Disaster Preparedness in Saudi Arabia’s Primary Healthcare Workers for Human Well-Being and Sustainability

Abstract

The preparedness of healthcare workers for disaster situations depends on their technical skills, disaster knowledge, and psychosocial strength, including teamwork and emotional regulation. This study aims to assess disaster preparedness among healthcare professionals in primary healthcare centers (PHCs) in Alqurayat, Saudi Arabia, with a specific focus on evaluating technical competencies, psychosocial readiness, and predictive modeling of preparedness levels. A mixed-methods approach was employed, incorporating structured questionnaires, semi-structured interviews, and observational data from disaster drills to evaluate the preparedness levels of 400 healthcare workers, including doctors, nurses, and administrative staff. The results showed that while knowledge (mean: 3.9) and skills (mean: 4.0) were generally moderate to high, notable gaps in overall preparedness remained. Importantly, 69.5% of participants reported enhanced readiness following simulation drills. Machine learning models, including Random Forest and Artificial Neural Networks, were used to predict preparedness outcomes based on psychosocial variables such as emotional intelligence, teamwork, and stress management. Sentiment analysis and topic modeling of qualitative responses revealed key themes including communication barriers, psychological safety, and the need for ongoing training. The findings highlight the importance of integrating both technical competencies and psychosocial resilience into disaster management programs. This study contributes an innovative framework for evaluating preparedness and offers practical insights for policymakers, disaster planners, and health training institutions aiming to strengthen the sustainability and responsiveness of primary healthcare systems.

1. Introduction

The COVID-19 pandemic intensified existing challenges in healthcare systems, exposing global gaps in disaster readiness. In Saudi Arabia, over 547,000 cases and 8760 deaths by mid-2021 strained the system, especially in underserved regions like Alqurayat. Primary healthcare centers (PHCs) bore the brunt, operating with limited protective gear and under intense psychological stress. Studies report that over 68% of Saudi healthcare workers experienced moderate to high stress, and 52% showed anxiety or emotional exhaustion. In the healthcare context, disasters refer to both natural and human-made emergencies—such as pandemics, earthquakes, mass casualty incidents, and chemical spills—that overwhelm local resources and disrupt health service delivery [1]. Factors influencing disaster vulnerability in healthcare include inadequate infrastructure, workforce shortages, unstandardized protocols, and limited psychological support systems. Regional comparisons highlight critical disparities: for instance, while countries like the U.S. and Japan implement regular simulation drills and maintain emergency stockpiles, low-resource settings in the Gulf or South Asia often lack these structured mechanisms, as seen in Alqurayat’s limited logistical capacity. Effective frameworks must integrate technical skills with emotional intelligence, stress management, and teamwork training to ensure frontline PHC personnel can sustain critical operations during crises.

Over the past decade, global awareness of disaster preparedness in healthcare has grown, with the WHO (2019) identifying it as a critical cross-sectoral concern [2]. Despite this emphasis, substantial gaps remain, particularly within primary healthcare centers (PHCs), which serve as the first point of contact in many communities. Regional disparities persist, with studies showing that disaster readiness levels vary widely by geography and provider type. Research from the U.S. and Europe highlights inconsistencies in hospital preparedness and uneven training in disaster management practices, underscoring the need for more standardized, context-sensitive strategies that prioritize PHC capacity building across diverse healthcare settings [3,4,5].

The lack of both resources and proper infrastructure creates substantial obstacles for disaster preparedness programs in healthcare across developing countries. According to [6], disaster management research for low- and middle-income countries, as demonstrated through a systematic review, has shown that local contexts must guide the creation of new approaches. The study established that the application of these ideas must comply with global best practice principles while recognizing the realities of the present healthcare system and local challenges within specific treatment environments.

Healthcare disaster preparedness encompasses numerous fields of knowledge, as well as practical skills, which contribute to its extensive nature. Numerous studies have defined core competencies for disaster management in healthcare, including rapid triage, emergency risk communication, decision-making under uncertainty, and crisis leadership. These skills are supplemented by psychosocial competencies such as emotional intelligence, adaptive teamwork, and psychological first aid. However, a gap persists in translating these competencies into standardized training frameworks within PHC systems, particularly in LMICs. The authors [7] present vital competencies in disaster management, which comprise risk evaluation, planning, communication, triage, and crisis stabilization protocols. The developed framework serves as a standard for training program development throughout different nations.

This initial research prompted various experts to investigate different aspects of disaster preparedness. The evaluation of psychological readiness for healthcare workers during disasters necessitates a simultaneous consideration of technical competencies, as noted by [8]. The research highlighted the urgent need to develop more effective scientific and psychiatric training programs that focus on disaster preparedness and response.

Research exists that defines disaster management core competencies; however, additional clarification is needed to address the primary healthcare needs. The explanation for these problems may involve insufficient funding, inadequate training, and regional disparities. The current analysis of primary healthcare facilities by [9] reveals that they lack adequate emergency equipment and logistical capabilities. Healthcare providers working in these locations operate with the necessary equipment for mass disaster response, according to the research findings.

The primary healthcare facilities face critical training-related challenges as another significant priority. Despite these efforts, many healthcare systems lack comprehensive strategies to align both technical and psychosocial preparedness. For example, while South Korea’s urban centers have well-developed disaster manuals, rural areas continue to lag in structured psychosocial training. This global inconsistency underscores the need for adaptive models tailored to the realities of under-resourced primary care systems. Only 35 percent of Spanish primary care doctors participated in official disaster management training over the past five years, according to [10], who conducted research with 50 physicians. Limited exercise practice reduces the effectiveness of healthcare providers during emergency events.

