Rehaf Abdulrhman Areeshi¹, Renad Ismail Madeni², Saja Mohammed Al Raddad³

Introduction

Intensive care units (ICUs) represent data-rich environments where complex decisions must be made rapidly based on multiple streams of information. The high-stakes nature of these decisions, coupled with the volume and complexity of data generated in critical care settings, creates both challenges and opportunities for enhancing decision-making processes. Artificial intelligence (AI), defined broadly as the capability of machines to perform tasks that typically require human intelligence, offers potential solutions to these challenges through applications including predictive analytics, clinical decision support, resource optimization, and workflow enhancement[2].

AI applications in intensive care have developed substantially over recent years, with demonstrated benefits in several domains [1]. Predictive models for clinical deterioration have shown promise in identifying patients at risk for adverse events such as cardiac arrest, sepsis, or acute kidney injury before traditional clinical signs become apparent. Clinical decision support systems have demonstrated value in areas such as mechanical ventilation management, sedation optimization, and antimicrobial stewardship [3]. Resource allocation tools have helped optimize bed management, staffing, and equipment utilization. These applications have the potential to enhance clinical outcomes, improve resource utilization, and reduce provider cognitive burden in intensive care settings.

The Saudi healthcare system is undergoing significant transformation under Vision 2030, with emphasis on technological innovation, quality enhancement, and efficiency improvement. The National Digital Health Strategy explicitly identifies AI as a priority area for development and implementation [4]. This strategic emphasis aligns with broader national initiatives to position Saudi Arabia as a leader in technological innovation and digital transformation across sectors [5].

Despite this strategic emphasis, the actual implementation of AI in Saudi intensive care units remains variable and poorly documented. Previous research on healthcare AI in Saudi Arabia has primarily focused on technical aspects, theoretical frameworks, or single-institution experiences, with limited comprehensive assessment of implementation patterns, barriers, and facilitators across multiple centers. This knowledge gap hinders evidence-based planning and policy development to optimize AI integration into critical care practice [6].

The implementation of AI in healthcare settings involves complex interactions between technology, clinical workflows, organizational structures, and human factors. International experience suggests that successful implementation requires attention to data infrastructure, staff engagement and training, workflow integration, ethical frameworks, and sustainability planning. These factors likely operate in the Saudi context as well, but potentially with unique manifestations reflecting the specific characteristics of the Saudi healthcare system, workforce composition, and cultural environment.

This study aims to address the knowledge gap by providing a comprehensive assessment of AI implementation in Saudi intensive care units [7]. Specifically, we seek to: 1. Document the current status of AI implementation across Saudi ICUs 2. Identify technical, organizational, and cultural barriers to implementation 3. Determine factors associated with successful implementation 4. Develop a framework for effective AI integration in Saudi critical care

The findings will inform healthcare administrators, policymakers, and clinicians seeking to implement or optimize AI applications in intensive care settings within Saudi Arabia and similar healthcare contexts.

Figure 1: Application of AI in Intensive Care1

Methodology

Study Design and Setting

This study employed a sequential explanatory mixed-methods design, combining quantitative and qualitative approaches to provide a comprehensive assessment of AI implementation in Saudi intensive care units. The study was conducted between September 2024 and January 2025, following approval from the Institutional Review Board of King Faisal University (approval number: KFU-IRB-2024-09-08).

The quantitative phase consisted of a cross-sectional survey of ICU directors across Saudi Arabia. The qualitative phase involved detailed case studies of five ICUs with established AI applications. This mixed-methods approach allowed for triangulation of findings and provided complementary insights into the complex phenomenon of AI implementation in critical care.

Survey Development and Distribution

The survey instrument was developed based on a comprehensive literature review and consultation with experts in critical care, health informatics, and implementation science. The initial questionnaire was reviewed by a panel of seven experts, and content validity was assessed using the Content Validity Index (CVI). Items with CVI scores below 0.78 were revised or eliminated. The final questionnaire was pilot-tested in five ICUs not included in the main study to assess clarity, comprehensibility, and time required for completion.

