Higher Risk of Acute Respiratory Distress Syndrome and Risk Factors among Patients with COVID-19: A Systematic Review, Meta-Analysis and Meta-Regression

Objective: To estimate the global risk and risk factors associated with acute respiratory distress syndrome (ARDS) among patients with COVID-19: Design: A systematic review, meta-analysis and meta-regression. Setting and Participants: Hospitals or nursing homes and patients with acute respiratory distress syndrome after COVID-19. Methods: The literature review was systematically conducted on Embase, MEDLINE, CINAHL, and Web of Science, in addition to manual searches and reference list checking from 1 January 2019 to 2 March 2022. The search terms included coronavirus, acute respiratory syndrome, acute respiratory distress syndrome and observational studies. Three reviewers independently appraised the quality of the studies and extracted the relevant data using the Joanna Briggs Institute abstraction form and critical appraisal tools. A study protocol was registered in PROSPERO (CRD42022311957). Eligible studies were meta-analyzed and underwent meta-regression. Results: A total of 12 studies were included, with 148,080 participants. The risk ratio (RR) of ARDS was 23%. Risk factors were age ≥ 41–64 years old (RR = 15.3%, 95% CI =0.14−2.92, p = 0.03); fever (RR = 10.3%, 95% CI = 0.03−2.03, p = 0.04); multilobe involvement of the chest (RR = 33.5%, 95% CI = 0.35–6.36, p = 0.02); lymphopenia (RR = 25.9%, 95% CI = 1.11–4.08, p = 0.01); mechanical ventilation with oxygen therapy (RR = 31.7%, 95% CI = 1.10–5.25, p = 0.002); European region (RR = 16.3%, 95% CI = 0.09–3.17, p = 0.03); sample size ≤ 500 (RR = 18.0%, 95% CI = 0.70–2.89, p = 0.001). Conclusions and Implications: One in four patients experienced ARDS after having COVID-19. The age group 41–64 years old and the European region were high-risk groups. These findings can be used by policymakers to allocate resources for respiratory care facilities and can also provide scientific evidence in the design of protocols to manage COVID-19 worldwide.


Introduction
The novel coronavirus disease 2019 (COVID-19) is associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and severe acute respiratory syndrome (SARS), which caused the global pandemic of COVID-19 in late 2019 and which is continuing to date [1][2][3]. The rapid increase in COVID-19 has critically influenced society, healthcare systems, and people worldwide [4,5]. The clinical spectrum of disease presentation could be asymptomatic, mild, moderate, or severe, with some cases leading to death. In addition, some patients develop severe lung failure (acute respiratory distress syndrome, ARDS) [3, 6,7]. Previous research data showed that 76·40% of cases in Greece [8] and 61.7%

Research Question
What are the global risk and risk factors associated with acute respiratory distress syndrome among patients with coronavirus infection?

Search Strategy
This study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO Reg No-CRD42022311957). Five databases (Embase, MEDLINE, CINAHL, and Web of Science) were searched for studies on the prevalence of acute respiratory distress syndrome among patients with coronavirus published between 1 January 2019 and 2 March 2022. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA) were followed [19][20][21][22]. English synonyms including coronavirus, acute respiratory disease, 2019-nCoV, COVID-19, SARS-CoV, respiratory syndrome, severe acute respiratory syndrome, severe acute respiratory infection, Middle East respiratory syndrome, hospital, hospitalization, hospitalized, inpatient, patient, and sufferer were used to search each database. Several control vocabularies for the Emtree and MeSH databases were also used. For Emtree, these included "coronavirus", "Wuhan seafood market pneumonia virus", "Wuhan coronavirus", "Hospital patient", "Adult respiratory distress syndrome", "severe acute respiratory syndrome", "ARDS", and "SARS". We supplemented the search results with the Endnote X9 bibliographical database. Publications that cited the papers identified during the search and the reference lists of relevant articles and previous systematic reviews were manually screened to confirm the sensitivity of the search strategy.