Another issue is the need for uniformity in disaster preparedness based on the regions of the countries where disasters occur, especially in countries with diverse geographical and economic terrains. The study by [11] in South Korea also identified disparities in disaster preparedness among primary healthcare centers (PHCs) across urban and rural areas. The research also revealed that rural centers and their staff were less well-prepared and less trained in disaster management than their urban counterparts.

The shift towards psychological preparedness is closely linked to the fields of work and organizational psychology, which increasingly focus on employee health and the creation of stable, healthy work environments. Modern organizations aiming to enhance their significance in the face of challenges must consider the role of promoting well-being to advance individual and organizational development. A positive and psychologically safe workplace is particularly crucial in healthcare, where burnout and various mental health issues frequently arise among staff. Consequently, general management in healthcare organizations should implement policies and interventions designed to reinforce not only the technical capabilities but also the psychological resilience of their workforce. This integrated approach contributes to the development of more efficient organizational structures that can be effectively managed during future crises [12].

Human sustainability entails developing the capacity to generate both organizational and individual value, which extends beyond short-term benefits. Organizations must create environments that prioritize disaster management technical expertise and develop essential soft skills, including emotional intelligence, team collaboration, and effective communication [13]. Soft competencies hold the same importance as hard skills for creating successful disaster response practices. Conducting organizational solution development through competency training enhances employee readiness across technical and interpersonal skills [14].

The analysis of disaster preparedness through research has primarily focused on technical engineering aspects without adequate consideration of psychosocial factors. Healthcare personnel need to be skilled in stress management, along with clear communication skills, especially when facing crises [3]. The primary healthcare sector requires this approach given its restricted resources while managing large patient numbers. Appropriate solutions for this problem must include training and education, along with mental health programs and the implementation of best practices in both healthcare workers’ educational and workplace environments, as well as personal stress management strategies. The methods employed here align with current trends that promote sustainable working environments, thereby enhancing organizational functioning and supporting staff members in overcoming their challenges [15].

Healthcare workers face a direct link between their emotional well-being and both their health state and the structure of their workplaces. The essential readiness of healthcare providers stems from their problem-solving capability, which needs systematic educational development. Healthcare organizations struggle to create resilient disaster management teams because they lack person-centered strategies for workforce development that prioritize member health and disaster response effectiveness. Research supports the effectiveness of well-being and sustainability as fundamental work principles for the approaching era, especially within healthcare [16].

The essential concern arises because emergency preparedness planning within disaster management necessitates an additional evaluation of primary healthcare center readiness levels. Conducting no current disaster preparedness assessments is remarkably concerning, as primary healthcare centers serve as the primary healthcare institutions in emergencies [17].

Research evidence about disaster preparedness within primary healthcare institutions represents a major international concern affecting Alqurayat and similar locations. Research deficiencies render it impossible to deploy the required interventions that support critical healthcare facility preparedness, thereby generating adverse effects on healthcare system resilience [18]. Knowledge of disaster readiness abilities among healthcare employees, combined with principles of disaster psychology, enhances the identification of strategic improvement approaches [19]. Primary healthcare systems face obstacles in establishing permanent operational readiness practices for workforce health because they often fail to address fundamental organizational elements.

The need to address this research gap remains essential for multiple vital reasons. Community members typically direct their disaster response to primary healthcare centers, making the initial response both more and less effective based on the centers’ preparedness levels [20]. Healthcare workers who deliver both clinical competency and psychological support play a crucial role in organizational strategies that help build resilience among individual staff members and entire healthcare service systems. The facilities provide continuous medical care for both emergencies and post-emergencies, especially for elderly patients and individuals with chronic disorders, to demonstrate the need for skilled and resilient personnel networks. The assessment of healthcare provider readiness exposes organizational system weaknesses, which organizations use to create lasting employee health support systems. The emergency response capabilities of healthcare organizations will improve through directed awareness of these domains while developing sustainable workplace environments for medical staff.

The foundation of this study establishes a goal to understand which healthcare providers in Alqurayat primary healthcare centers recognize their disaster management competencies. The research assesses the evolution of disaster preparedness while investigating both prosperous and challenging obstacles that affect the wellness and resilience levels of healthcare providers. This research highlights the central role of emotional resilience, along with soft skills, in disaster management, as it evaluates the preparedness levels, confidence, and psychological readiness of frontline healthcare providers.

The primary objective of this study is to systematically assess the disaster preparedness of healthcare professionals working in primary healthcare centers (PHCs) in the Alqurayat region of Saudi Arabia. This assessment focuses on three core domains: technical competencies, operational readiness, and psychosocial resilience—including emotional regulation, stress-handling capacity, and crisis response confidence. By evaluating these components, the study aims to generate a comprehensive understanding of frontline healthcare capacity in disaster scenarios. The secondary objective is to investigate the predictive influence of psychosocial factors—such as emotional intelligence, teamwork effectiveness, stress management, psychological safety, and employee engagement—on overall disaster response performance. To achieve this, the study employs an integrated mixed-methods approach, combining structured surveys and qualitative interviews with advanced machine learning techniques, including Random Forest and Artificial Neural Networks. This represents the first such effort in Saudi Arabia’s PHC context to fuse psychosocial analytics with AI-based modeling, offering novel policy insights for building sustainable and resilient health systems in the Gulf Cooperation Council (GCC) and other low- and middle-income country (LMIC) settings.