The final survey consisted of five sections: 1. Hospital and ICU characteristics (10 items) 2. Current status of AI implementation (15 items) 3. Applications and utilization patterns (12 items) 4. Barriers to implementation (15 items on a 5-point Likert scale) 5. Facilitators and success factors (15 items on a 5-point Likert scale)

The survey was distributed electronically to the medical directors of 55 intensive care units across Saudi Arabia, representing different geographic regions, hospital types (government, private, academic), and levels of care (secondary, tertiary). Three reminder emails were sent at two-week intervals to maximize response rate.

Table 1: Summary of the survey development and distribution process

Case Study Selection and Data Collection

For the qualitative phase, we conducted detailed case studies of five intensive care units that had successfully implemented AI applications. These sites were selected using purposive sampling to ensure diversity in terms of: - Geographic location (different regions of Saudi Arabia) - Hospital type (academic, government, private) - AI application type (predictive analytics, decision support, resource optimization) - Implementation maturity (recently implemented vs. established programs)

Data collection for each case study included: 1. Semi-structured interviews with key stakeholders (ICU director, AI implementation lead, physicians, nurses, IT staff, administrators) 2. Document review (implementation plans, protocols, evaluation reports, usage statistics) 3. Direct observation of AI applications in clinical use

The interview guide was developed based on preliminary survey findings and included open-ended questions exploring implementation processes, barriers, facilitators, success factors, and lessons learned. Interviews were conducted in person or via videoconference, lasted 45-60 minutes, and were audio-recorded with participant consent. All interviews were conducted in English or Arabic based on participant preference.

Table 2: Summary of the Case Study Selection and Data Collection process

Data Analysis Methods

Quantitative Analysis

Survey data were analyzed using SPSS version 28.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics were used to summarize hospital characteristics, implementation status, and application patterns. Continuous variables were presented as means and standard deviations, while categorical variables were presented as frequencies and percentages.

Barriers and facilitators were ranked based on mean scores, and factor analysis was conducted to identify underlying dimensions. Principal component analysis with varimax rotation was used, and factors with eigenvalues greater than 1 were retained.

Bivariate analyses including chi-square tests, t-tests, and one-way ANOVA were conducted to examine associations between hospital characteristics and implementation status, barriers, and facilitators. Multivariate logistic regression was performed to identify independent predictors of active AI implementation, controlling for potential confounders. Statistical significance was set at p<0.05 for all analyses.

Qualitative Analysis

Interview recordings were transcribed verbatim, and Arabic interviews were translated into English by bilingual researchers. Transcripts were analyzed using thematic analysis following Braun and Clarke’s six-step approach: familiarization with data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, and producing the report.

Two researchers independently coded the transcripts using NVivo 14 (QSR International, Melbourne, Australia). Initial codes were compared and discussed to develop a coding framework, which was then applied to all transcripts. Emerging themes were reviewed and refined through iterative discussion among the research team until consensus was reached.

Case study data were analyzed using both within-case and cross-case approaches. Within-case analysis focused on understanding the unique implementation journey of each site, while cross-case analysis identified common patterns, success factors, and challenges across sites.

Integration of Findings

Quantitative and qualitative findings were integrated using a weaving approach, where results from both methods were presented together by theme rather than by method. This integration allowed for a comprehensive understanding of AI implementation in Saudi ICUs, with qualitative findings providing context and depth to quantitative results.

Results

Current Implementation Status Across Saudi ICUs

Of the 55 ICU directors invited to participate, 40 completed the survey (response rate: 72.7%). Participating ICUs represented diverse settings, with 30.0% from academic medical centers, 52.5% from government hospitals, and 17.5% from private facilities. The majority (75.0%) were in urban areas, while 25.0% were in rural or semi-urban settings. Regarding level of care, 67.5% were tertiary and 32.5% were secondary care facilities.