Inclusion and Exclusion Criteria
The inclusion criteria were: (1) the study provided primary data on the prevalence of acute respiratory distress syndrome (ARDS) measured using validated assessment tools or coded medical report data within a population-based study after COVID-19 occurred; (2) participants were diagnosed with COVID-19; and (3) the studies were observational, such as cohort and cross-sectional studies, and were published in English, Chinese, or Sinhala from 2019 to 2022. The following types of studies were excluded: studies for which the study population did not include COVID-19 patients and studies that were qualitative research and review articles.
Titles and abstracts were independently screened by three researchers based on the inclusion and exclusion criteria after removing duplicates using the Endnote X9 bibliographical database. Then, the full texts of the selected studies were reviewed by three researchers independently, with any disagreement resolved by a fourth researcher to avoid selection bias.

Quality Assessment
All eligible studies were assessed for quality of evidence using the Joanna Briggs Institute (J.B.I.) Critical Appraisal for Checklist for Prevalence Studies Scale (CACPSS), which contains nine items and four responses (yes, no, unclear, and not applicable) [23]. Studies with a total score of 8 and above were considered acceptable quality evidence and were included in this systematic review and meta-analysis. Study quality and risk of bias were also independently assessed by the three reviewers, with any disagreement resolved by a fourth researcher.

Data Extraction
The following data were extracted: names of authors, year of publication, country, settings, study design, sample size, participant ethnicity, age, gender, and prevalence of acute respiratory distress syndrome. In addition, three authors independently assigned quality scores for the included studies according to the PRISMA guidelines [20][21][22], and any disagreements were resolved via a discussion among all four authors.

Statistical Analysis
A meta-analysis was conducted to identify statistical outcomes of higher risk of ARDS among patients with COVID-19 using the eligible studies. The pooled prevalence of acute respiratory distress syndrome was analyzed using several events converted to the risk ratio of 95% CI and p-values and a fitted model based on heterogeneity. Random or fixedeffects models were used based on the heterogeneity of results for acute respiratory distress syndrome among coronavirus infection patients. We transformed the proportions with the Freeman-Tukey double arcsine method before pooling the data for the prevalence rate of acute respiratory distress syndrome [24]. The heterogeneity value was assessed using I 2 , Cochran's Q test, and Tau2 for the included studies according to the DerSimonian-Laird estimator [25][26][27][28]. A zero value indicated the absence of heterogeneity, 25% indicated no significance or low significance, 50% indicated moderate heterogeneity, and 75% indicated significant heterogeneity. In the present study, 75-100% indicated significant heterogeneity, where the Q statistic and p-value were used to validate the heterogeneity results. In this meta-analysis, I 2 < 75% and p < 0.05 indicated statistical significance.
Publication bias was determined using funnel plots, and the Q statistic for Egger's test was used to determine the correlation between the effect estimate and the variances in the results for acute respiratory distress syndrome via C.M.A. software and a visual examination of the funnel plots [29,30]. A subgroup analysis and a meta-regression were performed to investigate potential sources of heterogeneity. For the meta-regression, we used the pool of effect size data as a single covariable introduced individually into the models. A simultaneous test was conducted to determine if all coefficients were zero in the model test. We used a null hypothesis model for the effect size comparison. Statistical analyses were conducted using Comprehensive Meta-Analysis Software version 3.0 (Biostat, Englewood, NJ, USA.) [23].

Study Identification
A total of 35,005 articles published from 1 January 2019 to 2 March 2022 were identified during the initial database search. Six thousand seven hundred eighty-seven articles were removed as duplicates using the Endnote X9 bibliographical database. The titles and abstracts of 28,220 articles were screened, and 1611 articles met the inclusion criteria.
The full text of each article was read to determine eligibility, and 1596 articles were excluded due to the following reasons: 469 articles did not have any relationship to COVID; 1099 articles did not mention ARDS behaviour among COVID-19 patients; 25 articles did not assess the outcome variables; and 3 articles were not available in a full-text format. In addition, three articles were removed after the quality assessment due to a low-quality score in the peer review. Finally, 12 articles were included in the systematic review and meta-regression ( Figure 1). Studies with quality scores of 8 and above were accepted as high quality (Supplementary Table S1). models. A simultaneous test was conducted to determine if all coefficients were zero in the model test. We used a null hypothesis model for the effect size comparison. Statistical analyses were conducted using Comprehensive Meta-Analysis Software version 3.0 (Biostat, Englewood, NJ, USA.) [23].

