Mediating effect of digital health literacy between preventive health service utilization and mental health among hospital personnel

Abstract

BACKGROUND

Preventive health service utilization (PHSU) is positively associated with the mental health of hospital employees. Digital health literacy (DHL), a core competence for utilizing digital health services, has been shown to influence health behaviors and outcomes. However, the mediating role of DHL in the relationship between PHSU and mental health among hospital personnel remains unclear.

AIM

To explore relationships among DHL, PHSU, and mental health in hospital employees, and to examine the mediating role of DHL.

METHODS

PHSU-related data, including completion rates of annual physical examinations, vaccination records, and participation in specific disease screenings between 2021 and 2023 were extracted from electronic health records. In the first quarter of 2024, standardized questionnaires were used to measure DHL (DHL scale), mental health status [12-item General Health Questionnaire (GHQ-12)], and occupational burnout [Maslach Burnout Inventory (MBI)]. The mediating effect of DHL was tested using correlation and hierarchical regression analyses and PROCESS macro model 4.

RESULTS

The mean (± SD) age of the cohort (n = 688; 69.2% female) was 35.6 ± 8.3 years. The mean comprehensive PHSU index was 6.8 ± 1.9, with an annual physical examination completion rate of 82.8%, and a recommended vaccination rate of 76.5%. Mean scores were distributed as follows: Total DHL, 28.8 ± 5.5; total GHQ-12, 14.0 ± 4.3; and MBI subscale scores, emotional exhaustion (22.5 ± 7.0), depersonalization (6.4 ± 3.7), and reduced personal accomplishment (32.6 ± 6.8). There were significant positive correlations among PHSU, DHL, and mental health indicators (all P < 0.01). Regression analysis revealed that PHSU predicted better mental health status (β = -0.26; P < 0.001) and higher DHL (β = 0.33; P < 0.001). DHL played a partial mediating role between PHSU and mental health [indirect effect = -0.10 (95% confidence interval: -0.15 to -0.06), while the direct effect remained significant].

CONCLUSION

PHSU among hospital personnel is directly associated with better mental health status, indirectly strengthened by improved DHL. Enhancing psychological well-being of healthcare workers encourages regular participation in preventive health services.

Key Words: Digital health literacy; Preventive health service utilization; Mental health; Hospital staff; Mediating effect; Retrospective cohort study

Core Tip: This retrospective cohort study investigated the relationships among preventive health service utilization (PHSU), digital health literacy (DHL), and mental health in 688 hospital employees. Data analysis, including correlation and regression methods, revealed a significant association between higher levels of DHL and better mental health. Crucially, DHL was identified as a key mediating variable in this relationship. The findings indicate that while PHSU directly promotes employee mental health, its positive effects are also indirectly enhanced through improved DHL.

INTRODUCTION

As the core of the healthcare system, hospital staff face prolonged work-related stress, operate in complex environments, and frequently interact with patients. Their mental health not only affects their own well-being but also directly impacts the quality of medical care and patient safety[1]. Existing research indicates that preventive health services are a vital approach to maintaining the health of occupational populations. For example, regular health check-ups, vaccinations, and disease screenings can help professionals detect potential health risks early, reduce disease burden, and significantly alleviate psychological stress caused by health concerns[2].

With the increasing integration of digital technology into the medical field, services such as online health consultations, electronic health record inquiries, and digital preventive health guidance have become indispensable components of the medical system[3]. Digital health literacy (DHL), which refers to an individual's ability to acquire, understand, evaluate, and apply digital health information and services, has been proven to be significantly related to the formation of healthy behaviors and the improvement of health outcomes[4]. Due to their professional characteristics, hospital staff have more opportunities to come into contact with and use digital health tools. However, most existing studies focus on the impact of DHL on the health behaviors of the general population, with few exploring the mechanism of preventive health service utilization (PHSU) and mental health among hospital employees.

There is a hypothesis suggesting that during the process of providing preventive health services, healthcare providers may gradually enhance their DHL by frequently interacting with digital health management tools (such as electronic health report interpretation, online vaccine appointments, and digital screening reminders). In turn, higher DHL can help them obtain health information more efficiently, make better health decisions, reduce psychological distress, and alleviate job burnout[5]. Therefore, DHL may play a mediating role in the association between PHSU and mental health.

