Assessing Occupational Work-Related Stress and Anxiety of Healthcare Staff During COVID-19 Using Fuzzy Natural Language-Based Association Rule Mining

Abstract

Background/Objective: Frontline healthcare staff who contend diseases and mitigate their transmission were repeatedly exposed to high-risk conditions during the COVID-19 pandemic. They were at risk of mental health issues, in particular, psychological stress, depression, anxiety, financial stress, and/or burnout. This study aimed to investigate and evaluate the occupational stress of medical doctors, nurses, pharmacists, physiotherapists, and other hospital support crew during the COVID-19 pandemic in Saudi Arabia. Methods: We collected both qualitative and quantitative data from a survey given to public and private hospitals using methods like correspondence analysis, cluster analysis, and structural equation models to investigate the work-related stress (WRS) and anxiety of the staff. Since health-related factors are unclear and uncertain, a fuzzy association rule mining (FARM) method was created to address these problems and find out the levels of work-related stress (WRS) and anxiety. The statistical results and K-means clustering method were used to find the best number of fuzzy rules and the level of fuzziness in clusters to create the FARM approach and to predict the work-related stress and anxiety of healthcare staff. This innovative approach allows for a more nuanced appraisal of the factors contributing to work-related stress and anxiety, ultimately enabling healthcare organizations to implement targeted interventions. By leveraging these insights, management can foster a healthier work environment that supports staff well-being and enhances overall productivity. This study also aimed to identify the relevant health factors that are the root causes of work-related stress and anxiety to facilitate better preparation and motivation of the staff for reorganizing resources and equipment. Results: The results and findings show that when the financial burden (FIN) of healthcare staff increased, WRS and anxiety increased. Similarly, a rise in psychological stress caused an increase in WRS and anxiety. The psychological impact (PCG) ratio and financial impact (FIN) were the most influential factors for the staff's anxiety. The FARM results and findings revealed that improving the financial situation of healthcare staff alone was not sufficient during the COVID-19 pandemic. Conclusions: This study found that while the impact of PCG was significant, its combined effect with FIN was more influential on staff's work-related stress and anxiety. This difference was due to the mutual effects of PCG and FIN on the staff's motivation. The findings will help healthcare managers make decisions to reduce or eliminate the WRS and anxiety experienced by healthcare staff in the future.

Keywords: burnout; depression; fuzzy association rule mining; mental health; occupational stress; resilience.

Figure 1

Fuzzy association rule mining model for WRS and anxiety prediction.

Figure 2

Distribution of research samples.

Figure 3

Percent of healthcare staff suffered psychological stress during COVID-19 pandemic.

Figure 4

Percent of staff suffered financial stress during COVID-19 pandemic.

Figure 5

Distribution of WRS during COVID-19 pandemic.

Figure 6

Normality plot of WRS (a), financial stress (b), and psychological stress (c).

Figure 7

Variations among socio-demographic factors (SCFs) related to WRS and anxiety.

Figure 8

Distribution of clusters among demographic characteristics.

Figure 9

Cluster analysis of PCG, FIN, and WRS and anxiety used for fuzzy rule set development.

Figure 10

Membership functions of fuzzy variables for association rule mining.

Figure 11

Fuzzy inference system for association rules mining of WRS and anxiety.

Figure 12

The stress level of staff measured by FARM approach.