Introduction

The world is still dealing with a pandemic that was first reported in Wuhan province in December 2019 and the etiological agent was recognized as severe acute respiratory syndrome coronavirus 2 (SARS CoV2)1. During this period of one and half years (till 31st August 2021), a total of 217,925,862 cases were identified with a total of 4,524,091 death cases reported across the world2. Many factors including advanced age, obesity, and comorbidities such as diabetes mellitus, hypertension, epilepsy, sarcopenia and schizophrenia were associated with severe COVID-193,4,5,6,7. Moreover, damages to the internal organs such as the heart, liver, and kidneys were identified as major factors linked with severe COVID-19. Other factors such as time to hospital admission, tuberculosis, inflammation disorders, and coagulation dysfunctions also contributed to the higher fatality and mortality in COVID-19 patients3. Factors like current malignant status, dyspnoea, neutrophil/lymphocyte ratio, elevated C-reactive protein, lactate dehydrogenase, creatinine levels, oxygen saturation at admission, and use of azithromycin have been associated with mortality in geriatric COVID-19 patients8.

Inflammatory Bowel Diseases (IBDs) are a group of chronic conditions which affect the small and large intestines. Sustained inflammation in the gut may contribute to permanent damage of the intestine, compromise the quality of life and increase healthcare costs9. In 2017 a total of 6·8 million (6.4–7.3) people were reportedly living with IBD worldwide. Between 1990 and 2017, the age-standardized prevalence rate of IBD hiked from 79.5 (75.9–83.5) to 84.3 (79.2–89.9) persons per 100,000 population, while the death rate decreased from 0.61 (0.55–0.69) to 0.51 (0.42–0.54) per 100,000 population10.

The incidence and risk of COVID-19 in patients with IBD is still inconclusive. Several studies indicate a higher risk of COVID-19 and mortality in patients with IBD along with other factors such as advanced age and comorbidities11,12. In contrast to this, studies conducted by Maconi et al.13 and Ardizzone et al.14 reported a lower risk of COVID-19 in IBD patients. Interestingly, several other studies observed no cases of COVID-19 in the IBD cohorts they investigated15,16,17,18. In presence of all these conflicts, we aimed to identify all the currently available literature and assess the risk and outcomes of COVID-19 in patients with IBD through a comprehensive systematic literature review process and meta-analysis.

Methodology

We followed a PECOS framework (Population, Exposure, Control, Outcome, and Study Design) for the inclusion of relevant studies and adapted the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Guidelines19 to report this systematic review. Two independent reviewers were involved in the study selection, data extraction and methodological quality assessment and any disagreements were resolved through discussion or consultation with another reviewer.

Criteria for considering the studies for this review

Participants

We only considered patients who were diagnosed with IBD (CD, UC or IBD-unclassified) or microscopic colitis (MC) as per the author’s discretion in adult patients. Studies involving patients aged less than 18 or a population that included any disease other than IBD were excluded.

Exposure

The exposure or the etiology of interest were diagnosed with IBD such as CD, UC or IBD-unclassified and MC as per the author’s discretion.

Control

The comparator group considered were those who did not have IBD in the case of cohort studies and non-COVID patients in the case of case–control studies.

Outcomes

The outcomes considered were incidence and risk of COVID-19, COVID-19 hospitalization, severity of COVID-19, and mortality. Any studies which did not provide outcomes that were specific to IBD patients were excluded. The outcomes were considered as per author’s discretion or based on the report from authors.

Study designs

The quantitative observational studies such as cohort, cross-sectional (Descriptive and analytical), and case–control studies that assessed the incidence, risk, and/or outcomes of COVID-19 among adult IBD patients were considered for this review. The descriptive cross-sectional studies which presents only the prevalence were termed as prevalence studies and analytical cross-sectional studies marked as cross-sectional studies. Only the studies with full text available in the English language were considered. Reviews, descriptive studies, clinical trials, commentary, guidelines, and qualitative analyses were excluded.

Search methods for identification of studies

A comprehensive literature search was performed in PubMed/Medline (https://pubmed.ncbi.nlm.nih.gov/), Scopus (https://www.scopus.com/), Embase (https://www.embase.com/#search) and Cochrane Library (https://www.cochranelibrary.com/advanced-search) using all the possible keywords and entry terms in July 2021. We also did a snowball search in Google, Google Scholar Research Gate, and MedRxiv (https://www.medrxiv.org/) to identify any relevant articles. The reference lists of potential articles were also screened to identify additional potentially relevant citations. A detailed search strategy in various databases is provided as Supplementary File S1.

Study selection

All articles identified from databases following the literature search were retrieved to an Excel sheet and screened against the pre-defined criteria. The studies were screened by first reading the title and abstracts followed by reviewing the full text. Only studies that were not excluded at this stage were considered for final inclusion in the review.

Data extraction

The data were abstracted to a comprehensive data extraction form by two independent reviewers. The author’s first name and year of publication were used to identify the studies. The data regarding the publication, study settings, participants, and outcomes were captured from the studies. The number of events and sample size were collected from the studies or calculated from the available data.

