Factors influencing antenatal corticosteroid administration for management of women imminent preterm birth: A cross-sectional study in two Kenyan county referral hospitals

Abstract

Antenatal corticosteroids (ACS) are an established, evidence-based intervention that significantly reduces morbidity and mortality associated with prematurity. However, despite strong global recommendations, including those from the World Health Organization (WHO), their utilization remains suboptimal in low- and middle-income countries (LMICs), where the burden of neonatal deaths is highest. Understanding the determinants of ACS use is therefore critical for improving implementation. This study was primarily guided by the Donabedian Model, which examines healthcare quality through structure, process, and outcomes. A cross-sectional design was employed between June and August 2025 at Nyahururu County Referral Hospital (Laikipia County) and Wamalwa Kijana County Referral Hospital (Trans Nzoia County). The study included 160 healthcare providers (80 per facility) involved in maternal and newborn care. Data were collected using structured questionnaires covering demographics, knowledge (20 items), practices (15 items), and perceived facility-level factors (10 items). Additionally, a data extraction tool was used to get information from the records of the 160 preterm birth cases that were conducted. A data extraction form was used to do record review, which was done to obtain information regarding preterm birth (n=160). Quantitative data were analyzed using chi-square tests, independent t-tests, and bivariate and multivariable logistic regression, while qualitative insights were obtained through thematic analysis of eight key informant interviews. Thematic analysis of key informant interviews (n=8) provided complementary qualitative insights. The mean knowledge score was 55.2% (SD 20.7), with only 49.4% of providers correctly identifying ACS indications and 46.9% knowing WHO-recommended gestational age. ACS was administered to only 30.6% (49/160) of eligible women. The primary reason for non-administration was failure to prescribe (82.5% of missed cases). Facility-level factors significantly associated with ACS administration included availability of visible standard operating procedures (adjusted OR [aOR]: 2.34, 95% CI: 1.28-4.27, p=0.006), access to WHO guidelines (aOR: 1.89, 95% CI: 1.04-3.43, p=0.037), and prior training on evidence-based practices including ACS (aOR: 2.67, 95% CI: 1.41-5.06, p=0.002). Provider knowledge score was associated in bivariate analysis (p=0.018) but not after adjusting for facility factors (aOR: 1.02, p=0.412). Qualitative findings revealed four major themes: knowledge and practice gaps, challenges in identifying eligible clients, lack of accessible guidelines and protocols, and inconsistent administration practices. Facility-level system factors, particularly visible protocols, guideline accessibility, and provider training, are stronger predictors of ACS administration than individual provider knowledge alone. Interventions to improve ACS utilization in resource-limited settings should prioritize system-level changes alongside educational programs. Policy recommendation: Hospital administrators and county health management teams should provide necessary facility-level actors, including displaying visible standard operating procedures (SOPs), ensuring accessible WHO guidelines, and supporting the Kenyan Ministry of Health (MOH) to implement regular evidence-based practice training, integrate ACS administration indicators, and develop a national ACS clinical decision support tool.