Study Identification
A total of 35005 articles published from 1 January 2019 to 2 March 2022 were identified during the initial database search. Six thousand seven hundred eighty-seven articles were removed as duplicates using the Endnote X9 bibliographical database. The titles and abstracts of 28220 articles were screened, and 1611 articles met the inclusion criteria.
The full text of each article was read to determine eligibility, and 1596 articles were excluded due to the following reasons: 469 articles did not have any relationship to COVID; 1099 articles did not mention ARDS behaviour among COVID-19 patients; 25 articles did not assess the outcome variables; and 3 articles were not available in a full-text format. In addition, three articles were removed after the quality assessment due to a lowquality score in the peer review. Finally, 12 articles were included in the systematic review and meta-regression ( Figure 1). Studies with quality scores of 8 and above were accepted as high quality (Supplementary Table S1).

Participant Characteristics
The total participants in the 12 studies were 148,080 individuals; 74,3851 were male, and 72,860 were female. The participant age range in 12 studies was 30-70 years of age, and one study [36] did not mention the participant's age.

Higher Risk of ARDS among Patients with COVID-19
Within the seven countries (the United States, Germany, Greece, India, China, Poland, and Korea), 12 studies analyzed the higher risk of acute respiratory distress syndrome among patients with COVID-19. The reported numbers were 7320 of 140760 participants, and four studies in the European region showed the highest rates of ARDS (Tables 1 and 2). After conducting a meta-analysis, we found that the pooled risk ratio of ARDS among patients with COVID-19 was 23% (95% CI = 14.3-34.7%, p = 0.001), with significant heterogeneity within the 12 studies (I 2 = 99.70, Q = 3685.601, Tau2 = 1.002, p = 0.001, Figure 2).

Participant Characteristics
The total participants in the 12 studies were 148,080 individuals; 74,3851 were male, and 72,860 were female. The participant age range in 12 studies was 30-70 years of age, and one study [36] did not mention the participant's age.

Higher Risk of ARDS among Patients with COVID-19
Within the seven countries (the United States, Germany, Greece, India, China, Poland, and Korea), 12 studies analyzed the higher risk of acute respiratory distress syndrome among patients with COVID-19. The reported numbers were 7320 of 140760 participants, and four studies in the European region showed the highest rates of ARDS (Tables 1 and 2). After conducting a meta-analysis, we found that the pooled risk ratio of ARDS among patients with COVID-19 was 23% (95% CI = 14.3-34.7%, p = 0.001), with significant heterogeneity within the 12 studies (I² = 99.70, Q = 3685.601, Tau2 = 1.002, p = 0.001, Figure 2).

Risk Factors of ARDS among COVID-19 through Meta-Regression Analysis
Based on the meta-analysis results, we identified significant heterogeneity within outcome variables of risk of ARDS. Therefore, a meta-regression analysis was conducted to identify factors affecting heterogeneity through the subgroups. The meta-regression model included the following risk factors for ARDS among patients with COVID-19: gender, age, smoking, cluster exposure history, fever, muscular soreness, cough, productive cough, sore throat, dyspnea, fatigue, headache, diarrhea, nausea and vomiting, lung infiltrates or consolidation, multilobe involvement in the chest, leucopenia, lymphopenia, underlying illnesses, diabetes mellitus, hypertension, chronic obstructive pulmonary disease, asthma, chronic kidney failure, chronic cardiac disease, coronary artery disease, thyroid disease, antiviral therapy, antibiotic therapy, nasal cannula oxygen therapy, mechanical ventilation oxygen therapy, WHO region, and sample size. The results regarding acute respiratory distress syndrome among coronavirus infection patients for statistical model 1, random effects, Z-distribution, and the log odds ratio. The model test was a simultaneous test to confirm that all coefficients (excluding the intercept) were zero (Q = 3685.601, df = 12, p = 0.00).