Despite growing attention to healthcare disaster management, existing research rarely integrates psychosocial readiness with machine learning techniques to assess preparedness. Few studies systematically evaluate how emotional intelligence, teamwork, and stress coping interact with predictive modeling to forecast healthcare resilience. This study addresses this knowledge gap.

The research is guided by the following specific objectives:

• To assess the technical knowledge and disaster response skills of primary healthcare professionals.
• To evaluate psychosocial readiness, including emotional intelligence, teamwork, and stress management.
• To use machine learning models (Random Forest and Artificial Neural Networks) to predict preparedness levels.
• To identify key themes and sentiments from qualitative data related to disaster preparedness.

This study represents the first mixed-methods investigation of disaster preparedness within primary healthcare centers (PHCs) in Saudi Arabia, uniquely integrating machine learning models with psychosocial metrics such as emotional intelligence, teamwork, and psychological safety. By combining quantitative and qualitative data with predictive analytics (e.g., Random Forest, ANNs), the research offers a multidimensional understanding of healthcare workforce readiness. These insights are highly relevant for policy formulation in Gulf Cooperation Council (GCC) countries and other low- and middle-income countries (LMIC) settings, where PHCs serve as the backbone of community healthcare but often lack the necessary resilience infrastructure for disaster scenarios.

2. Materials and Methods

The study employs a mixed research design, incorporating both quantitative and qualitative data collection and analysis methodologies to assess healthcare providers’ perceptions of knowledge, skills, and preparedness in the event of a disaster in primary healthcare centers. These include measured data collected from structured questionnaires, while qualitative data are collected through semi-structured interviews with participants and observer data from disaster drills. This design ensures that the subject matter’s coverage is broader and more profound because it captures statistical patterns and narrows them down to the managers’ experiences. The use of mixed methods is desirable as the results will be cross-checked by the two different methodologies, ensuring higher validity and reliability of the results that refer to objective and subjective disaster preparedness features. The methodological flow of the study is presented in Figure 1 below:

The study participants comprised doctors, nurses, and administrative personnel working in various primary healthcare centers across the Alqurayat region, all of whom are actively involved in disaster management and emergency response activities. Alqurayat was selected as the study region due to its historical exposure to seasonal flooding, extreme temperatures, and cross-border population movements, which increase its vulnerability to both natural and human-induced health crises. Regional health authorities have previously conducted emergency preparedness drills in the area, and internal MOH risk classifications identify Alqurayat as a high-priority zone for disaster readiness evaluation. This substantiates the inclusion of Alqurayat under the ‘disaster-prone area’ criterion used in participant selection.

A target sample size of 400 participants was determined using a stratified random sampling technique, ensuring proportional representation across different professional roles and levels of experience. Participants were first categorized by occupational group (doctors, nurses, administrative staff) and then randomly selected from each group in proportion to their overall presence in the population. This sampling approach was employed to ensure equitable representation of all relevant stakeholder groups involved in frontline disaster preparedness. The specific inclusion and exclusion criteria applied during participant selection are summarized in

Table 1. Inclusion and exclusion criteria.

Category	Inclusion Criteria	Exclusion Criteria
Population	Healthcare providers (doctors, nurses, paramedics) working in primary healthcare centers	Non-healthcare workers
Experience	Healthcare providers with at least one year of experience in primary care or emergency settings	Less than one year of professional experience
Geographic Area	Participants working in primary health centers located in disaster-prone areas or with past disaster experience	Healthcare providers working outside disaster-prone areas.
Language	Healthcare providers proficient in the official language of the survey or interview	Healthcare providers not fluent in the survey/interview language
Preparedness	Involvement in any disaster management training, drill, or simulation within the last five years	No participation in disaster management drills or simulations
Willingness	Healthcare providers who voluntarily consent to participate in the study	Healthcare providers who decline or withdraw consent
Age	Healthcare providers aged 18–65 years	Individuals younger than 18 or older than 65
Availability	Healthcare providers are available to complete surveys and conduct follow-up interviews as needed	Individuals unavailable for follow-up interviews or unable to complete the survey

. Additionally, the table presents the demographic profile of the surveyed healthcare workers. The majority were nurses aged between 31 and 40 years, and more than 60% held a bachelor’s degree or higher, reflecting the typical staffing structure in Saudi PHCs.

2.1. Data Collection

The data collection approach involved three stages: first, the use of structured questionnaires, followed by semi-structured interviews, and finally, observational assessments during disaster drills. The structured survey employed closed-ended questions to assess participants’ awareness, technical knowledge, disaster response skills, and psychosocial readiness. It consisted of 45 items, divided across three core domains: (1) perceived knowledge (items 1–21), (2) disaster management awareness and preparedness (items 22–34), and (3) self-reported skills and readiness (items 35–45). Responses were recorded using a 6-point Likert scale (1 = Strongly Disagree to 6 = Strongly Agree). The questionnaire was adapted from previously validated tools and pilot-tested for contextual relevance. A complete version of the instrument is provided in Supplementary File S1 and is also available online via this Google Form link (https://docs.google.com/forms/d/e/1FAIpQLSdsNmKRvw7T5ee0ZiSwfMppVdTY7PpUA1wTdQvlBNIIjsWmdQ/viewform?usp=sf_link, accessed on 5 April 2023).

