Methods

This review adhered to principles outlined in guidance published by PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analysis) and CHARMS (CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies). The protocol for this review has been published by PROSPERO and is available online.
A literature search was conducted during May 2018 of the following databases: Medline, Embase, CINAHL and the Cochrane Library. To inform the full search strategy a limited search of Medline was first conducted followed by an extensive search of the literature of the aforementioned databases. A copy of the search strategy for Medline and Embase is available in Table S1. The main search terms were ‘predict$’, ‘risk score’ and ‘postpartum haemorrhage’ with the appropriate synonyms adopted.
Inclusion criteria and exclusion criteria for this review are outlined in Table 1 . Titles and abstracts were independently screened by two reviewers (CN and SN) with any disagreements resolved by a third reviewer (MB).
Data extraction and quality assessment (at study level) were conducted independently in accordance with the CHARMS checklist (Table S2 ) to allow identification of potential bias in primary studies and identify limitations to applicability of the results. Items extracted were as follows: source of data; participants; outcome to be predicted; candidate predictors (or index tests); sample size; missing data; model development; model performance; model evaluation; results and interpretation (including whether authors deemed their model fit for purpose or nature of further research required before using). The findings were tabulated and a narrative synthesis performed. The findings address the baseline characteristics of the studies, the type of models included, evaluation of the models and the applicability of the models to clinical practice.