Systematic review and meta-analysis of the prevalence of frailty and pre-frailty amongst older hospital inpatients in low- and middle-income countries.

Background: As populations age, low- and middle-income countries (LMICs) are rapidly adapting hospital services to meet the needs of older populations. This systematic review aimed to establish the prevalence of frailty and pre-frailty amongst older people admitted to hospital in LMICs, and to compare levels with existing estimates for high-income settings.

Methods: Databases Medline, Embase, CINAHL and PsychInfo were searched, and results were manually screened by two researchers. Included studies were cross-sectional or cohort design and reported data from hospital inpatients, aged ≥60 years, in LMICs, using validated methods for identifying frailty. Risk of bias was assessed by two researchers using the Joanna Briggs Institute checklist. Where possible, results were synthesised using meta-analysis.

Results: Twenty-nine studies were included, all reporting data from middle-income countries. Physical frailty tools were the most common, with Fried's phenotype and the FRAIL Scale being the most often used methods of assessment. The pooled estimate of the prevalence of frailty was 39.1% [95% confidence interval (CI) 31.9-46.6%] comprising data from 23 311 older inpatients. For pre-frailty, prevalence was 40.0% (95% CI 25.1-51.4%) from 6954 individuals.

Discussion: Due to the paucity of eligible studies and their geographical distribution, these pooled estimates are only representative of the burden of frailty in select middle-income settings (particularly China). Future research should seek to establish the prevalence of frailty in hospitals in low-income countries, and to assess clinical outcomes by frailty status, in order to develop bespoke clinical tools and to aid the planning of future geriatric services.