The current status of AI implementation varied significantly across the surveyed ICUs (Figure 1). Among the 40 ICUs, 7 (17.5%) reported active implementation of at least one AI application, 13 (32.5%) were in planning or early implementation stages, and 20 (50.0%) had no immediate plans for AI implementation. Among ICUs with active implementation, the median duration of implementation was 1.8 years (range: 0.5-3.5 years).

Implementation status was significantly associated with hospital type (p=0.008) and level of care (p=0.003). Academic medical centers had the highest rate of active implementation (41.7%), followed by government hospitals (9.5%) and private hospitals (0.0%). Similarly, tertiary care facilities had higher rates of active implementation (25.9%) compared to secondary care facilities (0.0%).

Among ICUs with active implementation (n=7), the most common AI applications included predictive analytics for clinical deterioration (85.7%), medication error prevention (71.4%), resource optimization (57.1%), ventilator management support (42.9%), and sepsis detection (42.9%). The specific predictive models most frequently implemented were for acute kidney injury (83.3% of those using predictive analytics), cardiac arrest (66.7%), and ventilator-associated events (50.0%).

Regarding integration with existing systems, 71.4% of AI implementations were integrated with electronic health records, 57.1% with clinical decision support systems, and 42.9% with monitoring devices. The majority (57.1%) were commercial solutions from established vendors, while 28.6% were developed through academic-industry partnerships and 14.3% were developed in-house.

Table 3: Summary the Current Implementation Status Across Saudi ICUs

Application Types and Their Clinical Impact

The survey assessed perceived clinical impact of AI applications on a 5-point Likert scale (1=no impact, 5=significant impact). Among ICUs with active implementation, the highest-rated applications in terms of perceived impact were predictive analytics for acute kidney injury (mean score: 4.2±0.8), medication error prevention (4.0±0.9), and sepsis detection (3.8±1.0).

Case studies provided deeper insights into specific applications and their clinical impact. At Site A, an academic medical center in the Central region, a predictive model for acute kidney injury had been implemented for 3.5 years. The model analyzed laboratory values, vital signs, medication data, and comorbidities to identify patients at risk for AKI up to 48 hours before clinical manifestation. Implementation was associated with a 28% reduction in stage 3 AKI and a 15% reduction in renal replacement therapy initiation:

“The AKI prediction model has transformed our approach to kidney injury in the ICU. Before, we were often reacting to established AKI; now, we can intervene before significant damage occurs. This proactive approach has led to earlier nephrology consultation, more judicious use of nephrotoxic medications, and improved fluid management strategies.” (ICU Director, Site A)

At Site B, a government hospital in the Western region, an AI-based medication safety system had been implemented for 2.5 years. The system analyzed medication orders, patient characteristics, and laboratory values to identify potential errors or adverse drug events. Implementation was associated with a 35% reduction in medication-related adverse events:

“The medication safety system has significantly enhanced our ability to prevent medication errors. It identifies potential issues that might be missed during busy clinical workflows, such as dose adjustments needed for changing renal function or drug interactions that aren’t immediately obvious. This has been particularly valuable during night shifts and high-stress periods when cognitive overload is more likely.” (Clinical Pharmacist, Site B)

At Site C, a private hospital in the Eastern region, an AI-based resource optimization system had been implemented for 1.5 years. The system predicted ICU admission needs, length of stay, and resource requirements based on historical patterns and current hospital status. Implementation was associated with improved bed utilization and reduced emergency department boarding times:

“The resource optimization system has enhanced our ability to plan staffing, allocate beds, and manage patient flow. By predicting likely discharges and new admissions 24-48 hours in advance, we can better prepare for capacity challenges and avoid situations where emergency admissions are delayed due to resource constraints.” (ICU Nurse Manager, Site C)

Technical, Organizational, and Cultural Barriers

Respondents rated various barriers to AI implementation on a 5-point Likert scale (1=not a barrier, 5=major barrier). The highest-rated barriers are presented in Table 1. Data infrastructure limitations received the highest mean score (4.3±0.8), with 82.5% of respondents rating it as a significant or major barrier. Other prominent barriers included staff training needs (4.2±0.9, 77.5%), integration with existing systems (4.0±1.0, 72.5%), concerns about algorithm transparency (3.8±1.1, 67.5%), and cost considerations (3.7±1.2, 65.0%).