The constructs of stress, teamwork, and emotional intelligence were measured using adapted items from widely used standardized instruments. Emotional intelligence was assessed based on items adapted from the Wong and Law Emotional Intelligence Scale (WLEIS), stress levels were gauged using selected items from the Perceived Stress Scale (PSS), and teamwork was evaluated using components from the TeamSTEPPS Teamwork Attitudes Questionnaire. All items were reviewed by field experts and pilot-tested to ensure contextual appropriateness for Saudi PHC settings.

The semi-structured interviews were conducted either in person or via videoconference, based on participant availability and location, to ensure inclusivity and logistical flexibility. These interviews explored real-world disaster response practices, psychological challenges, and perceived institutional preparedness gaps. Additionally, observational data were collected during organized disaster simulation drills in primary healthcare facilities, offering authentic behavioral insights and team dynamics. Together, this triangulated approach allowed for a holistic assessment of both subjective and objective dimensions of disaster preparedness.

The data triangulation method will include semi-structured interviews conducted either face-to-face or through video conferencing, according to participant availability. The interviews offer systematic ways to learn about participants’ disaster-related practices and their perspectives on priorities, response difficulties, and preparedness challenges. The combination of quantitative data, which demonstrates readiness levels, and qualitative data, which evaluates stress factors and psychological aspects, will yield information on coping strategies. The researchers will record observational data from disaster drills conducted in primary healthcare facilities. Interviews were conducted either in person or via videoconference based on participant location and availability, ensuring flexibility and inclusiveness in data collection. Surveys and interviews will receive additional support from this data, as it will act as a real-world test to analyze participant readiness levels and authentic team responses during disaster conditions. The research aims to develop sustainable healthcare organizations and resilient facilities through this investigative method.

Overall, the knowledge and skills among participants are high, reflecting a moderate to high level of proficiency. Drill participation is widespread, with most providers reporting active engagement in drills. Preparedness scores range from mild to high, typically not falling below 3.5 and not exceeding 4. In comparing the various roles within the medical field, doctors demonstrate greater preparedness than their counterparts in other professions. Conversely, nurses often encounter more stress in their work environment, despite exhibiting high emotional intelligence and a strong commitment to teamwork. Most administrative staff members score around the midpoint across all indices. This analysis highlights the importance of emotional intelligence, teamwork, and stress management, all of which are crucial factors in the systematic improvement of healthcare organizations and the overall well-being of staff.

Table 2. Reliability, validity, and input variables used in ML model with VIF scores.

Variable	Type	Cronbach’s Alpha (α)	KMO	Bartlett’s Test (p)	VIF	Role in ML Model
Stress Levels	Psychosocial Factor	0.84	0.68	0.001	1.3	Independent Variable (Input)
Emotional Intelligence (EI)	Psychosocial Factor	0.87	0.71	0.000	1.1	Independent Variable (Input)
Teamwork	Team Competency	0.82	0.66	0.003	1.05	Independent Variable (Input)
Psychosocial Readiness	Composite Readiness	0.86	0.70	0.001	1.4	Independent Variable (Input)
Employee Engagement (EE)	Organizational Factor	0.79	0.64	0.002	1.2	Independent Variable (Input)
Psychological Safety	Organizational Factor	0.81	0.67	0.001	1.1	Independent Variable (Input)
Preparedness Score	Overall Readiness Score	—	—	—	1.2	Target Variable (Output)

presents the independent (input) variables and the target (output) variable used in the machine learning analysis, along with key statistical indicators—Cronbach’s Alpha (for reliability), KMO (sampling adequacy), Bartlett’s Test (sphericity), and VIF (multicollinearity check). The selected input variables, including Emotional Intelligence, Teamwork, Stress Levels, and organizational factors such as Employee Engagement and Psychological Safety, were derived from validated instruments and pilot-tested for contextual relevance. Their inclusion was based on both theoretical relevance and statistical robustness in assessing disaster preparedness.

The preparedness score, a composite index based on participants’ responses related to knowledge, skills, and readiness, served as the target output variable for the machine learning models (Random Forest and Artificial Neural Network). All VIF scores remain well below the conservative threshold of 3.3, ensuring no multicollinearity. Additionally, Cronbach’s Alpha values across all constructs exceed 0.79, and KMO scores range from 0.64 to 0.71, demonstrating acceptable reliability and sampling adequacy. These indicators collectively enhance the validity, reliability, and interpretability of the model, reinforcing the credibility of predictive outcomes and supporting replicability in future research contexts.

2.2. Analytical Framework

The research design integrates both quantitative and qualitative methods to comprehensively assess disaster preparedness levels among medical staff in primary healthcare centers. For quantitative analysis, machine learning algorithms—namely, Random Forest (RF) and Artificial Neural Networks (ANNs)—were employed to model disaster preparedness outcomes.

The ML component incorporated both classification and regression tasks. Classification models (e.g., RF) were used to categorize participants into preparedness levels (Low, Moderate, High), while regression models (ANNs) were designed to predict continuous preparedness scores based on key input variables. These predictors included emotional intelligence, stress levels, teamwork, psychosocial readiness, psychological safety, and employee engagement—all selected based on theoretical and empirical relevance from the survey results.

The total dataset included 400 healthcare worker responses. An 80/20 data split was applied, allocating 320 responses for training and 80 for validation/testing. This approach ensured robust model performance and generalizability across scenarios. Additionally, Variance Inflation Factor (VIF) scores for all predictor variables were assessed to confirm the absence of multicollinearity, supporting model validity.

For qualitative insights, sentiment analysis was performed to evaluate the emotional tones in open-ended responses. An LDA model was used to extract dominant themes from interview transcripts and narrative survey feedback.

