Association between severity of symptoms and minimum mental health treatment duration in humanitarian contexts: a retrospective observational study.

Background: Understanding and optimising mental health and psychosocial support (MHPSS) interventions in humanitarian crises is crucial, particularly for the most prevalent mental health conditions in conflict settings: anxiety, depression, and post-traumatic stress disorder. However, research on what is the most appropriate length of psychological intervention is lacking in this setting. We aimed to establish which factors are most closely related to improvement and to determine the required number of consultations needed to achieve this improvement. Methods: We retrospectively analysed records from 9028 patients allocated to treatment for anxiety, depression, and post-traumatic symptoms from the MHPSS programme in Borno State, Nigeria, from January 2018 to December 2019. Patient characteristics, severity (Clinical Global Impression of Severity Scale, CGI-S scale), and clinical improvement were assessed by an attending counsellor (CGI-I scale) and by the patient (Mental Health Global State, MHGS scale). Improvement was defined as scores 1, 2, and 3 in the Clinical Global Impression of Improvement (CGI-I) scale, and as a decrease of at least 4 points in the MHGS scale. We investigated the associations between the category of symptoms, the severity of illness, and improvement of symptoms using multivariable logistic regression. We used Kaplan-Meier (KM) curves to assess the number of consultations (i.e., time of treatment) needed to achieve improvement of symptoms, by symptom category and symptom severity. Findings: The patients included were referred to treatment for anxiety (n = 3462), depression (n = 3970), or post-traumatic symptoms (n = 1596). Median age was 31 years (range 16-103), and 84.3% were female. Patients categorised as severe were less likely to present improvement according to the CGI-I scale (OR 0.11, 95% CI 0.05-0.25), while none of the other categories of symptoms showed significant results. Overall, three or more consultations were associated with improvement in both scales (OR 3.55, 95% CI 1.47-8.57 for CGI-I; and OR 3.04, 95% CI 2.36-3.90 for MHGS). KM curves for the category of symptoms showed that around 90% of patients with anxiety, depression, or post-traumatic symptoms, as well as those with mild or moderate severity, presented improvement after three consultations, compared with six consultations for those with severe symptoms. Interpretation: Classification by severity among patients with anxiety, depression, or post-traumatic symptoms could predict the probability of improvement, whereas classification by symptoms could not. Our study highlights the importance of classifying patient severity in MHPSS programmes to plan and implement the appropriate duration of care. A major limitation was the number of patients lost to follow up after the first consultation and excluded from the logistic regression and KM analysis. Funding: The study was funded and staffed entirely by Médicos Sin Fronteras (Médecins Sans Frontières), Spain.