A meta-analysis of controlled studies comparing the association between clozapine and other antipsychotic medications and the development of neutropenia.

BACKGROUND: In most countries, clozapine can only be prescribed with regular monitoring of white blood cell counts because of concerns that clozapine has a stronger association with neutropenia than other antipsychotics. However, this has not been previously demonstrated conclusively with meta-analysis of controlled studies. METHODS: The aim of this study was to assess the strength of the association between clozapine and neutropenia when compared to other antipsychotic medications by a meta-analysis of controlled studies. An electronic search of Medline (1948-2018), PsycINFO (1967-2018) and Embase (1947-2018) using search terms (clozapine OR clopine OR clozaril OR zaponex) AND (neutropenia OR agranulocytosis) was undertaken. Random-effects meta-analysis using Mantel-Haenszel risk ratio was used to assess the strength of the effect size. RESULTS: We located 20 studies that reported rates of neutropenia associated with clozapine and other antipsychotic medications. The risk ratio was not significantly increased in clozapine-exposed groups compared to exposure to other antipsychotic medications (Mantel-Haenszel risk ratio = 1.45, 95% confidence interval = [0.87, 2.42]). This also applied to severe neutropenia (absolute neutrophil count < 500 per µL) when compared to other antipsychotics (Mantel-Haenszel risk ratio = 1.65, 95% confidence interval = [0.58, 4.71]). The relative risk of neutropenia associated with clozapine exposure was not significantly associated with any individual antipsychotic medication. CONCLUSION: Data from controlled trials do not support the belief that clozapine has a stronger association with neutropenia than other antipsychotic medications. This implies that either all antipsychotic drugs should be subjected to haematological monitoring or monitoring isolated to clozapine is not justified.

Meta-analysis of the strength of exploratory suicide prediction models; from clinicians to computers.

BACKGROUND: Suicide prediction models have been formulated in a variety of ways and are heterogeneous in the strength of their predictions. Machine learning has been a proposed as a way of improving suicide predictions by incorporating more suicide risk factors. AIMS: To determine whether machine learning and the number of suicide risk factors included in suicide prediction models are associated with the strength of the resulting predictions. METHOD: Random-effect meta-analysis of exploratory suicide prediction models constructed by combining two or more suicide risk factors or using clinical judgement (Prospero Registration CRD42017059665). Studies were located by searching for papers indexed in PubMed before 15 August 2020 with the term suicid* in the title. RESULTS: In total, 86 papers reported 102 suicide prediction models and included 20 210 411 people and 106 902 suicides. The pooled odds ratio was 7.7 (95% CI 6.7-8.8) with high between-study heterogeneity (I2 = 99.5). Machine learning was associated with a non-significantly higher odds ratio of 11.6 (95% CI 6.0-22.3) and clinical judgement with a non-significantly lower odds ratio of 4.7 (95% CI 2.1-10.9). Models including a larger number of suicide risk factors had a higher odds ratio when machine-learning studies were included (P = 0.02). Among non-machine-learning studies, suicide prediction models including fewer risk factors performed just as well as those including more risk factors. CONCLUSIONS: Machine learning might have the potential to improve the performance of suicide prediction models by increasing the number of included suicide risk factors but its superiority over other methods is unproven.