Data-based nurse staffing indicators with Bayesian networks explain nurse job satisfaction: a pilot study

Area of Science:

• Nursing Management
• Health Informatics
• Organizational Psychology

Background:

• Hospital information systems generate vast data, yet its application in nursing leadership and management remains limited.
• Effective utilization of healthcare data is crucial for optimizing nursing practices and improving work environments.

Purpose of the Study:

• To investigate the correlation between nursing intensity, work environment intensity, nursing resources, and nurse job satisfaction.
• To explore the predictive power of nurse staffing indicators on job satisfaction within a hospital setting.

Main Methods:

• A pilot study combining retrospective time series patient data (9704 patients) and cross-sectional nurse survey data (110 nurses) from six inpatient units.
• Bayesian networks were employed to model the relationships between identified nurse staffing indicators and unit-level nurse job satisfaction (n=98 nurses).

Main Results:

• Eighteen nurse staffing indicators were identified within the hospital data system.
• Four key indicators—patient acuity, diagnosis-related group volume, skill mix, and nurse turnover—significantly predicted nurse job satisfaction.
• A non-linear relationship between nurse job satisfaction and patient acuity was identified using Bayesian networks.

Conclusions:

• Nurse job satisfaction, measured via surveys, can be effectively modeled using data-derived nurse staffing indicators.
• The Bayesian approach offers a valuable methodology for nurse researchers to investigate the impact of nurse staffing on nursing outcomes.