Effectiveness of pharmacological treatments for severe agitation in real-world emergency settings: protocol of individual-participant-data network meta-analysis
Spyridon Siafis,Hui Wu,Nobuyuki Nomura,Johannes Schneider-Thoma,Irene Bighelli,Carolin Lorenz,Joseph E Dib,Prathap Tharyan,Leonie A Calver,Geoffrey K Isbister,Esther W Y Chan,Jonathan C Knott,Celene Y L Yap,Célia Mantovani,Marc L Martel,David Barbic,William G Honer,Wulf-Peter Hansen,Gisele Huf,Jacob Alexander,Nirmal S Raveendran,Evandro S F Coutinho,Josef Priller,Clive E Adams,Georgia Salanti,Stefan Leucht
DOI: https://doi.org/10.1186/s13643-024-02623-z
2024-08-02
Abstract:Background: Severe psychomotor agitation and aggression often require immediate pharmacological intervention, but clear evidence-based recommendations for choosing among the multiple options are lacking. To address this gap, we plan a systematic review and individual-participant-data network meta-analysis to investigate their comparative effectiveness in real-world emergency settings with increased precision. Methods: We will include randomized controlled trials investigating intramuscular or intravenous pharmacological interventions, as monotherapy or in combination, in adults with severe psychomotor agitation irrespective of the underlying diagnosis and requiring rapid tranquilization in general or psychiatric emergency settings. We will exclude studies before 2002, those focusing on specific reasons for agitation and placebo-controlled trials to avoid concerns related to the transitivity assumption and potential selection biases. We will search for eligible studies in BIOSIS, CENTRAL, CINAHL Plus, Embase, LILACS, MEDLINE via Ovid, PubMed, ProQuest, PsycINFO, ClinicalTrials.gov, and WHO-ICTRP. Individual-participant data will be requested from the study authors and harmonized into a uniform format, and aggregated data will also be extracted from the studies. At least two independent reviewers will conduct the study selection, data extraction, risk-of-bias assessment using RoB 2, and applicability evaluation using the RITES tool. The primary outcome will be the number of patients achieving adequate sedation within 30 min after treatment, with secondary outcomes including the need for additional interventions and adverse events, using odds ratios as the effect size. If enough individual-participant data will be collected, we will synthesize them in a network meta-regression model within a Bayesian framework, incorporating study- and participant-level characteristics to explore potential sources of heterogeneity. In cases where individual-participant data are unavailable, potential data availability bias will be explored, and models allowing for the inclusion of studies reporting only aggregated data will be considered. We will assess the confidence in the evidence using the Confidence in Network Meta-Analysis (CINeMA) approach. Discussion: This individual-participant-data network meta-analysis aims to provide a fine-tuned synthesis of the evidence on the comparative effectiveness of pharmacological interventions for severe psychomotor agitation in real-world emergency settings. The findings from this study can greatly be provided clearer evidence-based guidance on the most effective treatments. Systematic review registration: PROSPERO CRD42023402365.