Background and aims of the study: Mental disorders are frequently investigated in economic evaluations. However, measuring direct costs of mental disorders is complex, in particular in Germany. We conducted a systematic review that investigated the following research questions: Which cost categories are assessed? Is the used data source (patient-reported data from interviews or questionnaires; claims data from health insurance; medical records from patient charts) associated with specific study characteristics?
Methods: We conducted a systematic literature search in PubMed. We included cost-of-illness studies (COIs) and cost-effectiveness analyses (CEAs) that measured patient level data on direct costs of mental disorders in Germany.
Results: We found 31 COIs and 17 CEAs. Cost items could be assigned to the cost categories outpatient medical, outpatient non-medical, sheltered living, hospital, rehabilitation, nursing care, outpatient goods (medication and medical goods), other direct costs, and indirect costs. Most studies, in particular COIs and studies conducted from a societal perspective, used patient-reported data which suited to assess cost categories comprehensively. Studies based on claims data or medical records tended to include fewer cost categories.
Discussion: The sample size of 48 analyzed studies was rather small. However, our study investigated three methods to assess health care utilization -- using patient-reported data, claims data, or medical records -- and identified their respective strengths and weaknesses in the context of covered cost categories and further attributes.
Conclusion: Studies measuring the direct costs of mental disorders in Germany mostly relied on patient-reported data to assess resource consumption. Although more objective, claims data were used less frequently and covered less cost categories.
Implications for health policies: The direct costs of mental disorders are likely to be underreported as relevant cost factors such as non-medical services or sheltered living were often neglected in the studies. Policy makers should consider this when deciding on measures that aim at reducing the costs of care.
Implications for further research: Eliciting data from patients seems to constitute an adequate instrument to assess a broad range of cost categories, even though this bears various methodological challenges. Further research should (i) ameliorate the methods of collecting data from patients in ensuring that all relevant costs are covered and that methodological flaws are minimised (ii) develop or improve methods for the combination of claims data from different data holders.