Systematic reviews of case reports and patient level analysis as a new tool to study rare adverse events of medications: the case of neuroleptic malignant syndrome

Article type
Authors
Gastaldon C1, Guinart D2, Misawa F3, Rubio JM2, de Filippis R4, Schoretsanitis G2, Barbui C1, Kane JM2, Correll CU5
1WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation Department of Neuroscience, Biomedicine and Movement Sciences Section of Psychiatry, University of Verona
2Division of Psychiatry Research, Zucker Hillside Hospital, Northwell Health, New York; Center for Psychiatric Neuroscience, Feinstein Institute for Medical Research, Manhasset; Department of Psychiatry, Zucker School of Medicine at Nort.
3Yamanashi Prefectural KITA Hospital, Yamanashi
4Psychiatric Unit, Department of Health Sciences, University Magna Graecia of Catanzaro, Catanzaro
5Division of Psychiatry Research, Zucker Hillside Hospital, Northwell Health, New York; Department of Child and Adolescent Psychiatry, Charité-Universitätsmedizin Berlin, Berlin
Abstract
Background: Neuroleptic Malignant Syndrome (NMS) is a rare, potentially fatal condition associated with dopamine receptor blocking agents, such as antipsychotics (APs). Due to its relatively low incidence (~ 1% of antipsychotic users), epidemiological and experimental studies are almost absent. High severity low incidence conditions, such as NMS, may therefore require alternative research approaches. Using NMS as a paradigm, we propose to conduct a systematic review and assessment of published case reports, supplemented by an individual patient case report meta-analysis.
Objectives: To pilot the feasibility of developing a patient-level database based on a systematic literature review of NMS case reports/series, and to explore the applicability of individual patient case report data analytical approaches. In particular, we aim to address association of NMS with specific medications, and the clinical characteristics and outcomes of NMS.
Methods: At least 2 authors conducted a systematic and independent search in Medline, Embase, Cochrane, CINAHL and PsychINFO databases. Included were all case reports describing NMS that occurred during ongoing AP treatment or within one injection interval of a long-acting injection (LAI) in adults. Each included case report was reviewed and extracted by ≥2 physicians independently to extract demographic, clinical, treatment and outcome data. NMS severity was coded using the Francis-Yacoub scale.
Results: The database was created 10/2018 and last updated 01/2020. To date, it includes 690 case reports of NMS. Clinical characteristics and outcomes of NMS (frequency of complete recovery, incomplete recovery or death, duration of NMS, and length of hospital stay) were compared between different types of AP formulations and class as well as between monotherapy and polytherapy. The analyses were adjusted for between-group differences and potential confounders in a multivariable regression model.
Conclusions: To our knowledge, this is the first systematic review and individual patient case report meta-analysis of a severe rare adverse event, such as NMS. Since NMS is so infrequent, it is unlikely to observe a significant number of cases in alternative designs such as epidemiological or experimental studies. Hence, our approach provided more information than what could have possibly been made available by epidemiological studies. For example, the largest 11-year longitudinal register case-control study of NMS (based on the Psychiatric Danish Register) gathered only 83 cases of NMS. We retrieved 690 cases and compared characteristics and outcomes of NMS occurring during treatment with different AP classes and formulations. Therefore, the results from our study are clinically relevant, innovative and could be one of the few possible strategies to generate relevant clinical and treatment evidence for rare adverse events. Individual patient case report meta-analyses, although sensitive to reporting bias, can arguably be informative to researchers and clinicians, and can help guide the design of future research studies.