Are hospital management practices associated with enhanced quality of care for small and sick newborns? A nationwide cross-sectional study using linked inpatient admission records in Malawi

TL;DR

This study in Malawi found no clear link between hospital management practices and neonatal mortality, suggesting more research is needed to improve care for small and sick newborns.

Contribution

The study provides novel national evidence on hospital management practices and neonatal outcomes in a low-income country.

Findings

No significant association was found between hospital management scores and neonatal mortality.

Limited evidence showed a positive link between management practices and clinical care quality.

The average hospital management score was 3.35 out of 5.

Abstract

Improved quality of care is fundamental for reducing patient mortality and building sustainable health systems. Currently, there is a lack of research on the role of hospital management in improving the quality of care and health outcomes, particularly in low-income settings.

We examined associations between hospital management practices and neonatal quality of care in Malawi. We adapted the World Management Survey tool to measure 28 management practices across five domains – delivery of clinical care in the neonatal unit, human resource management, target setting, finances, and governance. In April 2022, we administered the tool to five clinical and administrative managers in each of the 36 central and first-level referral hospitals (n = 180 interviews). Further, we calculated a hospital-level management score (1 – poor, 5 – excellent) and linked these data to records of 20 831 neonatal admissions (February–July 2022). Our primary outcome was in-hospital neonatal mortality, and secondary outcomes included 14 clinical quality indicators. We examined associations between hospital-level management scores and individual-level patient outcomes using a multilevel mixed-effects Poisson regression.

The mean hospital-level management score across the 36 hospitals was 3.35 (standard deviation = 0.4). Among 20 831 neonatal admissions, 2590 (12.4%; 95% confidence interval (CI) = 11.9–12.8) died, representing a mortality rate of 27.2 deaths per 1000 person-days. We found no relationship between the management score and in-hospital neonatal mortality (adjusted incidence rate ratio per unit increase in the score = 1.08; 95% CI = 0.81–1.44). Five management domains were not associated with mortality, and we found limited evidence that management practice scores were positively associated with quality of clinical care.

This study presents novel, national evidence on the association between hospital management practices and neonatal mortality in a low-income country, with complementary data on quality of clinical care. We found no evidence that hospital management practices were associated with neonatal mortality and limited evidence of an association with the quality of clinical care. Rigorous impact evaluations of targeted management interventions, with embedded process evaluations, could address potential confounders and help understand how and under what circumstances management improvements could translate into better quality of care.

Figures