Challenges, Biases, and Limitations in Systematic Reviews and Meta-Analyses: Addressing Pitfalls in Medical Research Methodologies

Abstract

Systematic reviews and meta-analyses represent crucial tools in medical research, synthesizing vast arrays of data to inform clinical and policy decisions. While their merits are undeniable, potential pitfalls, biases, and limitations threaten the integrity of these methodologies. This paper explores common challenges faced in the execution and interpretation of systematic reviews and meta-analyses, emphasizing strategies to enhance the robustness and reliability of future research.

Introduction

I. INTRODUCTION

Systematic reviews and meta-analyses have risen to prominence as cornerstones of evidence-based medical research over the last few decades. Their capacity to collate, synthesize, and evaluate vast quantities of research data allows them to serve as invaluable tools for healthcare professionals, policymakers, and researchers. By combining the findings of multiple studies, these methodologies can provide more comprehensive and, arguably, more objective insights into a particular research question, often with greater statistical power than individual studies. However, like all scientific tools, they are not without their challenges.

The importance of systematic reviews and meta-analyses in the broader medical landscape cannot be overstated. In a rapidly evolving field like medicine, where vast amounts of research are generated at an ever-increasing pace, the ability to discern the overall 'message' from a plethora of individual studies is crucial. These reviews provide summarized evidence for clinicians, allowing for better-informed patient care decisions. For policymakers, they inform guidelines and protocols, ensuring that health systems deliver care based on the best available evidence. And for researchers, they highlight gaps in the current literature, guiding the direction of future studies.

Given the weight and influence that systematic reviews and meta-analyses often carry, their integrity and accuracy are paramount. If conducted poorly or with bias, the consequences can be far-reaching, potentially leading to misguided clinical decisions or ineffective policy directives. The foundational promise of these methodologies is their systematic and comprehensive approach, theoretically minimizing bias and maximizing the validity of the conclusions drawn. However, this promise can only be realized if researchers are acutely aware of the inherent challenges and take proactive steps to mitigate them.

One of the primary challenges faced in conducting systematic reviews and meta-analyses is the vast and diverse nature of medical literature. With countless studies published across various journals, languages, and regions, ensuring a truly comprehensive review is an arduous task. Moreover, the variations in study design, populations, interventions, and outcomes add layers of complexity to the synthesis process. Thus, while the goal is to obtain a unified perspective, the diverse nature of the primary research can sometimes yield more questions than answers. Another critical concern is the presence of biases. From publication and selection biases to data extraction biases, these can insidiously infiltrate the review process, compromising the validity of the results. It's important to recognize that while systematic reviews aim to be objective, they are not conducted in a vacuum. The choices made by reviewers at various stages, be it in the literature search strategy, the inclusion or exclusion of certain studies, or the statistical methods used, can all introduce elements of subjectivity.

Finally, while systematic reviews and meta-analyses strive for comprehensiveness, they are, by nature, retrospective. They rely on existing research, which means they are bound by the limitations of the primary studies they evaluate. If the original research was flawed or biased, those issues could be carried forward and amplified in the review.

In recent years, medical research landscape has seen an unprecedented surge in volume and complexity. This surge can be attributed to numerous factors, including advancements in technology, increased funding, and a heightened global emphasis on healthcare outcomes.

IV. ADDRESSING PITFALLS: BEST PRACTICES

A. Comprehensive Literature Search

A comprehensive literature search is a cornerstone of systematic reviews and meta-analyses. By ensuring that all relevant studies, irrespective of their results, are identified and considered for inclusion, this process reduces the risk of selection bias and increases the validity and generalizability of the findings. The objective is to capture as complete a picture as possible of the available evidence on the research question of interest.

Steps in a Comprehensive Literature Search:

1. Define the Research Question: Clearly specify the research question using frameworks like PICO (Population, Intervention, Comparator, Outcome) to ensure clarity about what you are looking for.
2. Identify Relevant Databases and Registers: Based on the topic, choose appropriate databases. Common choices include:

• PubMed/MEDLINE
• EMBASE
• Cochrane Central Register of Controlled Trials (CENTRAL)
• Web of Science
• PsycINFO (for psychological studies)

3. Develop a Search Strategy:

• Use relevant keywords, medical subject headings (MeSH), and Boolean operators (AND, OR, NOT).
• Be comprehensive in keyword choice, including synonyms, abbreviations, and alternative spellings.
• Consider consulting a librarian or expert in the field to refine the search strategy.

