Construction of a multifactorial prediction model for healthcare workers’ work ability

Focusing on the interaction and impact pathways of job burnout and sleep disorder

Datos Bibliográficos

ID22077249
AutoresNi Wang (0000-0003-4961-7477, Shanxi Medical University), Liang Shang (0000-0002-8962-7226, Shanxi Medical University), Ting Zhou (0009-0007-5297-1166, Shanxi Medical University, autor de correspondencia)
Año2026
Volumen14
Páginas1787439-1787439
Fecha de publicación2026-03-17
Open AccessSí
TipoARTICLE
RevistaFrontiers in Public Health (JOURNAL)
Identificadores de la revistaISSN: 2296-2565 • E-ISSN: 2296-2565
EditorialFrontiers Media SA (PUBLISHER • CH)
DOI10.3389/fpubh.2026.1787439
PMID41923757
OpenAlexW7137801990
IdiomaEN
Referencias citadas43

Objective To explore the latent profiles, core associated factors, and complex mechanisms of work ability among healthcare workers in large tertiary hospitals in China. Methods A cross-sectional study was conducted from July to October 2025. A convenience sample of 1,590 healthcare workers from a large tertiary hospital in Shaanxi Province was assessed using the Work Ability Index (WAI), the Maslach Burnout Inventory-General Survey (MBI-GS), and the Pittsburgh Sleep Quality Index (PSQI). Latent profile analysis (LPA) was employed to identify potential categories of work ability. Multivariable logistic regression analysis was performed to determine independently associated factors and to construct a nomogram prediction model. An additive interaction model and structural equation modeling (SEM) were used to analyze the joint effect and the influential pathways of job burnout and sleep disorder. Results LPA identified two distinct categories: “Good Work Ability” (73%) and “Poor Work Ability” (27%). Multivariable regression analysis indicated that job burnout (OR = 3.770, 95% CI: 2.510–5.661) and sleep disorder (OR = 2.890, 95% CI: 2.121–3.939) were the factors most strongly associated with poor work ability. Longer working years (≥21 years) and higher professional titles (intermediate/senior) were also associated with an increased likelihood of poor work ability. In contrast, higher education (master’s degree or above) and regular physical exercise were associated with a decreased likelihood. The predictive nomogram model demonstrated good discriminative ability (AUCs of 0.781 and 0.740 for the training and validation sets, respectively) and clinical utility. Interaction analysis revealed a significant positive additive interaction between job burnout and sleep disorder (RERI = 5.164, AP = 47.453%). SEM supported a model in which job burnout was not only directly and negatively associated with work ability ( β = −0.359, p < 0.01) but also showed an indirect association via impaired sleep quality ( β = −0.281, p < 0.01). Conclusion Among healthcare workers in large tertiary hospitals in China, job burnout and sleep disorder are two core and synergistic factors associated with work ability. The prediction model based on multiple factors can provide a practical tool for the early identification of high-risk individuals. Future occupational health intervention programs need to adopt integrated strategies, targeting both the alleviation of job burnout and the improvement of sleep quality as dual core objectives, and implement precise prevention and control for key populations such as those with long service years and high professional titles to maintain and enhance the work ability of healthcare workers

Burnout · Health care · Logistic regression · Musculoskeletal disorder · Nomogram · Pittsburgh Sleep Quality Index · Regression analysis · Structural equation modeling · Healthcare professionals’ stress and burnout · Sleep and Work-Related Fatigue · Workplace Health and Well-being

Herramientas

Abrir DOIOpen Access