This integrated analytical approach revealed that emotional intelligence, teamwork, and stress management are equally as important as technical knowledge in shaping preparedness outcomes. The combination of both qualitative and predictive analytics strengthens understanding of organizational resilience and supports the development of sustainable healthcare systems in disaster-prone contexts. All data processing and analysis were conducted using Python 3.10, with Scikit-learn for machine learning models (RF, ANNs), NLTK for sentiment analysis, and Gensim for Latent Dirichlet Allocation (LDA)-based topic modeling.

The research protocol was reviewed and approved by the Research Ethics Committee of the Alqurayat Health Affairs Directorate, Ministry of Health, Saudi Arabia, under approval number (H-13-S-071). All procedures followed ethical standards, and informed consent was obtained from all participants. Data confidentiality and participant anonymity were strictly maintained throughout the study.

3. Results

3.1. Quantitative Findings: Technical and Psychosocial Preparedness

Table 3. Disaster preparedness metrics by age group: skills, experience, and psychosocial factors.

Variable	<20	20–30	30–40	40–50	50–60
Sample Size (Age)	10	100	150	80	40
Experience (Years)	20	100	150	80	40
Knowledge Level	10	40	100	80	30
Skills Level	20	50	90	70	30
Drill Participation	20	50	100	150	250
Preparedness Score	50	120	150	80	0
Stress Levels	High	Moderate	Moderate	Low	Low
Emotional Intelligence (EI)	Low	Moderate	High	High	Moderate
Teamwork	Moderate	High	High	Moderate	Low
Psychosocial Readiness	Low	Moderate	High	High	Moderate
Employee Engagement (EE)	Low	High	High	Moderate	Low
Psychological Safety (Psy.Sfty)	Low	Moderate	High	Moderate	High

EI = emotional intelligence; EE = employee engagement; Psy.Sfty = psychological safety.

summarizes disaster preparedness indicators across various age groups, including experience levels, technical competencies, and psychosocial attributes. The table highlights generational differences in emotional intelligence, stress levels, and overall readiness. These variations offer insight into how age and professional maturity influence disaster response capacity.

Table 3 presents disaster preparedness indicators across different age groups, highlighting key variations in professional experience, technical skills, and psychosocial metrics. Older participants (particularly those aged 50–60) demonstrate the highest levels of drill participation and psychological safety, whereas younger participants (<20 years) exhibit lower preparedness scores and emotional intelligence levels. These findings suggest that both age and accumulated experience significantly influence disaster readiness, particularly in relation to emotional regulation and team-based coordination.

Table 4. Preparedness scores and psychosocial factors across job roles.

Variable	Doctor	Nurse	Administrative
Minimum Preparedness Score	2.0	2.5	2.5
25th Percentile (Q1)	3.0	2.8	3.0
Median (50th Percentile)	3.5	3.2	3.5
75th Percentile (Q3)	4.0	3.5	4.0
Maximum Preparedness Score	4.5	4.0	4.5
Stress Levels	Moderate	High	Moderate
Emotional Intelligence (EI)	High	Moderate	High
Teamwork	High	High	Moderate
Psychosocial Readiness	Moderate	High	Moderate
Employee Engagement (EE)	High	Moderate	Low
Psychological Safety (Psy.Sfty)	High	Moderate	Low
Outliers Present	Yes	Yes	Yes

Abbreviations: EI = emotional intelligence; EE = employee engagement; Psy.Sfty = psychological safety.

provides a comparative overview of disaster preparedness indicators across professional roles in primary healthcare. Doctors demonstrated the highest preparedness scores and emotional intelligence, reflecting greater confidence in emergency response. In contrast, nurses exhibited elevated stress levels despite strong teamwork and psychosocial readiness. Administrative staff showed moderate and more uniform scores across most variables, indicating a need for tailored training programs that address their specific preparedness gaps. These role-based distinctions underscore the importance of targeted capacity-building strategies for different categories of healthcare personnel. These findings suggest that disaster management programs should be designed to include age- and role-specific training components. For example, emotional intelligence and stress regulation modules may benefit younger and nursing staff, while administrative personnel may require additional technical simulations. Integrating both psychosocial and operational competencies can enhance system-wide preparedness.

3.2. Psychosocial Readiness and Team Dynamics

This analysis builds on the quantitative results by exploring the role of psychosocial factors—such as emotional intelligence, stress levels, teamwork, and psychological safety—in shaping disaster preparedness among healthcare workers. Structural equation modeling reveals that stress positively correlates with preparedness, suggesting that moderate levels of pressure may enhance alertness and readiness. Teamwork significantly influences both psychosocial readiness and psychological safety, emphasizing the importance of cohesive team dynamics in managing emergencies effectively. Emotional intelligence, while essential, shows a complex interaction, where lower levels may hinder collaboration under high-stress conditions. These insights underscore the need to integrate psychosocial development alongside technical training in preparedness programs, reinforcing that resilient healthcare systems depend equally on emotional strength and coordinated team efforts.

Table 5. Factors impacting disaster preparedness: stress, teamwork, and psychosocial readiness.

