Time to lithium: an electronic health records study

Background: Lithium monotherapy has been recommended as a first-line maintenance or long-term pharmacological therapy for bipolar disorder (BD) by numerous clinical guidelines including the UK National Institute for Health and Care Excellence (NICE guidelines), the British Association for Psychopharmacology (BAP) guidelines and the Canadian Network for Mood and Anxiety Treatments (CANMAT) guidelines. This is unsurprising given the existing evidence for its efficacy in prophylaxis of mood episodes [1] and prevention of completed suicide in mood disorders [2]. Previous research has linked early lithium use with better outcomes for patients with BD [3]. Despite its extensive evidence as an effective treatment for BD, lithium prescription has declined in the past years [4]. Potential reasons for this decline have been recently studied and include the widespread use of second-generation antipsychotics and a lack of commercial support due to not having a medication patent [5]. Objectives: With the present study, our primary aim was to determine the time between the initial assessment with a mental health professional and the initiation of lithium. Secondary objectives included determining the time between first assessment and recorded diagnosis of BD, number of mood episodes experienced by those with BD before lithium was prescribed, predominant mood polarity when lithium was prescribed and number of antipsychotics before lithium was initiated. Methods: Free-text clinical notes were extracted from a de-identified electronic health record database. Inclusion criteria were: a) having an ICD-10 diagnosis of F30 (Manic episode) or F31 (Bipolar affective disorder) recorded between 1 January 2015 and 01 January 2020, b) being aged 18 years on the date of the first recorded F30 or F31 diagnosis, c) having had at least one face-to-face assessment with a medically qualified practitioner, d) receiving treatment with lithium at a therapeutic dose, ascertained by a laboratory recorded lithium serum level above 0.0 mmol/L e) having a continuous first episode of care within SLaM for at least six months during which the F30 or F31 diagnosis was recorded. Exclusion criteria were: a) having any recorded treatment or episode of care outside of SLaM before lithium was prescribed, or b) unclear or incomplete clinical documentation. Data were extracted on 16 February 2021 and free-text clinical notes were chronologically read line-by-line by a psychiatrist. Language processing or other data algorithms were not used. Results: A cohort of 88 patients was identified based on inclusion and exclusion criteria. The median time between the first assessment and prescription of lithium was 588 days. The median time between assessment and diagnosis of BD was 220 days, and patients presented with a median of 2.5 mood episodes and were prescribed a median of 2 antipsychotics prior to lithium prescription. Around 30% of patients presented with manic polarity symptoms at the time of lithium prescription. Conclusions: Our study identified a significant delay between initial assessment with a clinician and the prescription of lithium. In our opinion, interventions are urgently needed to reinforce the prescription of lithium, given its proven efficacy and links to better prognosis.