4. Hand Searching: Manually review the reference lists of included studies or relevant reviews to identify additional studies missed in the database search.

5. Gray Literature Search: Gray literature refers to materials not published in traditional academic journals. This might include:

• Theses and dissertations
• Conference abstracts and proceedings
• Clinical trial registries (e.g., ClinicalTrials.gov) to identify unpublished or ongoing studies

6. Search Updates: Given the continuous publication of new studies, update the literature search periodically, especially if there's a significant time lag between the initial search and publication of the systematic review.

7. Document the Process: For transparency and reproducibility:

• Record databases searched, search terms used, and the number of hits.
• Use PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagrams to depict the study selection process.

8. Screening and Selection: Titles and abstracts are screened to eliminate irrelevant studies, followed by a full-text review to determine final inclusion. This process often involves multiple reviewers to minimize individual bias.

9. Inclusion and Exclusion Criteria: Clearly define criteria for study inclusion/exclusion based on study design, population, interventions, comparators, outcomes, and other relevant factors.

Challenges:

• Language Barriers: Some relevant studies might be published in languages other than English.
• Publication Bias: Positive results are more likely to be published, so it's essential to find unpublished studies or those with negative or neutral results.
• Database Limitations: Not all databases index every journal, so relevant studies might be missed if the search isn't broad enough.

For example, suppose a researcher is interested in the effects of yoga on chronic lower back pain. A comprehensive literature search would involve:

• Formulating a clear PICO question.
• Searching in relevant databases using keywords like "yoga," "chronic lower back pain," "pain relief," "RCT," and their synonyms/MeSH terms.
• Hand-searching references of relevant articles and systematic reviews on the topic.
• Exploring conference proceedings related to alternative therapies or musculoskeletal disorders.
• Documenting the entire process, screening results, and making final selections transparently.

A comprehensive literature search is pivotal in ensuring the robustness and credibility of systematic reviews and meta-analyses. It involves a systematic, exhaustive, and transparent approach to identifying all relevant studies on the topic, minimizing biases and maximizing the scope and depth of the review.

B. Transparent Inclusion/Exclusion Criteria

The establishment of clear and transparent inclusion and exclusion criteria is a fundamental step in systematic reviews and meta-analyses. These criteria determine which studies will be considered for inclusion in the review and which will be excluded, ensuring consistency and reducing bias in the selection process. By being explicit about these criteria, researchers provide a transparent blueprint for their study selection, allowing for reproducibility and critique of their methodological choices.

1. Purpose of Inclusion/Exclusion Criteria:

a. Reduce Bias: Systematic and transparent criteria ensure that study selection is based on predefined and objective standards rather than subjective judgment.

b. Ensure Relevance: The criteria ensure that only studies relevant to the research question are included.

c. Allow Reproducibility: Other researchers can reproduce the systematic review or meta-analysis using the same criteria.

d. Enhance Clarity: Clearly defining the scope of the review aids readers in understanding the context and applicability of the findings.

2. Components of Inclusion/Exclusion Criteria:

Often based on the PICO framework, the criteria typically cover:

a. Population: Specify characteristics of the study population, such as age, gender, diagnosis, or other relevant factors.

b. Intervention: Clearly define the intervention or exposure of interest. This could be a drug, therapy, procedure, risk factor, etc.

c. Comparator: Specify if studies must have a specific comparator (e.g., placebo, standard care) or if no comparator is required.

d. Outcomes: State the primary and secondary outcomes of interest. For instance, in a review about treatments for depression, the outcome might be symptom reduction measured by a specific scale.

e. Study Design: Determine which types of study designs will be considered. Common choices include randomized controlled trials (RCTs), cohort studies, case-control studies, etc.

f. Publication Date: If the review focuses on recent advancements, a date range might be specified.

g. Language: State if there are language restrictions, though limiting to certain languages might introduce bias.