Based on this, this retrospective cohort study aims to systematically explore the relationships among DHL, PHSU, and mental health in the healthcare provider group, and to test the mediating effect of DHL in this association, providing theoretical basis and practical references for formulating targeted mental health intervention strategies.

MATERIALS AND METHODS

Study participants

In this study, a convenience sampling method was used to select the research cohort, including all the staff who were continuously employed at a certain tertiary hospital since January 1, 2021.

Inclusion criteria: (1) Those who joined the company before January 1, 2021 and are still employed as of March 2024; (2) Having a complete electronic health record, including the relevant records of PHSU from 2021 to 2023; and (3) Voluntarily participate in this research and complete the questionnaire survey.

Exclusion criteria: (1) Resigned or transferred to another hospital during the study period; (2) Missing key information in electronic health records; and (3) Diagnosis of severe mental illness or cognitive impairment, confirmed by self-report or medical records.

A total of 688 participants' data were included in this study, among whom 212 were male (30.8%) and 476 were female (69.2%); the average age was 35.6 ± 8.3 years (range: 22-59 years).

The educational attainment distribution of the study cohort is as follows: 169 had a junior college degree (24.5%), 416 had a bachelor's degree (60.5%), and 103 had a master's degree (15.0%).

In terms of occupation, 313 were clinical nurses (45.5%), 206 were physicians (30.0%), and 169 were medical technicians and administrative staff (24.5%).

Measures

PHSU: This study extracted relevant data from the hospital's electronic health record system for the years 2021-2023, and constructed a comprehensive PHSU index. The total score of this index ranges from 0 to 10. The specific scoring scheme was as follows: Annual physical examination completion (0-3 points; 3 points for completion in all 3 years, 2 points for completion in 2 years, 1 point for completion in 1 year, and 0 points for no completion); recommended vaccination coverage (0-3 points; 3 points for full completion of all occupationally recommended vaccinations, including influenza vaccine, hepatitis B vaccine, 2 points for 70%-99% completion, 1 point for 30%-69% completion, and 0 points for < 30% completion); participation in specific disease screenings (0-4 points; total of 5 recommended screening items, our hospital's employee physical examinations conduct different cancer screenings based on gender. Four points were assigned for the completion of ≥ 3 items, 3 points for 2 items, 2 points for 1 item, and 0 points for no participation. This index was validated in a pilot study with a content validity index of 0.89, indicating satisfactory content validity.

DHL: The assessment was conducted using a Chinese version of the DHL scale revised based on the European Health Literacy Framework (i.e., the eHEALS scale). This scale consists of 8 items across 4 dimensions: Information acquisition (2 items), information evaluation (2 items), information application (2 items), and digital communication (2 items). The scale uses a 5-point rating system (1 = strongly disagree, 5 = strongly agree), with a total score range of 8-40 points. Higher scores indicate a higher level of DHL. In this study, the Cronbach's α coefficient of this scale was 0.87, indicating good internal consistency reliability.

Mental health status: The 12-item General Health Questionnaire (GHQ-12) was used to assess the general level of psychological distress among the research subjects. A four-point scoring system was used (0 = not at all, 3 = much more than usual), with the total score ranging from 0 to 36. A higher score represents more severe psychological distress, and the cut-off score was 12 (a score ≥ 12 suggests significant psychological distress). The Cronbach's α coefficient for this scale in this study was 0.82.

Job burnout: The Maslach Burnout Inventory (MBI) was used for assessment. This inventory consists of 3 dimensions with a total of 22 items: Emotional exhaustion (9 items), depersonalization (5 items), and reduced personal accomplishment (8 items). It uses a 7-point rating scale (0 = never, 6 = every day); the higher the scores for the emotional exhaustion and depersonalization dimensions, and the lower the score for the personal accomplishment reduction dimension, the more severe the degree of job burnout. In this study, the overall Cronbach's α coefficient of this inventory was 0.85, and the Cronbach's α coefficients for each dimension were 0.83, 0.78, and 0.81 respectively.

Demographic information: The age, gender, educational background, and occupation of the research subjects were collected through questionnaires, and these demographic data were used as control variables in the regression analysis.