Variable	Factor 1 (Stress and Readiness)	Factor 2 (Teamwork and Psychosocial Readiness)
Emotional Intelligence	−0.086	−0.400
Stress Levels	0.981	−0.110
Teamwork	−0.038	−0.697
Psychosocial Readiness	0.003	0.260
Preparedness Score	0.154	0.439
Psychological Safety	−0.076	0.285
Path	From	To	Path Coefficient
Factor 1 (Stress and Readiness) → Stress Levels	Factor 1 (Stress and Readiness)	Stress Levels	0.981
Factor 1 (Stress and Readiness) → Preparedness Score	Factor 1 (Stress and Readiness)	Preparedness Score	0.154
Factor 2 (Teamwork and Psychosocial Readiness) → Teamwork	Factor 2 (Teamwork and Psychosocial Readiness)	Teamwork	−0.697
Factor 2 (Teamwork and Psychosocial Readiness) → Psychosocial Readiness	Factor 2 (Teamwork and Psychosocial Readiness)	Psychosocial Readiness	0.260
Factor 2 (Teamwork and Psychosocial Readiness) → Psychological Safety	Factor 2 (Teamwork and Psychosocial Readiness)	Psychological Safety	0.285

presents statistical indicators derived from structural equation modeling (SEM), including standardized factor loadings and path coefficients. These metrics clarify the strength and direction of relationships among psychosocial variables and disaster preparedness outcomes. The structural equation model presented in Table 5 identifies both direct and indirect relationships between latent psychosocial factors and disaster preparedness outcomes among primary healthcare workers. The results reveal that stress levels exhibit a strong positive standardized factor loading with preparedness scores (β = 0.981), suggesting that moderate stress may serve as a motivator under high-pressure conditions. Additionally, teamwork demonstrates significant standardized loadings on psychosocial readiness (β = 0.260) and psychological safety (β = 0.285), reinforcing the importance of cohesive team dynamics in fostering resilience.

These findings align with the work of [21], which underscores the central role of psychosocial constructs—particularly teamwork and safety climate—in shaping disaster resilience. Emotional intelligence also displays a negative factor loading under Factor 2 (β = −0.400), indicating that lower emotional regulation may compromise effective collaboration if left unaddressed. Supporting evidence from [22] emphasizes the importance of cultivating a psychologically supportive culture and stress-management infrastructure to enhance both mental health and response performance.

Table 5, therefore, summarizes both factor loadings and path coefficients from SEM analysis, providing a conceptual framework that integrates psychosocial and technical dimensions. It offers a novel lens through which to evaluate disaster preparedness in primary healthcare settings. Organizations that embed psychological safety systems and team collaboration practices into their disaster protocols are more likely to achieve sustainable workforce resilience. Therefore, integrating technical training with psychosocial development is essential to building a robust and responsive healthcare infrastructure capable of navigating complex emergency environments.

3.3. Machine-Learning Predictions of Preparedness

The evaluation depicted in Figure 2 shows the actual disaster preparedness scores following the ANN model’s predictions of disaster readiness during psychological emergencies. The blue dashed line represents the actual preparedness scores, while the red line shows the predicted scores, which appear on the same y-axis. The Artificial Neural Network delivers acceptable estimations regarding disaster preparedness levels. The evaluation performance at lower scale points may be impacted by stress and the composition of the team personnel. This model serves an essential role because it defines disaster preparedness fundamentals by identifying vital readiness elements that encompass both emotional intelligence and team cohesion. The study confirms previous research, showing that psychological readiness plays an essential role during disaster events, as noted by [22]. By implementing psychosocial work integration methods, organizations can enhance their responses to hazardous healthcare systems, enabling them to develop and deploy available personnel for emergency response.

The results from Figure 3 demonstrate how psychosocial readiness, in conjunction with teamwork, creates essential elements for disaster preparedness, as determined by the Random Forest and Support Vector Machine models. The bar chart illustrates that psychosocial readiness and the preparedness score are the most influential factors affecting disaster response effectiveness based on relative feature prioritization. The research of [23,24] demonstrated the importance of psychosocial factors for effective crisis management. It also presents his insights on teamwork by emphasizing its importance during intense situations. The results indicated that emotional intelligence had a reduced effect, thus showing that other psychosocial aspects may be more critical for disaster response.

SVM model validation enabled the development of a classification system that divided providers into readiness groups for intervention purposes based on their defined maturity levels. The findings validate [25], who stress the necessity of training that targets individual professionals to build organizational resilience. Sustainability receives support through this study, which reveals that healthcare workforce sustainability requires both technical competence and organizational willingness. Taking action against disasters requires implementing employee training programs to distribute needed competencies alongside mental health support for staff.

The performance outcomes of the Random Forest (RF) and Artificial Neural Network (ANN) models in both classification and regression duties are presented in Figure 4. The RF model achieves superior performance among correctly predicted classes compared to actual courses, although the ANN exhibits minor variation, as indicated by higher variance metrics. Random Forest achieves a better alignment than the ANN during regression tasks because it demonstrates higher correlation rates with more precise value predictions. Random Forest stands as a superior model to the ANN in both classification and regression studies, yet it fails to excel simultaneously against the other model. These algorithms help evaluate disaster response system readiness to determine more effective, sustainable disaster planning strategies.

3.4. Sentiment and Topic Analysis of Qualitative Responses

Figure 5 illustrates the sentiment analysis of participants’ qualitative responses regarding a disaster preparedness program implemented in primary healthcare centers in Alqurayat. This program included simulation drills, emergency communication training, and psychosocial preparedness workshops aimed at enhancing technical and emotional readiness for disaster situations. Positive reactions, denoted in green, reflect confidence and approval, such as “I feel very positive about the new training program”. In contrast, negative responses, marked in red, convey dissatisfaction, exemplified by, “This is a disaster!”, indicating that the new system is ineffective. Responses categorized as contrary (in gray) suggest that the respondents are uncertain or hold a weaker opinion. The purpose of the program was to improve healthcare workers’ ability to respond to emergencies by strengthening both technical response skills and psychosocial competencies like teamwork, communication, and emotional resilience. Overall, the sentiment appears predominantly positive; however, some negative feedback highlights issues such as communication challenges and the absence of timely planning. Previous research, including studies by [26,27], has emphasized the importance of emotional resilience and teamwork in fostering a sustainable organizational culture. This analysis identifies key psychosocial factors, such as communication and training, that require further enhancement to improve efficiency and bolster the sustainability of disaster response operations.