Publication Bias
Publication bias was analyzed using a funnel plot and Egger's test on ARDS among patients with COVID-19 for the 12 studies. However, the funnel plot did not show evidence of asymmetry, and there was a minor probability of publication bias. Statistically, possible publication bias was observed based on Egger's test results (Q = 3685.6, p = 0.001, I 2 = 99.70%, Figure 3) due to the diversity of the sample sizes and the length of the publication time in the included studies.

Risk Factors of ARDS among COVID-19 through Meta-Regression Analysis
Based on the meta-analysis results, we identified significant heterogeneity within outcome variables of risk of ARDS. Therefore, a meta-regression analysis was conducted to identify factors affecting heterogeneity through the subgroups. The meta-regression model included the following risk factors for ARDS among patients with COVID-19: gender, age, smoking, cluster exposure history, fever, muscular soreness, cough, productive cough, sore throat, dyspnea, fatigue, headache, diarrhea, nausea and vomiting, lung infiltrates or consolidation, multilobe involvement in the chest, leucopenia, lymphopenia, underlying illnesses, diabetes mellitus, hypertension, chronic obstructive pulmonary disease, asthma, chronic kidney failure, chronic cardiac disease, coronary artery disease, thyroid disease, antiviral therapy, antibiotic therapy, nasal cannula oxygen therapy, mechanical ventilation oxygen therapy, WHO region, and sample size. The results regarding acute respiratory distress syndrome among coronavirus infection patients for statistical model 1, random effects, Z-distribution, and the log odds ratio. The model test was a simultaneous test to confirm that all coefficients (excluding the intercept) were zero (Q = 3685.601, df = 12, p = 0.00).

Publication Bias
Publication bias was analyzed using a funnel plot and Egger's test on ARDS among patients with COVID-19 for the 12 studies. However, the funnel plot did not show evidence of asymmetry, and there was a minor probability of publication bias. Statistically, possible publication bias was observed based on Egger's test results (Q = 3685.6, p = 0.001, I 2 = 99.70%, Figure 3) due to the diversity of the sample sizes and the length of the publication time in the included studies.

Discussion
This systematic review and meta-analysis evaluated the higher risk of ARDS among patients with COVID-19 during the recent pandemic of COVID-19. A comprehensive search for relevant studies yielded 12 studies from 7 countries across the four WHO regions (region of the Americas, the Southeast Asian region, the European region, and the Western Pacific region) from 2020 to 2022. Additionally, this study has identified risk factors for ARDS among patients with COVID-19 to provide scientific evidence for respective stakeholders to prepare and allocate resources for any future pandemics.
According to our findings, 23% of COVID-19 patients experienced ARDS. This means that nearly one in four patients had progressed to ARDS. Those who had COVID-19 needed to have advanced care plans for further treatments because severe respiratory failure with the progress of ARDS is a possible complication of COVID-19 infection. Additionally, it may relate to the cause of death in COVID-19. Therefore, well-organized and reliable observational systems are beneficial for the patient screening process to detect ARDS early among patients with COVID-19, especially in people in residential care facilities. This will aid in early transfer to specialized medical care units for the proper management of ARDS to minimize the risk of death and severe complications, such as lung fibrosis or permanent lung damage. Most importantly, the early identification of high-risk patients will be the better choice for timely, evidence-based treatments and approaches to prevent further post-COVID-19 complications [9,15,35,37]. Furthermore, this study analyzed possible risk factors associated with ARDS in COVID-19 patients. Age greater than or equal to 41-64 years was a significant risk factor for ARDS among patients with COVID-19. Our results were consistent with previous studies, with similar significant factors associated with the poor prognosis of COVID-19 [34,[38][39][40]. In addition, as per previous literature, medical co-mobilities were associated with the risk of ARDS in middle-aged adults who had COVID-19, specifically pre-existing respiratory disease. However, we did not find any significant association between medical co-morbidities and ARDS [3,40]. Therefore, a pre-preparedness lifestyle modification strategy needs to be applied for future prevention plans for COVID-19 for middle-aged adults around the globe. An example of this would be effective communication for information distribution during pandemics within adult communities to mitigate their myths about infectious diseases such as COVID-19 in the future [41,42].
According to our analysis, fever, multilobe involvement in the chest, lymphopenia, and mechanical ventilation with oxygen therapy were significant clinical risk factors for ARDS among patients with COVID-19. This is because most respiratory distress patients need multifaceted ventilation systems and frequent position changes, such as prone positioning and vital sign monitoring. Therefore, clinicians and healthcare people need to arrange facilities for a long-time inpatient care management strategy for their clinical units [9,15,31,32]. It should be more suitable for nursing home care facilities to arrange respiratory care resources for a future convention.