Data collection

The data collection for this study was carried out in two phases. The first phase (January 2024 to February 2024): Data related to the PHSU of 688 research subjects from 2021 to 2023 were extracted from the electronic health records by the hospital staff health office, and the comprehensive PHSU index was calculated. The second phase (March 2024): Standardized questionnaires were distributed through an online questionnaire platform (Questionnaire Star), including the DHL scale, the GHQ-12, the MBI, and demographic variables. Before distributing the questionnaires, the research purpose, content and confidentiality principles were explained in detail to the research subjects. After obtaining informed consent, the participants voluntarily filled out the questionnaires. A total of 800 questionnaires were distributed, and 688 valid questionnaires were retrieved, with an effective recovery rate of 86.0%. The questionnaire response rate was achieved through a series of strict quality control measures: Firstly, all the questions on the online questionnaire platform were set as mandatory fields to prevent participants from submitting incomplete questionnaires; Secondly, the research team conducted one-on-one follow-up verifications with each participant after the questionnaire distribution to ensure that all eligible participants had completed and submitted the questionnaires, and carefully checked the returned questionnaires to eliminate invalid responses, thereby achieving an 86.0% effective response rate.

Statistical analysis

Data analysis was conducted using SPSS 26.0 software (IBM Corporation, Armonk, New York, United States), and chart drawing was performed using Prism 9.0 software (GraphPad Inc., San Diego, California, United States). The specific analytical methods included the following.

Descriptive statistics: Measurement data are expressed as mean ± SD and enumeration data are expressed as n (%). Pearson’s correlation analysis was performed to explore correlations among PHSU, DHL, and mental health indicators.

Hierarchical regression analysis: Hierarchical regression was performed to examine the predictive effects of PHSU on DHL and mental health. The control variables (age, sex, educational background, and occupation) were entered in the first step, and PHSU was entered in the second step.

Mediation effect analysis: The PROCESS macro (model 4) developed by Hayes was adopted to test the mediating effect of DHL using the bootstrap method (5000 re-samplings). The mediating effect was considered to be statistically significant if the 95% confidence interval (95%CI) did not include zero. Differences with P < 0.05 were considered to be statistically significant.

RESULTS

Descriptive statistics

Table 1 presents the descriptive statistics for all main variables (n = 688). The mean comprehensive PHSU index was 6.8 ± 1.9. The annual physical examination completion rate was 82.8%, and the recommended vaccination rate was 76.5%. The mean total DHL score was 28.8 ± 5.5, and the mean total GHQ-12 score was 14.0 ± 4.3; 385 participants (56.0%; scored ≥ 12), indicating significant psychological distress. Mean MBI subscale scores were 22.5 ± 7.0 (emotional exhaustion), 6.4 ± 3.7 (depersonalization), and 32.6 ± 6.8 (reduced personal accomplishment).

Table 1 Descriptive statistics of main study variables (n = 688).

Variable	mean ± SD	Range
Comprehensive PHSU index (score)	6.8 ± 1.9	0-10
Annual physical examination completion rate (%)	82.8	-
Recommended vaccination rate (%)	76.5	-
Total DHL score	28.8 ± 5.5	16-40
Total GHQ-12 score	14.0 ± 4.3	3-28
MBI emotional exhaustion score	22.5 ± 7.0	5-40
MBI depersonalization score	6.4 ± 3.7	0-17
MBI reduced personal accomplishment score	32.6 ± 6.8	15-49

DHL: Digital health literacy; PHSU: Preventive health service utilization; GHQ-12: 12-item General Health Questionnaire; MBI: Maslach Burnout Inventory.

Correlation analysis

Pearson correlations among primary variables are shown in Table 2. All reported correlations were significant at P < 0.001. The comprehensive PHSU index correlated positively with total DHL score (r = 0.35) and negatively with total GHQ-12 score (r = -0.29), emotional exhaustion (r = -0.27), and depersonalization (r = -0.22), and positively with reduced personal accomplishment (r = 0.24). Total DHL score correlated negatively with total GHQ-12 (r = -0.32), emotional exhaustion (r = -0.30), and depersonalization (r = -0.25), and positively with reduced personal accomplishment (r = 0.26).

