Burnout In Overwhelmed Healthcare Providers: A Tool For Risk Identification
Overview
The aim of the Cross-sectional study was to create a predictive model for identifying high-burnout risk among nurses.
This research utilized an online survey, gathering data with the Chinese Maslach Burnout Inventory-General Survey (CMBI-GS) and self-administered questionnaires. These questionnaires covered demographic, behavioral, health-related, and occupational variables. The study participants were randomly grouped into a validation set and a development set. In the development set, multivariate logistic regression analysis was performed to identify factors linked to high burnout risk, and a nomogram was created based on significant factors. The nomogram’s discrimination, calibration, and clinical utility were assessed in both sets using ROC curve analysis, the Hosmer-Lemeshow test, and decision curve analysis. Data analysis was conducted using Stata 16.0 software.
A total of 2,750 nurses from 23 provinces in mainland China participated, with 1,925 in the development set (70%) and 825 in the validation set (30%). Key factors contributing to high burnout risk included workplace violence, shift work, weekly working hours, depression, stress, self-reported health, and alcohol consumption. A nomogram was developed using these factors. The ROC curve analysis showed an area under the curve (AUC) of 0.808 for the development set and 0.790 for the validation set. The nomogram demonstrated high net benefit in the clinical decision curve for both sets.
This study successfully developed and validated a predictive nomogram to identify high-burnout risk in nurses. The nomogram will help nursing managers identify at-risk nurses and understand related factors, enabling them to implement early and targeted interventions.
The study followed the EQUATOR reporting guidelines, specifically the TRIPOD Checklist for Prediction Model Development and Validation.
Introduction
Burnout is a chronic response to ongoing workplace stress, characterized by emotional exhaustion, depersonalization, and a diminished sense of personal accomplishment. While burnout can affect various professions, healthcare workers, particularly nurses, are highly susceptible due to factors such as workforce shortages, high workloads, and shift work. The resulting burnout significantly impacts nurses’ physical and mental health, thereby compromising the quality of healthcare services and patient safety. Addressing this issue requires further research to identify effective strategies for early detection and intervention of burnout in nurses.
Burnout, classified by the World Health Organization’s ICD-11 as an occupational phenomenon, is especially prevalent among nurses who endure substantial demands and strive to deliver empathetic care. This issue affects nurses globally, with high rates of emotional exhaustion and depersonalization reported, particularly in low- and middle-income countries. For example, in China, approximately 50% of nurses experience moderate to high levels of burnout. The consequences are severe, including increased risks of cardiovascular diseases, musculoskeletal pain, and mental health disorders such as depression, anxiety, substance abuse, and suicide. Additionally, burnout negatively affects patient safety and the quality of care provided, underscoring the necessity for early identification and prevention.
Various risk factors contribute to burnout among healthcare professionals, including high workloads, shift work, job stress, and workplace violence. Demographic factors such as age, gender, and education level, as well as mental health conditions like depression, anxiety, and stress, are also significant contributors. Health-related behaviors such as smoking and drinking further exacerbate burnout. These factors must be considered when assessing the risk of burnout in nurses.
Previous research on burnout has focused on prevalence, associated risk factors, and interventions to alleviate it. However, these studies often fall short in providing a solid foundation for early detection and proactive prevention. Most studies utilize the Maslach Burnout Inventory Scale to assess burnout levels, but this approach does not directly identify influencing factors and can be time-consuming and impractical for clinical use. In countries like China, where the nurse-to-population ratio is lower than in developed countries, a brief, user-friendly predictive tool for burnout is necessary.
A nomogram, a graphical tool combining various factors to predict outcomes, offers a practical solution for assessing burnout. Nomograms have been successfully used to predict health risks and work-related outcomes in nurses, demonstrating their effectiveness in facilitating targeted interventions. This study aims to develop and validate a nomogram-based predictive model for high burnout among nurses, providing valuable evidence to support the development of science-based early interventions to mitigate burnout.
In conclusion, this study highlights the critical need for more effective tools and strategies to detect and prevent burnout in nurses. The development of a nomogram-based predictive model represents a promising approach to address this issue, potentially improving the well-being of nurses and the quality of care they provide.
Method
This study was conducted following the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines. Data collection occurred between January and February 2019 through an online survey using snowball sampling. The survey, hosted on the Chinese platform Questionnaire Star, began with 40 nurses from a university-affiliated hospital. These initial participants were asked to share the survey’s QR code within their WeChat groups, encouraging their colleagues and classmates to participate and further distribute the survey.
The primary measurement tool was the 15-item Chinese Maslach Burnout Inventory-General Survey (CMBI-GS), which evaluates burnout across three dimensions: emotional exhaustion, depersonalization, and reduced personal accomplishment. Each item was rated on a scale from 0 (never) to 6 (daily). The overall burnout score ranged from 0 to 18, with higher scores indicating more severe burnout. Participants were categorized into low, median, and high burnout levels based on their scores. The internal consistency of the CMBI-GS in this sample was strong (Cronbach’s α = 0.86).
Candidate factors for the predictive model were grouped into demographic characteristics, behavioral variables, health-related variables, and occupational variables. Demographic factors included age, gender, education level, and marital status. Behavioral variables encompassed exercise, smoking, drinking, and daily sleep duration. Health-related variables included health status, stress levels, depressive symptoms (assessed with the PHQ-2), and sleep quality (assessed with the Pittsburgh Sleep Quality Index). Occupational variables covered shift work, hospital level, work department, professional title, years of experience, weekly working hours, and workplace violence.
Ethical approval for the study was obtained from the national clinical research center’s ethics committee at the Second Xiangya Hospital of Central South University. Participants provided informed consent and had the right to withdraw at any time before data analysis.