3. Examples of Inclusion/Exclusion Criteria:

Consider a systematic review examining the effects of aerobic exercise on cognitive function in elderly individuals:

a. Inclusion Criteria:

• Population: Individuals aged 65 and above.
• Intervention: Aerobic exercise interventions (e.g., walking, jogging, cycling).
• Comparator: Non-exercise groups or different exercise modalities.
• Outcomes: Measures of cognitive function, such as memory, attention, or executive function tests.
• Study Design: RCTs and cohort studies.
• Publication Date: Studies published from 2000 to 2023.
• Language: English and Spanish.

b. Exclusion Criteria:

• Studies that include participants with neurodegenerative diseases.
• Studies where the primary intervention is not aerobic in nature (e.g., strength training).
• Case reports, case series, or cross-sectional designs.
• Studies without a clear measure of cognitive function.

c. Challenges and Considerations:

• Overly Narrow Criteria: While specificity is vital, excessively narrow criteria might exclude relevant studies, leading to a less comprehensive review.
• Potential for Bias: If criteria are set post hoc or after preliminary exploration of the available literature, there's a risk of biasing the review towards certain findings.
• Updating Criteria: As the field advances, previously set criteria might need updating in subsequent reviews.

A transparent inclusion/exclusion criteria are crucial for the methodological rigor and validity of systematic reviews and meta-analyses. They provide a roadmap for the systematic identification and selection of relevant studies, ensuring that the review's findings are based on a comprehensive and unbiased assessment of the available evidence.

C. Standardized Data Extraction

In the context of systematic reviews and meta-analyses, standardized data extraction is the process of systematically and consistently gathering relevant information from the included studies. This step ensures that the data used for synthesizing evidence and drawing conclusions are accurate, complete, and comparable across studies. A standardized approach minimizes errors, reduces potential biases, and enhances the reproducibility and credibility of the findings.

1. Importance of Standardized Data Extraction:

a. Minimize Bias: A systematic and consistent approach prevents selective extraction of data that might favor a particular outcome.

b. Ensure Completeness: Ensures that all pertinent information is captured from each study.

c. Facilitate Synthesis: Provides a structured dataset ready for statistical analysis or qualitative synthesis.

d. Enable Verification: Allows other researchers to verify findings by following the same extraction procedures.

2. Key Components of Standardized Data Extraction:

a. Data Extraction Form/Tool: This is a predefined template or digital tool that guides the extraction process. The form usually includes:

• Bibliographic details (e.g., authors, publication year)
• Study characteristics (e.g., design, sample size, setting)
• Participant demographics (e.g., age, gender, health status)
• Details of the intervention and comparator
• Outcomes and measurement instruments
• Results (e.g., effect sizes, confidence intervals, p-values)

b. Training and Calibration: Reviewers should be trained on the extraction process. In larger teams, calibration exercises can ensure consistency in extraction among reviewers.

c. Dual Extraction: Ideally, two independent reviewers should extract data from each study. Discrepancies between reviewers are then identified and resolved, either through discussion or consultation with a third reviewer. This approach reduces errors and biases.

d. Pilot Testing: Before extracting data from all studies, the extraction form/tool should be pilot-tested on a few studies to identify potential challenges and refine the process.

e. Document Decision Rules: Clearly define rules for handling missing data, extracting data from graphs, or dealing with multiple publications from the same study.

f. Contact Authors: If essential data are missing or unclear, consider contacting study authors for clarification or additional information.

3. Examples of Standardized Data Extraction:

Consider a systematic review assessing the impact of a specific diet on blood pressure reduction. The data extraction form might include:

a. Bibliographic Information: Author names, publication year, journal name.

b. Study Design: e.g., Randomized Controlled Trial, cohort study.

c. Sample Size: Total participants, number in intervention, and control groups.

d. Participant Details: Age range, gender distribution, baseline blood pressure.

e. Intervention Details: Specifics of the diet, duration, compliance measures.

f. Comparator Details: Type of control (e.g., standard diet, placebo).

g. Outcome Measures: Instruments or methods used to measure blood pressure.

h. Results: Mean blood pressure reduction, standard deviation, p-values.

4. Challenges in Data Extraction:

a. Incomplete or Ambiguous Data: Not all studies provide clear or complete data, making extraction challenging.

b. Variability in Reporting: Different studies may report results in diverse formats or use different measurement units.

c. Subjectivity: Some data, especially in qualitative studies, may be open to interpretation, leading to potential discrepancies among reviewers.

In conclusion, standardized data extraction is a critical step in systematic reviews and meta-analyses, ensuring that evidence synthesis is based on accurate, consistent, and comprehensive data [28, 29]. By following a systematic approach, researchers can enhance the quality, credibility, and reproducibility of their findings, facilitating evidence-based decision-making in the respective fields.