Table 2 Results of Pearson correlation analysis of main study variables (r value, n = 688).

Variable	Comprehensive PHSU index	Total DHL score	Total GHQ-12 score	MBI emotional exhaustion	MBI depersonalization	MBI reduced personal accomplishment
Comprehensive PHSU Index	1	0.35a	-0.29a	-0.27a	-0.22a	0.24a
Total DHL score	0.35a	1	-0.32a	-0.30a	-0.25a	0.26a
Total GHQ-12 score	-0.29a	-0.32a	1	0.68a	0.53a	-0.49a
MBI emotional exhaustion	-0.27a	-0.30a	0.68a	1	0.50a	-0.42a
MBI depersonalization	-0.22a	-0.25a	0.53a	0.50a	1	-0.36a
MBI reduced Personal accomplishment	0.24a	0.26a	-0.49a	-0.42a	-0.36a	1

^aP < 0.001.

DHL: Digital health literacy; PHSU: Preventive health service utilization; GHQ-12: 12-item General Health Questionnaire; MBI: Maslach Burnout Inventory.

Regression and mediation analyses

Hierarchical regression results are detailed in Tables 3 and 4. After controlling for demographics, the comprehensive PHSU index significantly predicted higher DHL (β = 0.33, P < 0.001, ΔR² = 0.10) and lower psychological distress (GHQ-12: β = -0.26, P < 0.001, ΔR² = 0.08).

Table 3 Results of Hierarchical Regression Analysis of preventive health service utilization on digital health literacy.

Variable	β	t value	P value	R2	ΔR2
Step 1 (demographic variables)	-	-	-	0.07	-
Age	0.09	2.35	0.02	-	-
Gender (female = 1)	0.06	1.58	0.11	-	-
Educational background (bachelor’s degree and above = 1)	0.16	3.47	0.001	-	-
Occupation (nurse = 1)	-0.06	-1.52	0.13	-	-
Step 2 (including PHSU)	-	-	-	0.17	0.10a
Age	0.08	2.12	0.03	-	-
Gender (female = 1)	0.05	1.32	0.19	-	-
Educational background (bachelor’s degree and above = 1)	0.14	3.01	0.003	-	-
Occupation (nurse = 1)	-0.05	-1.28	0.20	-	-
Comprehensive PHSU index	0.33	7.95	< 0.001	-	-

^aP < 0.001.

PHSU: Preventive health service utilization.

Table 4 Results of hierarchical regression analysis of preventive health service utilization on total 12-item General Health Questionnaire score.

Variable	β	t value	P value	R2	ΔR2
Step 1 (demographic variables)	-	-	-	0.06	-
Age	-0.10	-2.63	0.009	-	-
Gender (female = 1)	0.12	3.18	0.002	-	-
Educational background (bachelor’s degree and above = 1)	-0.13	-2.87	0.004	-	-
Occupation (nurse = 1)	0.11	2.95	0.003	-	-
Step 2 (including PHSU)	-	-	-	0.14	0.08a
Age	-0.09	-2.35	0.02	-	-
Gender (female = 1)	0.11	2.92	0.004	-	-
Educational background (bachelor’s degree and above = 1)	-0.11	-2.48	0.013	-	-
Occupation (nurse = 1)	0.10	2.68	0.007	-	-
Comprehensive PHSU index	-0.26	-6.38	< 0.001	-	-

^aP < 0.001.

PHSU: Preventive health service utilization.

Mediation analysis (Table 5 and Figure 1) revealed a significant total effect of PHSU on GHQ-12 (β = -0.26, 95%CI: -0.35, -0.17, P < 0.001), a significant direct effect (β = -0.16, 95%CI: -0.25 to -0.07, P = 0.001), and a significant indirect effect through DHL (β = -0.10, 95%CI: -0.15 to -0.06, P < 0.001), which accounted for 38.5% of the total effect. These results indicate that DHL partially mediates the relationship between PHSU and psychological distress.

