Journal Design Clinical Emerald
African Food Systems Research (Interdisciplinary - incl Agri/Env) | 11 January 2015

A Bayesian Hierarchical Model for Cost-Effectiveness Analysis of Community Health Centres in Ethiopia

A Methodological Case Study
A, b, e, b, e, T, a, d, e, s, s, e, ,, M, e, r, o, n, A, s, s, e, f, a, ,, S, e, l, a, m, a, w, i, t, M, e, n, g, e, s, h, a, ,, T, e, w, o, d, r, o, s, G, e, t, a, c, h, e, w
Bayesian ModellingCost-EffectivenessHealth SystemsSub-Saharan Africa
Model accounts for clustering at regional and district levels in sparse data.
Quantifies probabilistic inferences with posterior distributions via MCMC sampling.
Reveals high variability in cost-effectiveness between districts.
Provides wide credible intervals reflecting substantial system heterogeneity.

Abstract

{ "background": "Evaluating the cost-effectiveness of community health centres is critical for resource allocation in low-resource settings, yet standard frequentist methods often struggle with sparse, hierarchical data and do not fully quantify uncertainty.", "purpose and objectives": "This case study presents a novel Bayesian hierarchical model for cost-effectiveness analysis and demonstrates its application to assess the efficiency of a national network of primary care facilities.", "methodology": "We developed a Bayesian cost-effectiveness model using a hierarchical structure to account for clustering at the regional and district levels. The core effectiveness measure was disability-adjusted life years averted. The model, $\\text{log}(\\text{Cost}{ij}) = \\alpha + \\beta \\text{Effectiveness}{ij} + u{i} + v{ij} + \\epsilon{ij}$, where $ui \\sim N(0, \\sigma^2{region})$ and $v{ij} \\sim N(0, \\sigma^2_{district})$, explicitly modelled uncertainty in all parameters, with posterior distributions estimated via Markov chain Monte Carlo sampling.", "findings": "The model successfully synthesised sparse data, providing robust probabilistic inferences. A key finding was a high degree of variability in cost-effectiveness between districts, with the posterior probability that the incremental cost-effectiveness ratio exceeded a willingness-to-pay threshold being 0.78 for the central region. The 95% credible interval for the national average cost per DALY averted was wide, reflecting substantial heterogeneity.", "conclusion": "The Bayesian hierarchical approach offers a superior methodological framework for economic evaluations in complex health systems, formally incorporating variability and uncertainty that are often overlooked.", "recommendations": "Health economists and policy analysts should adopt Bayesian hierarchical modelling for cost-effectiveness analyses where data are hierarchically structured and sparse. Future research should focus on integrating prior evidence from similar settings to strengthen inferences.", "key words": "Bayesian hierarchical model, cost-effectiveness analysis, health economics, primary health care, uncertainty quantification, sub-Saharan Africa", "