Introduction ============ Bipolar disorder (BD) is a severe mental disorder, characterized by a significant level of disability and suffering. According to the European Bipolar Disorders and Other Disorders Network, BD is one of the leading causes of disability across all age groups.^[@bib1]^ The World Health Organization reported that mental and behavioral disorders, which include bipolar disorders, were one of the leading causes of years lived with disability (YLDs) in 2010 and 2013.^[@bib2],\ [@bib3]^ It is important to realize that the number of people affected by this disorder is on the rise and, therefore, it has been predicted that this disease will be the second leading cause of disability by the year 2020.^[@bib4]^ The literature provides several estimates of the total number of people suffering from bipolar disorder and the costs generated by the disease. Bipolar disorder is not simply a psychiatric disorder, but is associated with substantial comorbidity^[@bib5]^ and comorbidity affects the prognosis of BD and the patients\' functioning.^[@bib6]^ There is no doubt that the economic and social impact of bipolar disorder generates substantial costs for healthcare systems and society.^[@bib7]^ Bipolar disorder leads to a heavy economic burden for health services and society because of its impact on morbidity, mortality, life expectancy, treatment costs and quality of life.^[@bib8]^ Some of the major factors that lead to the economic burden of bipolar disorder are the frequency of recurrences, suicide rates, number of hospitalizations, disability and reduced quality of life. Studies have also focused on the impact of bipolar disorder on work productivity, the risk of premature death, costs of lost productivity due to disability and absenteeism, and costs related to premature mortality.^[@bib9],\ [@bib10],\ [@bib11]^ The objective of this paper is to review the published economic evidence regarding bipolar disorder and provide estimates on the direct and indirect costs, productivity loss and burden of illness of bipolar disorder in the countries of the European Union (EU) over the past 15 years. Materials and methods ===================== Search strategy --------------- A systematic review was carried out in four steps. First, we searched for relevant studies in the following databases: PubMed (from 1966), the Cochrane Database of Systematic Reviews, the database of abstracts of reviews of effectiveness (DARE), EconLit, CINAHL, Science Direct, ISI Web of Science, LILACS and Embase. Second, we searched for relevant studies in: PsycInfo and MEDLINE, covering the period from January 2003 to December 2013. Third, a search on health technology assessment (HTA) agencies\' websites was conducted to identify economic evaluations published by HTA agencies. A manual search for economic evaluations was also performed in various HTA reports. Fourth, the study was complemented by screening of the reference lists of the included publications and other reviews that met our inclusion criteria. Inclusion and exclusion criteria -------------------------------- The studies included in the review were comparative studies with a focus on the economic burden of the disease. Data from studies published until December 2013 in English, German, Spanish and French languages were included. There was no limit to year of publication. Studies were included in the analysis if they: (1) included data from countries within the EU, (2) included patients older than 18 years of age, (3) described an economic model with data from at least one sample of a population with bipolar disorder and (4) compared health outcomes between patients and controls (healthy people, other disease groups). A list of search terms for all studies included in the search are provided in [Table 1](#tbl1){ref-type="table"}. A systematic review was carried out in four steps. First, a search was conducted on relevant studies in electronic databases. Second, a manual search for relevant studies was performed, covering the period from January 2003 to December 2013. Third, a search on health technology assessment (HTA) agencies\' websites was conducted to identify economic evaluations published by HTA agencies. Finally, a manual search for economic evaluations was also performed in various HTA reports. We used the World Health Organization International Statistical Classification of Diseases and Related Health Problems, 10th edition (ICD-10) criteria to identify papers of economic evaluations that compared patients with bipolar disorder with healthy controls. The criteria are presented in [Table 2](#tbl2){ref-type="table"}. Results ======= Only four studies were included in the systematic review: one study presented a cost-minimization analysis, one study included cost-effectiveness analysis, one included cost-utility analysis and one study provided a probabilistic cost-effectiveness analysis. The total costs from