and patients costs for lost productivity contributed substantially. Our results indicate inequalities in costs caused by ADEs, by sex and income.