Using Machine Learning to Predict Resilience Among Nurses in a South African Setting.

Area of Science:

• Nursing
• Psychology
• Data Science

Background:

• Nursing is a high-stress profession, impacting mental health.
• Resilience, a positive stress response, is crucial for mental well-being.
• Understanding resilience factors is vital for supporting nurses' mental health.

Purpose of the Study:

• To identify predictors of high versus low resilience in South African nurses.
• To utilize machine learning for analyzing secondary survey data.
• To assess the effectiveness of a brief resilience screening tool.

Main Methods:

• Secondary data analysis from five surveys (2022-2023) of 1134 South African nurses.
• Random forest analysis applied to demographic variables, years of experience, and a 4-item resilience screen.
• Predictive modeling to identify factors associated with resilience levels.

Main Results:

• The random forest model achieved an 86.41% classification accuracy for resilience levels.
• Demographic variables provided limited additional predictive value.
• A brief 4-item resilience screening demonstrated high predictive accuracy.

Conclusions:

• A brief resilience screening tool is effective in identifying resilience levels in nurses.
• Machine learning models can accurately predict resilience in nursing populations.
• Further research can refine interventions to enhance nurse resilience.