Factor analysis of barrier items identified four underlying dimensions: technical barriers (explaining 30.5% of variance), organizational barriers (explaining 25.3% of variance), human factors barriers (explaining 22.8% of variance), and ethical/legal barriers (explaining 12.4% of variance).

Table 4: Summary of the Technical, Organizational, and Cultural Barriers to AI Implementation

Case studies provided deeper insights into these barriers. Regarding data infrastructure, participants highlighted challenges with data quality and integration:

“One of our biggest challenges was the quality and consistency of our historical data. We discovered significant variations in documentation practices, missing data points, and inconsistent terminology that complicated model development and validation. Substantial data cleaning and standardization were required before we could proceed with implementation.” (Data Scientist, Site D)

Staff engagement and training emerged as significant challenges:

“Initially, there was skepticism among clinical staff about the value and reliability of AI predictions. Some viewed it as a threat to clinical judgment, while others were concerned about additional alert burden. Developing trust in the system required transparent communication about how the algorithms worked, their limitations, and how they should complement rather than replace clinical decision-making.” (ICU Director, Site B)

Integration with existing workflows was frequently mentioned:

“The technical integration with our electronic health record was challenging, but the workflow integration was even more complex. We had to carefully consider how AI-generated insights would be communicated to clinicians, who would be responsible for responding, and how responses would be documented and evaluated. Poor workflow integration would have resulted in alert fatigue and system avoidance.” (Implementation Lead, Site A)

Ethical and legal considerations also influenced implementation:

“There were significant concerns about liability and responsibility when clinical decisions were influenced by AI recommendations. Clarifying that AI tools were decision support rather than decision-making systems was essential, as was developing clear protocols for how to document the use of AI insights in clinical decision-making.” (Legal Advisor, Site E)

Success Factors and Implementation Strategies

Respondents rated various facilitators of successful AI implementation on a 5-point Likert scale (1=not important, 5=very important). The highest-rated facilitators are presented in Table 2. Leadership commitment received the highest mean score (4.5±0.7), with 82.5% rating it as important or very important. Other key facilitators included dedicated AI champions (4.3±0.8, 77.5%), clear clinical use cases (4.2±0.8, 75.0%), phased implementation approach (4.1±0.9, 72.5%), and robust data governance (4.0±1.0, 70.0%).

Multivariate logistic regression analysis identified several factors independently associated with active AI implementation. These included leadership commitment (OR=3.6, 95% CI: 1.8-7.2, p<0.001), dedicated AI champions (OR=2.9, 95% CI: 1.5-5.8, p=0.002), phased implementation approaches (OR=2.5, 95% CI: 1.3-4.9, p=0.007), and academic affiliation (OR=2.3, 95% CI: 1.2-4.5, p=0.014).

Case studies revealed several successful implementation strategies. A problem-first rather than technology-first approach was emphasized:

“Our successful implementations started with a clearly defined clinical problem that AI could potentially address, not with a technology looking for an application. This ensured relevance and facilitated clinician buy-in.” (ICU Director, Site A)

Table 5: Key Facilitators of AI Implementation in ICUs

Table 6: Independent Predictors of Active AI Implementation

Engaging frontline clinicians throughout the development and implementation process was consistently highlighted:

“We established a clinical advisory group that participated in every phase from initial needs assessment through post-implementation evaluation. Their input shaped the user interface, alert thresholds, and workflow integration, which was crucial for adoption.” (Implementation Lead, Site B)

