Measurement and risk adjustment of the perioperative mortality rate in low and middle income countries: a scoping review.
Oluwatosin Gabriel Afolabi, Damilola Timothy Ishola, Mayowa Emmanuel Oluwajuyigbe, Adedamola Benjamin AdegbamigbeObjective
The perioperative mortality rate (POMR) is one of the core indicators endorsed by the Lancet Commission on Global Surgery, yet its value for comparing surgical safety across settings depends on how it is defined, measured and adjusted for differences in case-mix. This scoping review mapped the literature on the measurement and risk adjustment of POMR in low- and middle-income countries (LMICs), and identified the resulting evidence gaps.
Methods
The review followed the Arksey and O’Malley framework with the Levac refinements and was reported in accordance with the PRISMA extension for Scoping Reviews. Five electronic databases (MEDLINE via PubMed, Embase via Ovid, Scopus, the Cochrane Library and Web of Science) were searched from inception, supplemented by citation searching and grey literature, with no date or language restriction. Eligible sources reported the definition, measurement, reporting or risk adjustment of POMR in adult or paediatric surgical populations in one or more LMICs. Records were deduplicated in Rayyan. Two reviewers independently screened titles/abstracts and full texts, with inter-rater agreement quantified using Cohen’s kappa; disagreements were resolved by discussion with a third reviewer. Data were charted on a standardised form and synthesised narratively.
Results
Of 1,812 records identified across five databases (PubMed 894, Scopus 461, Embase via Ovid 365, Web of Science 73, Cochrane Library 19), 978 duplicates were removed, leaving 834 unique records. Title and abstract screening excluded 798 records (κ = 0.786, substantial agreement), leaving 36 reports for full-text assessment. Twenty-nine sources met the inclusion criteria (κ = 0.906, almost perfect agreement); seven were excluded, most commonly for high-income-only data without disaggregated LMIC results or for representing an overlapping cohort already captured by a more complete source. Four recurring themes emerged: heterogeneous numerator, denominator, and follow-up definitions that limited comparability; predominance of crude, in-hospital estimates with sparse and unevenly reported case-mix adjustment; feasibility of parsimonious, locally derived risk-adjustment models built on routinely available variables; and persistently uncommon standardised national reporting of POMR, even where indicator collection was otherwise feasible.