Figure 1 Pathway diagram of the mediating effect of digital health literacy between preventive health service utilization and psychological distress among hospital staff. Path coefficients are shown along the arrows: Preventive health service utilization (PHSU) to digital health literacy (DHL): β = 0.33, P < 0.001; DHL to 12-item General Health Questionnaire (GHQ-12): β = -0.30, P < 0.001. Indirect effect through DHL: β = -0.10, 95% confidence interval (95%CI): -0.15 to -0.06 (accounting for 38.5% of the total effect). ^aDirect effect of PHSU on GHQ-12 after controlling for DHL: β = -0.16, 95%CI: -0.25 to -0.07; ^bTotal effect of PHSU on GHQ12: β = -0.26, 95%CI: -0.35 to -0.17. DHL: Digital health literacy; PHSU: Preventive health service utilization; GHQ-12: 12-item General Health Questionnaire.

Table 5 Results of the mediating effect test of digital health literacy between preventive health service utilization and total 12-item General Health Questionnaire Score (Bootstrap = 5000).

Type of effect	β	SE	95%CI	P value
Total effect	-0.26	0.04	-0.35 to -0.17	< 0.001
Direct effect	-0.16	0.04	-0.25 to -0.07	0.001
Indirect effect (PHSU-DHL-GHQ-12)	-0.10	0.02	-0.15 to -0.06	< 0.001

Control variables: Age, gender, education level, occupation. DHL: Digital health literacy; PHSU: Preventive health service utilization; GHQ-12: 12-item General Health Questionnaire; 95%CI: 95% confidence interval.

DISCUSSION

This retrospective cohort study systematically explored the relationship between hospital staff's PHSU, DHL and mental health, and for the first time confirmed that DHL played a partial mediating role between PHSU and mental health. This finding not only enriches the theoretical research in the field of health behaviors and mental health, but also provides important practical basis for formulating mental health intervention strategies for hospital staff. The research results showed that during the three-year period, higher PHSU levels were significantly associated with lower levels of psychological distress and job burnout. This conclusion was consistent with the results of previous studies[6]. In this study, 56.0% of hospital staff had significant psychological distress, which was similar to the data from previous reports on the prevalence of psychological distress among healthcare workers[7]. It is worth noting that the 56.0% prevalence rate we observed in the hospital sample is comparable to the 52%-58% range reported in multi-center surveys conducted on Chinese medical professionals during and after the coronavirus disease 2019 pandemic[8]. This indicates that psychological distress, as a long-term occupational health challenge rather than a short-term crisis phenomenon, persists. This cross-study consistency further confirms the general applicability of our research conclusions regarding the psychological burden of this population. Additionally, the burnout of healthcare workers may also be related to professional traits such as empathy[9], which further highlights the urgency of implementing intervention measures for the mental health of hospital staff. Hospital employees face high levels of work pressure and higher occupational health risks; specific stressors (such as moral dilemmas) may also exacerbate burnout[10].

Regular participation in preventive health services (including health check-ups, vaccinations, and disease screenings) helps to promptly identify potential health issues and intervene early, thereby reducing negative emotions such as anxiety and worry caused by unknown health risks[11], and also reducing work avoidance behaviors related to stress[12]. In addition, the process of engaging in preventive health services is itself a positive health behavior that enhances individuals' sense of control over their own health and improves psychological resilience, an enhancement closely linked to personal and work-related factors[13], thereby alleviating emotional exhaustion and depersonalization and promoting personal accomplishment. This process may also be achieved by reinforcing positive interactions among occupational value appraisal, self-esteem, and self-perception[14]. Meanwhile, improvements in health status resulting from PHSU contribute to higher job satisfaction, which is significantly and positively associated with mental health[15]. Furthermore, hospitals generally provide employees with convenient access to preventive health services, and active participation in these services reflects their attention to their own health; this elevated health consciousness may further extend to the psychological domain, facilitating the adoption of mental health-promoting behaviors.

