Developing a predictive model for nurse, patient, and practice environment outcomes through data analytics

Abstract

Accurate forecasting of nurse, patient, and practice environment outcomes is essential to understand trends and inform policies in South African hospitals. Although the relationships between these outcomes are known, accurate and directive data analytics models that use historical data and trends to predict outcomes and inform policies are limited, resulting in a gap in the predictive capability required for data-driven decision-making in the South African context. This research addresses this gap by applying data analytics methods, including statistical analysis and machine learning methods, to develop predictive models based on nurse survey responses. The RN4CAST (2009-2010) and SANOPSys (2021-2022) nurse survey datasets were utilised to capture nurse, patient, and practice environment outcomes under both stable and crisis-driven conditions, including the COVID-19 pandemic. The research aims to develop predictive models for accurately forecasting nurse, patient, and practice environment outcomes based on survey data. A systematised literature review investigated and compared existing statistical analysis and machine learning methods used to forecast nurse, patient, and practice environment outcomes, while descriptive and diagnostic analytics identified key trends and relationships within the datasets that influence nurse, patient, and practice environment outcomes. Predictive models were developed and validated for performance and accuracy, providing insights into key factors influencing nurse, patient, and practice environment outcomes. Prescriptive analytics translated predictive models results into actionable insights, informing policies and decisions to improve nurse, patient, and practice environment outcomes. The research demonstrates that utilising statistical analysis and machine learning methods within a data analytics framework enables accurate forecasts, directive modelling, and actionable insights, offering a replicable approach to support data-driven decision-making and improve nurse, patient, and practice environment outcomes in South African hospitals.