D. Quality Assessment

Quality assessment, often termed 'risk of bias assessment', is an essential step in systematic reviews and meta-analyses. It involves a thorough evaluation of the methodological quality of the included studies to gauge the likelihood of bias that might affect the study results. By assessing study quality, researchers can understand the strength and reliability of the evidence presented, guiding more informed interpretations and conclusions.

1. Importance of Quality Assessment:

a. Assess Credibility: Understanding study quality aids in determining how much weight or trust to place in the findings.

b. Interpret Results: Allows for nuanced interpretations based on the quality of the studies, rather than purely on statistical outcomes.

c. Guide Synthesis: Helps in making decisions about study weighting in meta-analyses or excluding studies of extremely poor quality.

d. Inform Recommendations: A systematic review's conclusions and recommendations are bolstered when based on high-quality studies.

e. Key Aspects of Quality Assessment:

f. Randomization: Were participants randomly assigned to intervention and control groups? Proper randomization minimizes confounding.

g. Allocation Concealment: Were participants and investigators unaware of the group assignments in advance?

h. Blinding: Were participants, care providers, and evaluators blind to group assignments? Blinding can reduce performance and detection biases.

i. Incomplete Data: Were there any dropouts or missing data, and were they addressed appropriately (e.g., intention-to-treat analysis)?

j. Selective Reporting: Were all pre-specified outcomes reported, or were some omitted based on the results?

k. Other Biases: Consider other potential sources of bias relevant to the specific topic, such as conflicts of interest or baseline imbalances.

2. Tools for Quality Assessment:

Several standardized tools exist to facilitate the quality assessment:

a. Cochrane Risk of Bias Tool: Widely used for randomized controlled trials, this tool assesses risk across domains like randomization, blinding, and selective reporting.

b. ROBINS-I (Risk Of Bias In Non-randomized Studies - of Interventions): Used for non-randomized studies to assess factors like confounding, participant selection, and missing data.

c. QUADAS (Quality Assessment of Diagnostic Accuracy Studies): Specific for studies evaluating diagnostic tests.

d. CASP (Critical Appraisal Skills Programme): Provides checklists for various study designs, including randomized controlled trials, cohort studies, and qualitative studies.

3. Examples of Quality Assessment:

Consider a systematic review of the efficacy of a drug for migraine prevention. A study within the review might be assessed as:

a. Randomization: Adequately randomized using computer-generated sequences.

 b. Allocation Concealment: Double-blinded with sealed envelopes.

c. Blinding: Both participants and evaluators were blind to treatment assignment.

d. Incomplete Data: 5% dropout rate, but utilized intention-to-treat analysis.

e. Selective Reporting: All pre-specified outcomes reported.

f. Other Biases: No conflicts of interest declared.

Based on these criteria, the study might be rated as having a "low risk of bias."

4. Challenges in Quality Assessment:

a. Subjectivity: Different reviewers might assess the quality differently, so consensus methods and calibration exercises are crucial.

b. Insufficient Reporting: If studies do not provide adequate methodological details, assessing quality becomes challenging.

c. Over-reliance on Tools: While tools guide the assessment, critical thinking and expert judgment are also necessary.

In conclusion, quality assessment is crucial in understanding the reliability and credibility of the evidence presented in systematic reviews and meta-analyses. By carefully assessing the risk of bias in the included studies, researchers can provide more nuanced and trustworthy conclusions, promoting evidence-based decision-making in healthcare and other domains [29].