Inclusion Criteria
Eligibility for participation included nurses working in comprehensive hospitals for at least six months, while exclusions were made for those on extended leave or not working in frontline departments. The first page of the questionnaire outlined these criteria and provided informed consent information. Measures to ensure data quality included limiting each IP address to one response and checking for logical inconsistencies.
Statistical Analysis
Statistical analyses were performed using STATA software, dividing the dataset into a 70% development set and a 30% validation set. Univariate analysis identified potential risk factors for high burnout, which were then tested in multivariable logistic regression models. A nomogram was created to visualize the predictive model. The model’s discrimination was evaluated with ROC curve analysis, achieving good performance (AUC > 0.70). Calibration was assessed using the Hosmer–Lemeshow test and a calibration curve, while decision curve analysis evaluated clinical practicability.
Result
After excluding 18 respondents during quality review, data from 2,750 nurses (94.2% female, 5.8% male) across 23 provinces in mainland China were analyzed. The average age of participants was 30.6 ± 7.0 years, and the mean CMBI-GS score was 9.9 ± 3.7 points. The dataset was randomly split into a development set (1,925 participants) and a validation set (825 participants) at a 7:3 ratio. No significant difference in high-burnout prevalence was found between the sets (13.9% in the development set vs. 13.3% in the validation set, p = 0.681). Additionally, CMBI-GS scores and its three dimensions showed no significant differences between the two datasets.
Univariate analyses identified factors such as age, exercise, alcohol consumption, sleep duration, health status, stress levels, depressive symptoms, sleep quality, shift work, work department, professional title, work experience, weekly working hours, and workplace violence experience as associated with high burnout. Multivariate logistic regression revealed independent risk factors for high burnout: absence of drinking habits, normal or poor health status, low or high stress levels, a PHQ-2 score above 4 points (indicating depressive symptoms), shift work, working 40 or more hours per week, and experiencing workplace violence.
Using these seven risk factors, a nomogram for predicting high burnout in nurses was developed. To facilitate its use, these factors were converted into seven questions, and an Excel-based calculator was created for computing risk points.
The model’s performance was assessed using ROC curve analysis, which showed an AUC of 0.808 (95% CI 0.781–0.834) in the development set and 0.790 (95% CI 0.747–0.833) in the validation set, indicating good discrimination. In the development set, a cut-off value of 0.145 (14% risk) with an optimal score of 22.5 points yielded a sensitivity of 0.743 and specificity of 0.725. In the validation set, the sensitivity was 0.700 and specificity was 0.754 with a cut-off value of 0.166. Hosmer-Lemeshow tests produced p-values of 0.697 and 0.640 for the development and validation sets, respectively, indicating good calibration. Calibration curves indicated a slight overestimation of high-burnout probabilities above 0.40 in both sets, but overall calibration was deemed good.
Decision curve analysis evaluated the nomogram’s clinical utility, showing a net benefit across threshold probabilities from 3% to 68% in the development set and from 3% to 56% in the validation set. Predictions were beneficial within these ranges, whereas predictions below 3% or above 56% offered no benefit in diagnosing high burnout. To test real-world applicability, threshold probabilities for clinical decision curves were calculated using validation set data, allowing for accurate predictions in clinical scenarios. High-burnout risk was classified as low (<3%), moderate (3%–56%), and high (>56%).
Conclusion
Given the severe effects of high burnout among nurses on the healthcare system, addressing this issue is critical. To aid nursing managers in identifying and mitigating burnout risk, a predictive model based on a nomogram has been developed. This model has been translated into an accessible Excel tool, demonstrating effective predictive capability and clear interpretation of contributing factors.
The nomogram identifies several predictors of high burnout, including workplace violence, shift work, weekly working hours, depression, stress, self-reported health, and alcohol consumption. These factors have been linked to burnout in previous research. By calculating the cumulative score of these risk factors, the likelihood of high burnout can be assessed. Consistent with earlier studies, nurses who have faced workplace violence are more prone to high burnout. Shift work, especially shifts exceeding 12 hours, occurring more than three times weekly, night shifts, and irregular shifts, also heightens burnout risk by disrupting circadian rhythms and increasing stress. The number of hours worked per week contributes to burnout risk by limiting time for relaxation and preventive measures.
Depression serves as a significant predictor of burnout, aligning with past findings that link early-stage burnout with later depression. This study also notes an inverted-U curve relationship between stress and burnout, suggesting that while moderate stress can be motivating, excessive stress increases burnout risk. This insight highlights the need for nursing managers to monitor and manage stress levels carefully. Interestingly, light alcohol consumption is identified as a protective factor against burnout, a finding that may reflect cultural differences in drinking habits among Chinese nurses.
This study represents a novel approach by employing a nomogram to predict burnout, potentially addressing the limitations of existing evaluation tools. The nomogram classifies nurses into risk categories—low (<3%), moderate (3%-56%), and high (>56%)—enabling targeted interventions. Nurses at moderate or high risk may benefit from assertiveness training, mindfulness, and other support measures. Organizational strategies, such as a zero-tolerance policy for violence and better shift management, are recommended to prevent burnout.
While the nomogram shows promise, the study has limitations. Its cross-sectional design limits causal inference, and the focus on burnout’s total score rather than specific dimensions may affect the selection of tailored interventions. Additionally, the study predates the COVID-19 pandemic, so pandemic-specific factors are not included in the model. However, burnout factors such as workload and stress remain relevant.
In conclusion, this study provides a validated predictive nomogram for high burnout among nurses, integrating key predictors into a practical tool for nursing managers. It allows for early identification and intervention, offering a hierarchical approach to managing burnout risk and enhancing overall nurse well-being. Future research should consider longitudinal studies and updates to incorporate the latest data, including any pandemic-related impacts.
Oncology Related Tools
Other
Latest Research