A phased implementation approach was universally employed:

“We started with a limited pilot in one pod of the ICU, refined based on feedback, then gradually expanded. This allowed us to identify and address issues at a manageable scale before wider deployment.” (Project Manager, Site C)

Robust evaluation frameworks were established from the outset:

“We defined clear metrics for technical performance, clinical outcomes, workflow impact, and user satisfaction before implementation. This provided objective data to demonstrate value and guide refinements.” (Quality Improvement Director, Site D)

Continuous education and support were emphasized:

“We developed a multi-modal education program including workshops, simulation sessions, at-elbow support, and just-in-time resources. This ongoing support was critical for maintaining engagement and addressing evolving needs.” (Clinical Educator, Site A)

Discussion

This study provides the first comprehensive assessment of AI implementation in Saudi intensive care units, revealing both the current landscape and a path forward for effective integration [8]. The findings highlight significant variability in implementation status, with academic and tertiary care centers leading adoption. This pattern reflects a common trajectory for healthcare innovations, where academic centers often serve as early adopters due to their research orientation, educational mission, and typically greater resources for innovation.

The predominance of predictive analytics applications, particularly for clinical deterioration events like acute kidney injury and cardiac arrest, aligns with international trends in critical care AI [9]. These applications address high-priority clinical challenges where early intervention can significantly impact outcomes, and where traditional approaches to risk identification may be suboptimal due to the complexity and volume of relevant data.

The barriers identified in this study—data infrastructure limitations, staff training needs, integration challenges, algorithm transparency concerns, and cost considerations—echo findings from healthcare AI implementation studies in other settings [10]. However, the specific manifestations of these barriers reflect the unique Saudi healthcare context. For example, data infrastructure challenges may be more pronounced in certain regions or facility types, while staff training needs may be influenced by the multinational composition of the Saudi healthcare workforce.

The success factors identified—leadership commitment, dedicated champions, clear clinical use cases, phased implementation, and robust data governance—provide valuable guidance for implementation efforts [11]. The critical importance of leadership commitment aligns with broader literature on healthcare innovation, which consistently identifies leadership as a key determinant of successful change. The significance of dedicated champions highlights the value of individuals who can bridge clinical and technical domains, advocate for the innovation, and support colleagues through the adoption process [12].

The implementation strategies revealed through our case studies offer practical insights for intensive care units considering AI adoption. The problem-first approach ensures that AI applications address genuine clinical needs rather than representing technology in search of an application [13]. This approach aligns with user-centered design principles and increases the likelihood of meaningful impact and sustained adoption. Similarly, the phased implementation approach allows for iterative refinement and manageable change, reducing the risks associated with large-scale technology deployment.

Based on our findings, we propose a framework for effective AI integration in Saudi intensive care units. This framework emphasizes five interconnected domains: needs assessment, data infrastructure, workforce preparation, implementation approach, and evaluation [14]. Within each domain, specific strategies are recommended based on the identified barriers and facilitators.

For needs assessment, we recommend a systematic approach to identifying high-priority clinical challenges amenable to AI solutions, considering both local needs and organizational readiness [15]. For data infrastructure, we advocate for investment in data quality, integration capabilities, and governance structures as foundational elements for successful AI implementation.

Regarding workforce preparation, our findings support a comprehensive approach including education about AI capabilities and limitations, training in critical appraisal of AI outputs, and attention to the human-AI interface [16]. For implementation approach, we recommend a phased strategy with pilot testing, iterative refinement, and gradual expansion, coupled with clear protocols for how AI insights should influence clinical decision-making.

For evaluation, we emphasize the importance of comprehensive assessment across multiple dimensions, including technical performance, clinical outcomes, workflow impact, and user experience [20]. This multifaceted evaluation provides a more complete understanding of value and guides ongoing refinement.