Conclusion

Systematic reviews and meta-analyses serve as invaluable tools in modern research, acting as the bridges between the ever-expanding body of individual studies and the actionable insights needed in real-world settings. These methodologies, by integrating, synthesizing, and critically evaluating existing evidence, provide researchers, clinicians, policymakers, and other stakeholders with comprehensive insights into specific research questions. However, as with any research approach, their value is deeply intertwined with the rigour of their methodology. Publication bias, the inclination toward publishing studies with significant or favorable outcomes, remains a major challenge. It can distort the overall picture of evidence, potentially overemphasizing beneficial effects or underplaying adverse outcomes. Mathematical tools, such as funnel plots and Egger\'s test, allow for detection of such biases, ensuring a more balanced understanding of the existing literature. Another facet, selection bias, arises when systematic reviewers unintentionally introduce discrepancies in study selection. To ensure that the chosen studies are truly representative and pertinent, transparency in the selection process becomes paramount. During data extraction, the accuracy, completeness, and standardization of the process determine the reliability of the derived data. Without standardized extraction, the subsequent analysis can be plagued with inconsistencies, rendering the final conclusions questionable. Variability in study quality is yet another significant concern. Not all studies are created equal. Their methodological rigor, transparency, sample size, and other factors can greatly influence the trustworthiness of their findings. Tools such as the Cochrane Risk of Bias Tool and ROBINS-I serve as critical aids in evaluating this quality, helping researchers discern the wheat from the chaff. Furthermore, the issue of heterogeneity cannot be understated. Variability among studies in terms of design, populations, interventions, or outcomes can significantly impact the results of a meta-analysis. The choice between fixed and random-effects models, based on the degree of heterogeneity, can thus dictate the robustness of the pooled estimates. Additionally, the importance of a comprehensive literature search and transparent inclusion/exclusion criteria is pivotal. The former ensures that all relevant studies, irrespective of their language, publication status, or region, are identified, while the latter provides a clear roadmap for study selection, mitigating potential biases. Standardized data extraction facilitates a systematic and uniform collection of information, ensuring consistency across various studies. Tools, training, and pilot tests become imperative in ensuring that this extraction is free from errors and biases. Finally, the assessment of study quality or risk of bias, helps in determining the credibility of the evidence. This step is essential for making informed conclusions, ensuring that the synthesized evidence is both reliable and applicable. As we reflect upon these myriad facets of systematic reviews and meta-analyses, a few key takeaways emerge. First, while these methodologies offer an unparalleled depth of insight, their efficacy is deeply rooted in the meticulousness of their process. Every step, from literature search to quality assessment, holds the potential to introduce bias or error. Thus, rigor, transparency, and critical evaluation become the cornerstones of a trustworthy review. Second, with the evolution of research methods and tools, the methodologies for systematic reviews and meta-analyses too must adapt. Continuous updating of guidelines, tools, and practices is crucial to ensure that these reviews remain relevant and rigorous in the face of changing research landscapes. Lastly, as consumers of these reviews—be it clinicians, policymakers, researchers, or the general public—it becomes our responsibility to critically assess them. Understanding their methodologies, recognizing their limitations, and questioning their conclusions is crucial for informed decision-making. In essence, systematic reviews and meta-analyses, when conducted with diligence and critical acumen, have the power to transform the vast ocean of individual studies into distilled insights. These insights, in turn, serve as beacons, guiding evidence-based practices and policies, and ultimately, elevating the standards of care, intervention, and understanding across various domains. As we move forward, let us harness the potential of these methodologies, while continually refining and challenging them, ensuring that they remain the gold standards in evidence synthesis.