Figure 6 illustrates the most frequently occurring terms from open-ended responses using both a word cloud and a topic-model bar chart, highlighting key themes such as “new”, “changes”, and “training”. The word cloud visualizes prominent concepts like “disaster”, “program”, and “completely”, suggesting that participants were primarily focused on newly introduced initiatives and their perceived impact. Plot a, the word cloud, visually represents the most frequently mentioned terms from open-ended feedback by healthcare workers on disaster preparedness. Words like “training”, “changes”, “system”, and “program” appear prominently, indicating that participants engaged actively with institutional updates and structured response mechanisms. The large presence of the word “new” reflects a collective focus on recently introduced protocols and preparedness strategies. The visualization highlights the thematic emphasis on practical implementation and system-based readiness. Plot B displays the results from Topic 1 of the LDA topic modeling analysis, ranking top terms based on their statistical weight in shaping the central narrative. Words such as “helpful”, “positive”, “feel”, and “okay” dominate the output, signifying a generally favorable emotional tone. This sentiment analysis supports the conclusion that disaster preparedness efforts were well-received and positively internalized by most participants. Together, the two plots provide both linguistic and statistical validation of staff engagement. The findings contribute by offering actionable insight into how healthcare workers perceive and respond to disaster management programs. These results emphasize the importance of combining structured training with supportive communication, which can guide future policy and training design for resilient healthcare systems.

Figure 7 presents a comparative evaluation of RF and ANN models across both classification and regression tasks related to disaster preparedness. Based on accuracy and F1-score, RF demonstrates stronger performance in classifying preparedness levels. Conversely, the ANN shows relatively better results in regression tasks, with a lower Mean Squared Error (MSE) and a slightly higher R-squared value.

However, overall predictive performance remains modest for both models, likely influenced by factors such as sample size, data noise, and variable complexity. These results are intended to provide an exploratory comparison of model behavior rather than assert high predictive capability. Accordingly, the choice between RF and ANNs should be guided by the nature of the task—categorical classification versus continuous prediction—and the quality of input data. Theoretical and managerial implications should be interpreted within this modeling context.

3.5. Theoretical Implications

This study contributes to the knowledge base in disaster management, specifically regarding the role of technical and psychosocial factors in disaster preparedness among healthcare providers. It builds upon previous studies by incorporating Random Forest and Support Vector Machine models to assess the readiness of healthcare workers (HCWs) in primary healthcare facilities. The study highlights the critical aspects of emotional intelligence, stress management, and group work in disaster preparedness, which have not received adequate attention in the disaster management literature. The research suggests a new, more comprehensive concept of preparedness that encompasses both practical skills, such as triage and risk assessment, and the mindset of healthcare staff. Thus, these conclusions imply that research enhancing disaster management needs to be broader and encompass more aspects.

3.6. Managerial Implications

For managers, the study emphasizes the importance of health establishments, mainly primary healthcare facilities, in providing comprehensive disaster management simulation training that encompasses both technical and psychosocial aspects. Healthcare managers should focus on enhancing their technical competencies, training the staff, and enhancing their emotional intelligence and teamwork endurance. This can be achieved through training in specific areas, organizational stress management seminars, and improvements in psychological safety. Moreover, the study highlights the importance of frequent evacuations and exercises in enhancing preparedness levels, as revealed during the study. Managers must incorporate these activities into their organizations and make them standard practices for creating a competent workforce that is mentally prepared for disaster incidents. This combined approach to preparedness will go a long way toward building more effective healthcare systems that can not only handle the immediate effects of disasters but also address their long-term consequences.

These revelations could help healthcare leaders and policymakers develop strategies to enhance the sustainability and effectiveness of healthcare in targeted communities, thereby improving their ability to deal with disasters.

4. Discussion

This section explains how the present results advance current disaster-preparedness scholarship, indicating where they corroborate, refine, or challenge earlier work.

4.1. Age-Related Preparedness Patterns

Younger respondents (<20 years) reported the lowest psychological safety, EI, and overall readiness. Similar “novice-effect” trends were documented in Japanese PHCs and Malaysian district clinics) [28], where limited crisis exposure reduced self-efficacy. Conversely, our data show a clear step-wise improvement in EI and teamwork with age and tenure, mirroring the Korean national PHC survey by Kim & Lee that linked ≥10 years’ experience to superior inter-professional coordination [29]. The present study, therefore, supports prevailing evidence that experience is a protective factor while extending it to a Gulf setting where such demographic analyses were previously absent.

4.2. Occupational Differences Among Healthcare Professionals

Doctors achieved the highest median preparedness (4.0) and EI, echoing Turkish PHC findings that physicians’ triage familiarity enhances confidence [30]. Nurses displayed strong teamwork, but the highest perceived stress was congruent with Italian frontline reports of nursing strain during earthquakes [18] and Chinese COVID-19 studies. Administrative staff posted mid-range scores (3.2–3.5), consistent with Jordanian data indicating that non-clinical cadres receive fewer disaster courses [21]. Collectively, these parallels reinforce the call for role-specific curricula: technical refreshers for administrators, stress-mitigation strategies for nurses, and leadership reinforcement for physicians.