the four studies are shown in [Table 3](#tbl3){ref-type="table"}. The cost-minimization analysis by Sharma and colleagues^[@bib13]^ was conducted in 2004 and aimed to compare costs and consequences of lithium treatment versus treatment with an antidepressant for bipolar affective disorder. The results showed that the expected cost for lithium treatment was estimated at £913 compared with £1499 for the antidepressant treatment. However, the antidepressant treatment showed a gain in quality-adjusted life years of 0.07, which led to a difference of £25 000/QALY in favor of the antidepressant treatment. Therefore, the authors concluded that lithium treatment did not dominate antidepressant treatment. The cost-effectiveness analysis by Schmitt *et al.*^[@bib14]^ compared monotherapy with aripiprazole to combination therapy with lithium plus valproate. This study was conducted in 2008 and used the societal perspective and time frame of 30 years. A total of 38% of the patients were considered to be responders to treatment, that is, they did not experience any recurrence, morbidity or mortality. The total mean direct cost was significantly lower for aripiprazole than for lithium plus valproate. The results showed an additional average cost for the lithium plus valproate therapy of €28 734 per patient and year. For each patient who gained response to lithium, the extra costs generated was €27 649. The cost-effectiveness analysis by Alonso *et al.*^[@bib15]^ was conducted in 2011 and compared the use of lamotrigine and lithium as maintenance treatment for bipolar disorder. The primary endpoint was to identify the most cost-effective treatment, given that lithium was slightly less expensive and had better outcomes on quality of life. However, lithium was not significantly more effective than lamotrigine. The results showed that costs per quality-adjusted life year gained were €13 320 with lamotrigine, €22 869 with lithium and €25 110 with quetiapine. However, according to the deterministic and probabilistic sensitivity analysis, lithium was significantly less expensive and more effective than lamotrigine or quetiapine. The results also showed a greater effect of the treatment on the quality of life in patients treated with lithium compared with lamotrigine or quetiapine. The cost-utility analysis by Witter *et al.*^[@bib16]^ used the societal perspective and was conducted in 2009 and measured the differences between valproate monotherapy and quetiapine monotherapy. The outcomes were the mean cost per quality-adjusted life year gained, costs per treatment success and overall costs. Results showed that the mean cost per quality-adjusted life year was estimated at €9 000 for quetiapine compared with €27 000 for valproate. When combining the cost per treatment success (remission and response) and total costs, lithium showed the best results with a mean value of €14 100 per patient and year. For valproate, the costs were €42 000 and for quetiapine, it was €48 000. The cost-effectiveness analysis by Geddes and colleagues^[@bib17]^ presented results from a meta-analysis. The first part of the analysis by the authors used a systematic review and meta-analysis of randomized clinical trials of treatment with lamotrigine versus other standard treatments for bipolar disorder. The base case considered studies of lithium, divalproex sodium and carbamazepine. The results showed that lamotrigine was associated with lower costs compared with lithium, divalproex sodium and carbamazepine. Overall, the mean difference in costs between the treatments was −€1152 per patient. The second part of the analysis included cost comparisons from different studies to compare lamotrigine with other therapies for bipolar disorder. A cost-utility analysis of a study conducted by Bauer *et al.*^[@bib18]^ showed a cost-effectiveness ratio of −€943 per quality-adjusted life year. The economic impact of bipolar disorder was also evaluated from the viewpoint of patients\' time lost due to relapses. A loss of productivity is estimated to be at a cost of £40 811 per year for an individual patient and at a cost of £27.3 billion for the UK economy. Results from the review showed that the total direct and indirect costs were estimated at £8.9 billion. The authors concluded that this study confirmed the substantial cost associated with bipolar disorder, based on studies carried out in Italy and Spain. In conclusion, lithium and valproate are the two most commonly prescribed drugs for bipolar disorder. From the review, these drugs showed the lowest costs and best outcomes for quality of life. Despite the lack of available evidence in this field, it is important to realize that treatment with lithium or valproate provides greater value for money. Discussion ========== The main aim of this review was to provide an estimate of the costs associated with bipolar disorder in countries within the EU. The results of this