Hospital Staff Assessment of IT Infrastructure Effectiveness in Facilitating Predictive Analysis in Ghanaian Hospitals

Abstract

This study assessed the Effectiveness of IT Infrastructure in Facilitating Predictive Analysis in Ghanaian Hospitals. Employing a descriptive survey design, the research targeted three hospitals within the Catholic Diocese of Goaso: St. John of God Hospital, St. Elizabeth Hospital, and St. Edward Hospital. A purposive sampling technique was used to select 90 participants, comprising clinical and administrative staff, including doctors, nurses, IT personnel, and health information officers. Data were collected using a structured web-based questionnaire and analyzed through logistic regression and multicollinearity testing using standard statistical software. Using a descriptive survey and logistic regression analysis, it found that gender significantly influenced adoption, with males more likely to use predictive tools. Other demographics like age, job title, and experience were not significant. While IT infrastructure aided adoption, it was not a sole predictor. The logistic regression model showed strong robustness, and though XGBoost had higher accuracy, it suffered from poor recall and overfitting. Logistic regression was recommended for its balanced performance and interpretability in healthcare, while XGBoost could be optimized for better generalization.

Keywords: Demographic Factors, Predictive Analytics, Healthcare Professionals, Ghana, Technology Adoption, Digital Health, Hospital settings, UTAUT, TAM.

DOI: 10.7176/JHMN/119-09

Publication date: September 30th 2025