Strengths
This study has several strengths. First, we have included studies from four WHO regions with different populations and ethnicities, such as black, white, Chinese, and Indian. This was the first systematic review and meta-regression for ARDS among patients with COVID-19 to include risk factors. Therefore, we recommend future studies should include the Eastern Mediterranean region with Arabic and Islamic populations and the African region with the black African community to see if there is any difference in risk of ARDS and risk factors. This is because, in our findings, the European region is one of the significant risk factors for ARDS among patients with COVID-19 [3, 8,9]. In our study, when to start mechanical ventilation and how mechanical ventilation processes proceeded during the clinical management were also found to be risk factors for ARDS. Therefore, it is essential to identify the oxygenation peak flow measurement during ventilation periods, such as from the starting point until the end. Additionally, continuous blood saturation monitoring and advanced technological methods for blood oxygenation, such as extracorporeal membrane oxygenation (ECMO), are recommended [3,14,43-45].

Limitations
There were some limitations to our study. First, we noticed a possible publication bias due to the diversity of the sample sizes and the length of the publication time in the included studies. Therefore, we need to include a large sample with long-term observational studies, such as prospective cohort studies, throughout the pandemic. At this time, the available published studies were limited to 12 within seven countries due to early publication. Future analysis can be focused more on studies with new treatment strategies, such as ECMO, for ARDS and their survival rate. Most of the patients in the included studies are still in the hospital under treatment. Therefore, post-COVID co-morbidity for ARDS among patients with COVID-19 need to be analyzed in the future rehabilitation of patients with ARDS.

Conclusions
One in four patients has a risk of ARDS after acquiring COVID-19. The risk factors included middle-aged adults older than or equal to 41-64 years old, fever, multilobe involvement in the chest, lymphopenia, and mechanical ventilation with oxygen therapy. Additionally, the European region is at a high risk of ARDs among COVID-19 patients. Therefore, this study's findings are beneficial for frontline clinicians, healthcare clinical decision-makers, and health policymakers to precisely justify the healthcare system and government of COVID-19 to arrange early interventions and suitable treatment strategies.
This study provides scientific evidence to support clinical practice and the design of protocols to prevent ARDS. It is also a reference for future researchers who plan to examine ARDS and the risk factors among diverse populations. We recommend that future studies focus on the Eastern Mediterranean and African regions with multiple co-morbidities.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph192215125/s1, Table S1. The Joanna Briggs Institute score for quality of evidence for prevalence studies. Funding: This research was supported by the An Nan Hospital, China Medical University, Tainan, Taiwan (ANHRF109-30). The content is solely the responsibility of the authors and does not necessarily represent the official views of the An Nan Hospital. The funding sources did not have any role in the study design, collection, analysis and interpretation of data or in the writing of the manuscript.

Informed Consent Statement: Not applicable.
Data Availability Statement: Data sharing does not apply to this article as no datasets were generated or analyzed during the current study.