Studies have shown that the PHSU level has a significant positive predictive effect on the DHL level. This means that hospital staff who are more actively involved in preventive health services have a higher DHL level. Although previous studies on DHL mainly focused on social demographic predictors such as age and education level[3,4], this study is the first to confirm that actual medical service utilization behaviorn especially PHSU, is an independent and important predictor of the increase in healthcare workers' DHL levels. This finding expands the existing literature framework, shifting the research focus from static demographic association factors to modifiable behavioral determinants (i.e., digital health capabilities), and reveals a new way for institutional health services to actively enhance employees' digital literacy. Potential mechanisms underlying this relationship are as follows. With the digital transformation of healthcare, digital health tools have been widely integrated into hospital-based preventive health services, such as electronic health record inquiries, online appointments for physical examinations, digital vaccination reminders, and electronic report interpretation[16]. When healthcare workers participate in the aforementioned preventive health services, they need to frequently come into contact with various digital tools and gradually accumulate experience in obtaining, evaluating, and applying digital health information[17]. For instance, behaviors such as querying physical examination reports through online platforms and interpreting relevant indicators, making vaccine appointments, and receiving health reminders through digital systems can promote the development of digital health skills and thereby enhance the level of DHL. Moreover, the professional characteristics of healthcare workers make them urgently in need of health information; during the participation in preventive health services, they are more inclined to actively learn the usage methods of digital health tools, which further strengthens the positive correlation between PHSU and DHL[18]. It is crucial that this study has confirmed that DHL plays a partial mediating role between PHSU and mental health among healthcare workers, meaning that PHSU not only directly promotes mental health but can also indirectly improve mental health by enhancing the level of DHL. The internal mechanism of this mediating path can be explained from two aspects. First, a higher level of DHL enables healthcare workers to obtain, evaluate, and apply digital health information more efficiently[19]. After participating in preventive health services, employees with high DHL can quickly interpret physical examination results, obtain personalized health guidance through digital platforms, and adjust their lifestyle in a timely manner to reduce health risks, thereby alleviating psychological distress[20]. For example, recording health data, receiving professional advice through health management applications, and communicating screening results with physicians via online consultation platforms enhance a sense of health control and relieve anxiety caused by health information asymmetry[21]. Second, employees with high DHL are more adept at using digital tools for psychological adjustment, such as emotion management training and joining online peer support groups through mental health applications. Such digital self-help mental health interventions have been proven to exert positive effects on preventing occupational burnout[22], further reducing burnout and improving mental health levels[23]. Compared with prior studies that have examined either the direct link between PHSU and mental health (e.g., Almutairi et al[6] found a protective association of lifestyle changes with mental well-being among emergency staff) or the independent association between health literacy and psychological outcomes[4], the present study provides a novel contribution by demonstrating the triadic mediating mechanism of “PHSU-DHL-mental health”.

Results of the present investigation have significant theoretical and practical implications. Theoretically, this is the first study to investigate the mediating role of DHL in the relationship between PHSU and mental health among hospital personnel, enriching the theoretical model of health behaviors and mental health. Most existing studies have focused on the direct effect of PHSU on mental health, whereas this study uncovers the indirect pathway of “PHSU-DHL-mental health”, providing a new perspective for understanding the influencing factors of mental health in occupational populations. It also verifies the critical role of DHL among healthcare professionals, expands the application scenarios of DHL, and offers a theoretical reference for future research. Targeted interventions can be developed based on these findings. First, hospitals should further optimize the process of preventive health services by providing more convenient access to physical examination appointments, vaccinations, and disease screenings for employees, such as increasing the frequency of physical examinations, expanding screening items, and optimizing digital appointment platforms to encourage regular participation in preventive health services. Simultaneously, references can be made to the key determinants of mental health promotion programs in healthcare workplaces[24] to ensure the accessibility and pertinence of services. Second, personalized DHL training courses should be designed according to the occupational characteristics of hospital personnel, including skills in retrieving digital health information, methods of information evaluation, use of digital health tools, and the cultivation of digital resilience. Notably, digital resilience training has been proven effective for healthcare workers in randomized controlled trials[25] and can improve employees' DHL levels. Third, relying on the existing digital platforms of hospitals, resources, including mental health assessment, emotion management training (e.g., mindfulness intervention[26]), and online psychological counseling should be integrated, and risk communication interventions should be introduced to promote help-seeking behaviors[27], providing a one-stop digital mental health service for employees. Precise interventions should be implemented, especially for high-stress occupational groups such as clinical nurses and emergency department physicians.

