Prescription sequences in bipolar disorder – A nationwide Danish register-based study of 19,927 individuals followed for 10 years

Evidence-based use of pharmacological interventions in bipolar disorder is of paramount clinical importance. We aimed to uncover precription sequences in a large cohort of patients from the first diagnosis of bipolar disorder. Using Danish nationwide registers, we identified individuals with a first-time hospital diagnosis of bipolar disorder between January 1^st, 2001, and December 31^st, 2016. Redemeed prescriptions of litihum, anticonvulsants, antipsychotics, and antidepressants from five years before to five years after diagnosis were retreived. The data were analysed with descriptive statistics, sunburst plots, and Cox proportioal hazard models. The full study population consisted of 19,927 individuals. Before diagnosis, antidepressants were the predominantly prescribed group (46.9 % as first drug). After diagnosis, a major trend towards mood stabilising strategies was observed. although only 18.7 % received lithiumas first prescription. In analyses stratified for illness phase, lithium was more frequently prescribed as first drug after depression than after hypomania/mania, in which antidepressants were used as first drug in 10-15 % of the cases. Treatment sequences were highly heterogeneous (2,459 distinct sequences for the 19,927 individuals under investigation). Lithium appeared to carry the overall highest risk of treatment shift. We conclude that in accordance with national and international guidelines, a diagnosis of bipolar disorder leads to a relevant change of treatment strategy towards mood stabilising drugs. However, lithium continues to be underused;antidepressants probably used too frequently, and treatment sequences are highly heterogeneous and not adjusted according to illness phase. These results point to a potential for optimising the real-world pharmacological management of bipolar disorder.