This framework aligns with the objectives of Saudi Vision 2030 and the National Digital Health Strategy, suggesting potential synergies between AI implementation in critical care and broader healthcare transformation efforts [17]. The emphasis on data infrastructure development, workforce capacity building, and evidence-based implementation particularly resonates with national strategic priorities.

This study has several strengths, including its mixed-methods design, diverse sample of intensive care units, and detailed case studies of successful implementations. However, certain limitations should be acknowledged [19]. First, despite efforts to ensure representativeness, the sample may not fully reflect all Saudi intensive care units, particularly those in very remote areas. Second, the cross-sectional design provides a snapshot of implementation status but cannot capture the dynamic nature of implementation over time [18]. Third, self-reported data may be subject to social desirability bias, although anonymity was maintained to minimize this effect.

Conclusion

This survey manuscript reports varied AI implementation in participating ICUs. A single cross-sectional survey does not establish national trends or growth; site and clinical outcome records require verification.

The reported barriers and facilitators reflect respondents’ accounts. Associations from the small sample cannot establish effective implementation determinants.

The proposed framework is exploratory and requires independent assessment, with appropriate clinical safety evaluation and local authorization.

Future research should explore the long-term impacts of AI implementation on clinical outcomes, cost-effectiveness, and staff experience in Saudi intensive care units. Additionally, comparative studies examining different implementation approaches would provide valuable insights into optimal strategies for the Saudi healthcare context. As AI technologies continue to evolve, ongoing assessment of implementation status, barriers, and facilitators will be essential to guide evidence-based integration into critical care practice.

References

Al-Sofyani, K. A. (2025). Role of artificial intelligence in pediatric intensive care: a survey of healthcare staff perspectives in Saudi Arabia. Frontiers in Pediatrics, 13, 1533877.

Kumar, R., Singh, A., Kassar, A. S. A., Humaida, M. I., Joshi, S., & Sharma, M. (2025). Leveraging Artificial Intelligence to Achieve Sustainable Public Healthcare Services in Saudi Arabia: A Systematic Literature Review of Critical Success Factors. Computer Modeling in Engineering & Sciences (CMES), 142(2).

Alanazi, A., Aldakhil, L., Aldhoayan, M., & Aldosari, B. (2023). Machine learning for early prediction of sepsis in Intensive Care Unit (ICU) Patients. Medicina, 59(7), 1276.

Moafa, K. M. Y., Almohammadi, N. F. H., Alrashedi, F. S. S., Alrashidi, S. T. S., Al-Hamdan, S. A., Faggad, M. M., ... & Al-Anzi, A. K. (2024). Artificial Intelligence for Improved Health Management: Application, Uses, Opportunities, and Challenges-A Systematic Review. Egyptian Journal of Chemistry, 67(13), 865-880.

Almutairi, H. N., Mansouri, R. M., Alsulami, S. K., & Alsolmi, B. R. (2024). Effectiveness of Artificial Intelligence Applications in Nursing Practice and Workflow: An Integrative Literature Review. Journal of International Crisis & Risk Communication Research (JICRCR), 7(3).

Al Harbi, M., Alotaibi, A., Alanazi, A., Alsughayir, F., Alharbi, D., Bin Qassim, A., ... & Alsabani, M. (2024). Perspectives toward the application of Artificial Intelligence in anesthesiology‐related practices in Saudi Arabia: A cross‐sectional study of physicians views. Health Science Reports, 7(10), e70099.

Hassanein, S., El Arab, R. A., Abdrbo, A., Abu-Mahfouz, M. S., Gaballah, M. K. F., Seweid, M. M., ... & Alzghoul, H. (2025). Artificial intelligence in nursing: an integrative review of clinical and operational impacts. Frontiers in Digital Health, 7, 1552372.