Identifying DHL as a partial mediating variable between PHSU and mental health provides a specific and operational framework for hospital-level intervention measures. Based on the “PHSU-DHL-mental health” path confirmed by this study, three-level implementation strategies can be proposed: Firstly, at the service optimization level, hospitals should redesign preventive health service processes to enhance accessibility and participation, including simplifying online appointment systems for physical examinations and vaccinations, embedding automated digital reminder functions for disease screening, and providing user-friendly electronic report interpretation tools. These improvements not only help promote continuous participation in PHSU but also provide natural opportunities for healthcare professionals to practice digital health skills; Secondly, at the training and education level, customized DHL training plans should be developed for different occupational roles - for example, clinical nurses with heavy workloads can receive short, modular digital health training through mobile applications, while doctors need to learn advanced modules on digital data interpretation and remote medical communication; Additionally, incorporating the recently validated digital resilience training[25] that has been effective through randomized controlled trials targeting healthcare professionals will further enhance employees' ability to effectively manage digital health information. Thirdly, at the digital resource integration level, hospitals can utilize existing digital infrastructure to build a comprehensive mental health service platform, which integrates psychological assessment (such as regular screening with the GHQ-12 scale), self-help intervention measures (such as mindfulness-based stress reduction modules), online peer support groups, and direct access to professional psychological counseling services. The platform can also incorporate risk communication intervention measures to promote patients' proactive behavior of seeking help[27], especially in high-stress departments such as emergency departments and intensive care units. These strategies have clear application value in routine hospital management: The occupational health department can incorporate DHL indicators into annual health assessments; the human resources department can incorporate PHSU participation into employee health incentive plans; hospital managers can evaluate the effectiveness of intervention measures through longitudinal tracking of PHSU indicators and mental health outcomes. Ultimately, this comprehensive approach is expected to expand from individual hospitals to regional medical networks, providing a replicable model for systematic improvement of the mental health of medical workers, while also helping to enhance patient safety and medical service quality.

This study has several limitations. Firstly, the single-center design is limited to a single tertiary hospital, which may lead to selection bias and limit the generalizability of the research results; the majority of the participants are female and their occupational composition is specific, further restricting the external validity of the research results - the conclusions may not directly apply to healthcare workers in primary or secondary hospitals, or to populations with different demographic characteristics. Secondly, although the PHSU data comes from electronic health records, its retrospective nature may lead to incomplete information (such as unrecorded self-screening behaviors), which will affect the accuracy of the comprehensive PHSU index. Thirdly, DHL and mental health outcomes are both evaluated through cross-sectional studies; combined with the retrospective PHSU data, no causal inferences can be drawn, and the time series relationship assumed in the mediating model cannot be clearly verified. Future research should adopt a longitudinal research design to verify the causal path direction proposed in this study. Fourthly, all psychological measurement indicators (DHL, GHQ-12, and MBI) rely on self-report questionnaires, and such methods have inherent limitations due to being susceptible to social expectation bias, recall bias, and subjective interpretation bias. Additionally, the measurement of mediating variables and outcome variables using self-report tools may lead to the risk of common method variance and may exaggerate the observed association strength. Fifthly, several important confounding variables, including social support, work stress intensity, structural empowerment, and professional competence, were not included in the control. Previous studies have shown that social support may mediate the relationship between PHSU and mental health[28], and structural empowerment and professional competence are closely related to stress symptoms[29]; future research should incorporate these factors into a more comprehensive regulatory mediating model. Sixthly, although the 100% questionnaire response rate indicates a high level of participation, it may also reflect the design of mandatory questionnaire completion on an online platform, which may affect the quality of responses; however, all questionnaires have undergone completeness verification. Finally, this cohort is entirely composed of personnel who have been continuously employed since 2021, which may have the "health worker effect", leading to a more positive health outcome bias in the observed association.

CONCLUSION

The results of this retrospective cohort study show that the previous PHSU behavior of healthcare workers is not only directly related to their current better mental health status, but can also indirectly improve their mental health level by enhancing DHL. DHL plays a partial mediating role between PHSU and mental health. Therefore, intervention measures aimed at improving the mental health of healthcare workers should focus on promoting participation in preventive health services and enhancing DHL levels. By optimizing service processes, conducting targeted training, and integrating digital resources, the mental health levels of healthcare workers can be comprehensively improved, thereby ensuring the quality of medical services and patient safety.