References

[1] Soleimani, E., et al., The interactive relationship of dietary choline and betaine with physical activity on circulating creatine kinase (CK), metabolic and glycemic markers, and anthropometric characteristics in physically active young individuals. BMC Endocr Disord, 2023. 23(1): p. 158. [2] Smela, B., et al., Systematic literature reviews over the years. J Mark Access Health Policy, 2023. 11(1): p. 2244305. [3] Siri, G., et al., A high Diabetes Risk Reduction Score (DRRS) is associated with a better cardio-metabolic profile among obese individuals. BMC Endocr Disord, 2023. 23(1): p. 31. [4] Rahmani, E., et al., Point-Of-Care Ultrasonography for Identification of Skin and Soft Tissue Abscess in Adult and Pediatric Patients; a Systematic Review and Meta-Analysis. Arch Acad Emerg Med, 2023. 11(1): p. e49. [5] Rahmani, E., et al., Diagnostic Accuracy of Ultrasonography for Detection of Intussusception in Children; a Systematic Review and Meta-Analysis. Arch Acad Emerg Med, 2023. 11(1): p. e24. [6] Mokhtari-Ardekani, A., Fayyazishishavan, E., Akhavanfar, R. Abbasalizad-Farhangi, M., Circulating Advanced Oxidation Protein Products (AOPPs) increases the risk of metabolic syndrome among adults: A systematic review and meta-analysis. International Journal of Diabetes in Developing Countries, 2023. [7] Mahmoudinezhad, M., et al., The integrative panel of fatty acid desaturase-2 (FADS2) rs174583 gene polymorphism and dietary indices (DQI-I and HEI) affects cardiovascular risk factors among obese individuals. BMC Endocr Disord, 2023. 23(1): p. 41. [8] Kazemi, S.M., et al., Diagnostic Accuracy of Ottawa Knee Rule for Diagnosis of Fracture in Patients with Knee Trauma; a Systematic Review and Meta-analysis. Arch Acad Emerg Med, 2023. 11(1): p. e30. [9] Ghasemirad, M., et al., The association between screen use and central obesity among children and adolescents: a systematic review and meta-analysis. J Health Popul Nutr, 2023. 42(1): p. 51. [10] Taghizadeh, S., et al., Barriers and facilitators of childhood obesity prevention policies: A systematic review and meta-synthesis. Front Pediatr, 2022. 10: p. 1054133. [11] Chang, Y., et al., Is there a weekend effect in spine surgery?: A systematic review and meta-analysis of postoperative outcomes. Medicine (Baltimore), 2023. 102(42): p. e35640. [12] de Araújo, P.S.P., P.H.V. Pinto, and R.H.A. da Silva, Age estimation in adults by canine teeth: a systematic review of the Cameriere method with meta-analysis on the reliability of the pulp/tooth area ratio. Int J Legal Med, 2023. [13] Ding, C., et al., Association between diabetes and venous thromboembolism: A systematic review and meta-analysis. Medicine (Baltimore), 2023. 102(42): p. e35329. [14] Donnelly, R.R., et al., A Systematic Review and Meta-Analysis Investigating Head Trauma in Boxing. Clin J Sport Med, 2023. [15] Gajurel, B.P., et al., Utilization rates of intravenous thrombolysis for acute ischemic stroke in Asian countries:: A systematic review and meta-analysis. Medicine (Baltimore), 2023. 102(42): p. e35560. [16] Hu, W., et al., Efficacy of acupuncture or moxibustion in treating senile insomnia compared with a control group: A systematic review and meta-analysis. Medicine (Baltimore), 2023. 102(42): p. e34842. [17] Potdukhe, S.S., J.M. Iyer, and J.B. Nadgere, Evaluation of accuracy between extraoral Gothic arch tracing and various other methods assessing horizontal condylar guidance angle in completely edentulous patients: A systematic review and meta-analysis. J Indian Prosthodont Soc, 2023. 23(4): p. 322-334. [18] Sumedha, D. and K. Anita, Prevalence of beta thalassemia carriers in India: a systematic review and meta-analysis. J Community Genet, 2023. [19] Taneja, S., et al., Survival of tooth-implant connections: A systematic review and meta-analysis. J Indian Prosthodont Soc, 2023. 23(4): p. 310-321. [20] Veneri, F., et al., Fluoride Exposure and Skeletal Fluorosis: a Systematic Review and Dose-response Meta-analysis. Curr Environ Health Rep, 2023. [21] Wang, X., et al., It can be safe to discontinue oral anticoagulants after successful atrial fibrillation ablation: A systematic review and meta-analysis of cohort studies. Medicine (Baltimore), 2023. 102(42): p. e35518. [22] Rokhshad, R., et al., Deep learning for diagnosis of head and neck cancers through radiographic data: a systematic review and meta-analysis. Oral Radiol, 2023. [23] Lin, L. and H. Chu, Quantifying publication bias in meta-analysis. Biometrics, 2018. 74(3): p. 785-794. [24] Hernán, M.A., S. Hernández-Díaz, and J.M. Robins, A structural approach to selection bias. Epidemiology, 2004. 15(5): p. 615-25. [25] Nohr, E.A. and Z. Liew, How to investigate and adjust for selection bias in cohort studies. Acta Obstet Gynecol Scand, 2018. 97(4): p. 407-416. [26] Mithany, R.H., et al., Operative and Hepatic Function Outcomes of Laparoscopic vs. Open Liver Resection: A Systematic Review and Meta-Analysis. Cureus, 2023. 15(10): p. e47274. [27] Patel, J.J., et al., Critical Appraisal of a Systematic Review: A Concise Review. Crit Care Med, 2022. 50(9): p. 1371-1379. [28] Umbrello, M., et al., Music therapy reduces stress and anxiety in critically ill patients: a systematic review of randomized clinical trials. Minerva Anestesiol, 2019. 85(8): p. 886-898. [29] Cheung, M.W. and R. Vijayakumar, A Guide to Conducting a Meta-Analysis. Neuropsychol Rev, 2016. 26(2): p. 121-8.

Copyright

Copyright © 2023 Maryam Bigham. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.