Alabdulhafith, M., Saleh, H., Elmannai, H., Ali, Z. H., El-Sappagh, S., Hu, J. W., & El-Rashidy, N. (2023). A clinical decision support system for edge/cloud ICU readmission model based on particle swarm optimization, ensemble machine learning, and explainable artificial intelligence. IEEE Access, 11, 100604-100621.

Alafif, T., Tehame, A. M., Bajaba, S., Barnawi, A., & Zia, S. (2021). Machine and deep learning towards COVID-19 diagnosis and treatment: survey, challenges, and future directions. International journal of environmental research and public health, 18(3), 1117.

Alafif, T., Tehame, A. M., Bajaba, S., Barnawi, A., & Zia, S. (2021). Machine and deep learning towards COVID-19 diagnosis and treatment: survey, challenges, and future directions. International journal of environmental research and public health, 18(3), 1117.

Justinia, T. (2022). Transforming health care in Saudi Arabia. Nursing informatics. New York, NY, USA.

Aleid, A. M., Alrasheed, A. S., Aldanyowi, S. N., & Almalki, S. F. (2024). Advanced magnetic resonance imaging for glioblastoma: Oncology-radiology integration. Surgical Neurology International, 15, 309.

Aleid, A. M., Aleid, M. S., Alzimami, L. Z., Ajwah, Y. M. H., Alamrani, F. M. O., Alshehri, R. F. Z., & Kharaba, A. (2024). Impact of Mechanical Ventilation on COVID-19 Patients in the Intensive Care Unit: A Retrospective Study. Journal of Advanced Trends in Medical Research, 1(2), 680-685.

Aleid, A. M., Saeed, H. S., Aldanyowi, S. N., Albinsaad, L., Alessa, M., AlAidarous, H., ... & Almutair, A. (2024). Efficacy of high-dose versus low-dose tranexamic acid for reduction of blood loss in adolescent idiopathic scoliosis surgery: A systematic review and meta-analysis. Surgical Neurology International, 15, 473.

Aleid, A. M., Alshehri, F., Alasiri, N., Alhomoud, F., Alsaegh, S., Alrasheed, M., ... & Aldanyowi, S. N. (2025). Efficacy of Duloxetine for Postspine Surgery Pain: A Systematic Review and Meta‐Analysis. Brain and Behavior, 15(1), e70217.

Alrasheed, A. S., Alqadhibi, M. A., Khoja, R. H., Alayyaf, A. S., Alhumoudi, D. S., Aldawlan, M. I., ... & Alessa, A. A. (2024). Emerging therapies for immunomodulation in traumatic brain injury: A systematic review and meta-analysis. Surgical Neurology International, 15, 327.

Aleid, A. M., Al-Ghanim, H. A., Aldanyowi, S. N., Albinsaad, L. S., Alessa, M. Y., & Alhussain, A. A. (2025). A comparative meta-analysis between mini-transverse and longitudinal techniques in the management of carpal tunnel syndrome. Surgical Neurology International, 16(78), 1.

Aleid, A. M., Aljabr, A. A., Aldanyowi, S. N., AlAidarous, H. A., Aleid, Z. M., Alharthi, A. S., ... & Al Mutair, A. (2025). Dry needling for mechanical neck pain: A systematic review and meta-analysis of randomized controlled trials. Surgical Neurology International, 16, 44.

Alrasheed, A. S., Aleid, A. M., Alharbi, R. A., Alamer, M. A., Alomran, K. A., Maan, S. A. B., & Almalki, S. F. (2025). Stereotactic radiosurgery versus whole-brain radiotherapy for intracranial metastases: A systematic review and meta-analysis. Surgical Neurology International, 16, 18.

Aleid, A. M., Aldanyowi, S. N., Aleid, A. J., Alessa, A. A., Alhussain, A. A., Albinsaad, L. S., & Al Mutair, A. S. (2024). Which is better in the management of chronic subdural hematoma: Irrigation, or no irrigation? A systematic review and meta-analysis of randomized controlled trials. Surgical Neurology International, 15, 435.