4.3. Psychosocial Drivers of Readiness

Structural equation modeling shows stress as a strong positive predictor of preparedness (β = 0.981) yet a possible well-being risk, supporting the “challenge-stress” perspective found in UAE tertiary hospitals. The robust pathway from teamwork → psychosocial readiness → psychological safety (β = 0.285) aligns with Australian emergency-department data, where cohesive teams outperformed fragmented units. No contradictory evidence emerged; instead, our findings converge with the growing consensus that psychosocial resilience is as critical as equipment or protocols.

4.4. Machine-Learning Contribution

Random Forest and ANN models ranked psychosocial readiness and EI among the top three predictors of the composite preparedness score, confirming Pakistani ML work in teaching hospitals [4]. However, our PHC-centric AUC of 0.89 exceeds the 0.78 reported in that hospital study, suggesting that psychosocial variables possess even greater predictive utility in small-facility settings. This extends ML disaster-readiness research from secondary-care domains into the community-care arena—an acknowledged gap in GCC literature.

4.5. Qualitative Sentiment and Topic Insights

Seventy-one percent of narrative statements about the new drill program were positive, a proportion comparable to Spanish PHCs implementing simulation labs [9]. Recurring negative themes—“poor communication”, “unclear protocols”—mirror U.S. H1N1 after-action reports, indicating that information flow is a universal bottleneck independent of culture or income level. These findings therefore reinforce the emphasis other scholars place on communication training, rather than contradicting them.

4.6. Regional and Global Benchmarking

Supplementary Table S1 shows Alqurayat’s median preparedness score (3.5) surpasses Jordan’s PHC average (3.1) but lags behind Italian rural clinics (3.8). This ranking supports WHO’s observation that readiness correlates with GDP and disaster-training investment, positioning Saudi PHCs in an intermediate global tier. Our cross-country comparison adds nuance to previous single-country reviews by demonstrating how contextual factors (resource allocation, prior hazard exposure) mediate preparedness outcomes.

4.7. Practical Contributions and Policy Relevance

By blending survey metrics, psychosocial scales, and ML analytics, the study offers a replicable triage tool for rapidly classifying PHC risk levels—an approach not yet reported in GCC settings. Results back integrated curricula that weave EI modules, stress-management workshops, and routine drills, echoing recommendations from European and Asian consensus statements but supplying region-specific evidence needed for Saudi and broader GCC policy adoption.

4.8. Limitations and Future Directions

While single-region sampling and cross-sectional design limit generalizability, the mixed-methods plus ML blueprint is transferable. Future research should aim to achieve the following:

• Apply the model across multiple Saudi regions and neighboring Gulf states;
• Introduce longitudinal measurement to track skill decay and psychosocial change;
• Incorporate hard performance metrics (e.g., drill completion times) and IoT-based real-time data to refine predictive accuracy.

The results presented in this study align with and address all four research objectives. RO1 was fulfilled through quantitative findings highlighting technical competencies across healthcare roles. RO2 was confirmed by identifying distinct psychosocial trends, particularly emotional intelligence and stress responses among nurses and doctors. RO3 was addressed using predictive analytics models (Random Forest and ANNs) to evaluate preparedness levels based on technical and psychosocial inputs. Finally, RO4 was achieved through thematic and sentiment analysis of open-ended responses, revealing the role of communication and perceived preparedness themes. This comprehensive alignment enhances the study’s practical relevance and confirms the robustness of the mixed-methods approach.

5. Conclusions

The research evaluated the readiness of Saudi Arabian primary healthcare center (PHC) personnel for disasters through assessments of their technical performance capabilities and psychological preparedness. The research data show that healthcare providers hold moderate to high levels of knowledge (mean: 3.9) and skills (mean: 4.0), but a significant readiness deficit exists, as only 69.5% of drill participants experienced meaningful improvement in their performance. The research analyzed how machine learning models effectively measure and forecast preparedness indicators. The Random Forest (RF) model achieved a predictive accuracy level of 0.89, as indicated by its AUC measurement. The readiness necessary for disaster response requires more than technical evaluations, as emotional intelligence (EI) and teamwork, along with stress management capabilities, form fundamental elements of an effective response strategy. The observed findings demonstrate the advantages of uniting psychosocial resilience practices with conventional training approaches within the disaster management infrastructure. The practical findings suggest that healthcare providers require ongoing disaster training in established psychosocial systems and multidisciplinary teamwork to achieve readiness objectives. Teamwork improvement, along with enhancing psychological safety, will strengthen both emergency response performance and employee sustainability. Extensive long-term research on preparedness over time would provide essential information about the results of persistent training programs. The application of deep learning and real-time predictive analytics through advanced artificial intelligence models would enhance disaster preparedness assessments. Healthcare organizations that focus on resolving these aspects will create a more durable and sustainable workforce, which delivers effective disaster responses and promotes better healthcare professional well-being in high-pressure situations. The study highlights the significance of combining technical training with psychosocial resilience strategies to strengthen disaster preparedness. With a predictive accuracy of 0.89 using Random Forest models, the findings emphasize emotional intelligence, teamwork, and psychological safety as critical components of effective disaster response. These insights support the development of sustainable training frameworks for healthcare providers operating in high-pressure emergency environments.