Dataset Information

Using supervised learning to select audit targets in performance-based financing in health: An example from Zambia.

ABSTRACT: Independent verification is a critical component of performance-based financing (PBF) in health care, in which facilities are offered incentives to increase the volume of specific services but the same incentives may lead them to over-report. We examine alternative strategies for targeted sampling of health clinics for independent verification. Specifically, we empirically compare several methods of random sampling and predictive modeling on data from a Zambian PBF pilot that contains reported and verified performance for quantity indicators of 140 clinics. Our results indicate that machine learning methods, particularly Random Forest, outperform other approaches and can increase the cost-effectiveness of verification activities.

SUBMITTER: Grover D

PROVIDER: S-EPMC6350980 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Using supervised learning to select audit targets in performance-based financing in health: An example from Zambia.

Grover Dhruv D Bauhoff Sebastian S Friedman Jed J

PloS one 20190129 1

Independent verification is a critical component of performance-based financing (PBF) in health care, in which facilities are offered incentives to increase the volume of specific services but the same incentives may lead them to over-report. We examine alternative strategies for targeted sampling of health clinics for independent verification. Specifically, we empirically compare several methods of random sampling and predictive modeling on data from a Zambian PBF pilot that contains reported a ...[more]

PMID: 30695057

Similar Datasets

Project description:BackgroundHealth systems' weakness remains one of the primary obstacles towards achieving universal access to quality healthcare in low-income settings. Performance-based financing (PBF) programs have been increasingly used to increase access to quality care in LMICs. However, evidence on the impacts of these programs remains fragmented and inconclusive. We analyze the health system impacts of the PBF program in the Democratic Republic of the Congo (DRC), one of the largest such programs introduced in LMICs to date.MethodsWe used a health systems perspective to analyze the benefits of PBF relative to unconditional financing of health facilities. Fifty-eight health zones in six provinces were randomly assigned to either a control group (28 zones) in which facilities received unconditional transfers or to a PBF program (30 zones) that started at the end of 2016. Follow-up data collection took place in 2021-2022 and included health facility assessments, health worker interviews, direct observations of consultations and deliveries, patient exit interviews, and household surveys. Using multivariate regression models, we estimated the impact of the program on 55 outcomes in seven health system domains: structural quality, technical process quality, non-technical process quality, service fees, facility management, providers' satisfaction, and service coverage. We used random-effects meta-analysis to generate pooled average estimates within each domain.ResultsThe PBF program improved the structural quality of health facilities by 4 percentage points (ppts) (95% CI 0.01-0.08), technical process quality by 5 ppts (0.03-0.07), and non-technical process by 2 ppts (0-0.04). PBF also increased coverage of priority health services by 3 ppts (0.02-0.04). Improvements were also observed for facility management (9 ppts, 0.04-0.15), service fee policies, and users' satisfaction with service affordability (14 ppts, 0.07-0.20). Service fees and health workers' satisfaction were not affected by the program.ConclusionsThe results suggest that well-designed PBF programs can lead to improvements in most health systems domains relative to comparable unconditional financing. However, the large persisting gaps suggest that additional changes, such as allocating more resources to the health system and reforming the human resources for health management, will be necessary in DRC to achieve the ambitious global universal health coverage and mortality goals.Trial registrationAmerican Economics Association Trial registry AEARCTR-0002880.

Dataset Information

Using supervised learning to select audit targets in performance-based financing in health: An example from Zambia.

Publications

Using supervised learning to select audit targets in performance-based financing in health: An example from Zambia.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets