Methods for modelling composite indices of access to healthcare facilities: a systematic literature review

BACKGROUND: 

Access to quality healthcare services is key to achieving Universal Health Coverage (UHC). The multidimensional nature of access (availability, accessibility, accommodation, affordability and acceptability) makes it challenging to quantify the level of access. Current approaches focus predominantly on single dimensions, limiting the comprehensive monitoring and evaluation of access to healthcare facilities. Here, we conduct a systematic literature review on the methodological approaches and data used to construct multidimensional composite indices of healthcare facility access, globally.

METHODS: 

We undertook a literature search in eight databases including EBSCOhost (CINAHL), Google Scholar, Ovid (Embase and MEDLINE), PubMed, Scopus, Web of Science and Web of Science (MEDLINE) adhering to the Preferred Reporting Items for Systematic Reviews (PRISMA) guidelines. Studies that incorporated multiple dimensions of access to healthcare facilities to construct a composite index were considered and quality assessment performed. Methodological approaches to measuring access and their supporting conceptual frameworks were synthesised using descriptive summaries and thematic analysis.

RESULTS: 

Out of 4,291 articles retrieved,19 met inclusion criteria with an average quality score of 19.6 out of 26. Most of the studies (68%) were conducted in 2021-2024, mainly in India (53%) or USA (16%); none in Africa. The composite indices of access combined two (32%), three (42%), four (5%) or all five dimensions (21%), with affordability (84%) being the most frequent dimension. There was significant heterogeneity on the definition, data (survey-based or retrospective) and representation of indicators. There were four weighting techniques ranging from simple (equal weighting) to complex approaches (Principal Component Analysis and Analytical Hierarchy Process). Studies used four different approaches to combine indicators; arithmetic mean (ten studies), summation (six studies), Adjusted Mazziotta-Pareto Index (two studies) and geometric mean (one study). Only 63% validated their output.

CONCLUSIONS: 

There is diversity in the approaches used for multidimensional assessment of access to healthcare facilities. To ensure robust, context-specific and more comprehensive composite indices, the use of clearly defined frameworks, dimension weights that reflect context-specific access barriers and penalised aggregation methods will be required.