Risk of bias and quality assessment

The Joanna Briggs Institute critical appraisal checklist was used to assess the methodological quality and risk of bias of included prevalence studies, cross-sectional studies, case–control studies and cohort studies20. The Joanna Briggs appraisal for prevalence studies addresses the appropriateness of sample frame, study participants, sample size, measure of condition, study setting, data analysis, and response rate. The checklist for cross-sectional studies assessed study aspects such as inclusion criteria, study subjects and setting, measure of exposure, measurement of the condition, confounding factors, outcomes measurement, and the statistical analysis used. The checklist for the case–control studies assessed study aspects such as comparability and matching of population, participant criteria, measurement of exposures, confounding factors, strategies to deal with confounding factors, outcome measurement, follow-up time, and statistical analysis The checklist for the cohort studies assessed study aspects such as recruitment of population, group assignment, measurement of exposures, confounding factors, strategies to deal with confounding factors, outcome measurement, follow-up time, incomplete data, and statistical analysis20.

Evidence synthesis and meta-analysis

All the evidence extracted through the systematic process was summarized narratively and presented in tabular form. Review Man 5.3 was used to conduct the meta-analysis21. The number of events and the total number of participants was used calculate prevalence rates and results were presented in terms of percentage with the 95% confidence interval (CI). The odds ratio (OR) was captured or calculated for the risk outcomes and the results were presented in terms of OR with 95% CI. The I2 statistics were used to estimate the heterogeneity in the analysis. We used the random effect model in case of substantial heterogeneity (I2 > 50%; P < 0.10) during all analyses. To explore the sources of heterogeneity, we performed subgroup analysis based on the type of IBD, wherever possible22.

Publication bias and sensitivity analysis

Visual inspection of the funnel plot generated through RevMan 5.321 was used to analyse the publication bias wherever feasible, i.e., analyses with minimum of 10 studies22,23,24. Whereas, statistical tests such as Egger’s and Begg’s test using comprehensive meta-analysis (trial version) were performed for all the analyses to check the statistical significance of publication bias. A probability of less than 0.05 was considered statistically significant. The sensitivity analysis was performed to check the robustness of the findings by removing the study with the lowest weight in each analysis and results were provided22.

Results

Study selection process

A total of 3733 records were identified from the electronic databases and 40 articles from other resources. Then, a total of 2828 non-duplicate records were initially screened by their title and abstracts, in which 2260 studies were excluded for appropriate reasons. The remaining 568 full-text articles were screened for their eligibility and 86 studies11,12,13,14,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106 were considered for final inclusion in this systematic review and meta-analysis. A detailed study selection is depicted in Fig. 1.

Figure 1
figure 1

Study selection process.

Full size image

Characteristics of included studies

Among the included studies, 50 studies (58.1%) were published in 2021 and 36 studies (41.9%) in 2020. The majority of studies emerged from the USA (30.23%; n = 26), followed by Italy (20.93%; n = 18), Spain (13.95%; n = 12), Denmark (3.49%; n = 3), France (3.49%; n = 3), the United Kingdom (3.49%; n = 3) Germany (2.32%; n = 2), Iran (2.32%; n = 2), Israel (2.32%; n = 2), and Norway (2.32%; n = 2). The remaining studies were from countries such as Chile (1.16%; n = 1), Egypt (1.16%; n = 1), China (1.16%; n = 1), France & Italy (1.16%; n = 1), India (1.16%; n = 1), Italy & Germany (1.16%; n = 1), Netherlands (1.16%; n = 1), Poland (1.16%; n = 1), Romania (1.16%; n = 1), Saudi Arabia (1.16%; n = 1), Serbia (1.16%; n = 1), and Sweden (1.16%; n = 1); information regarding the country was not available from one study (1.16%; n = 1) as it was a conference abstract. The studies included were single or multi centre; national or international; retrospective or prospective cohort studies, population-based cohort studies, case–control studies, registry analysis, and direct or web-based cross-sectional studies. A total of 34,059,455 participants were included among the studies, out of which 814,633 were IBD patients. The data were collected from IBD centres, medical colleges, hospitals or from the general public. A detailed assessment of characteristics for each included study is illustrated in Table 1.

Table 1 Characteristics of the included studies and patients.
Full size table

Risk of bias in the included studies

The quality assessment or risk of bias of included studies for the prevalence and outcomes of COVID-19 patients with IBD is presented in Supplementary File S2. We did not provide a score to the studies as Joanna Briggs's guidance discourages the use of a score cut-off for quality assessment107. Most of the studies were observed as being of good quality with an acceptable risk of bias. Among the prevalence studies, some studies failed to report the method used for the identification of the condition and its reliability. In the case of cross-sectional studies, the risk of bias was attributed to factors such as the identification and dealing with confounding factors in the study. The methodological quality of cohort studies was observed to be good and free of bias.

COVID-19 in patients with IBD

Prevalence of COVID-19

A pooled estimate of 63 studies11,12,14,25,29,31,32,36,37,39,40,41,42,43,49,50,51,52,53,54,58,59,61,66,67,68,69,72,75,77,78,79,80,81,82,85,86,88,89,90,91,92,93,94,95,96,98,99,102,103,104,106 indicated an overall prevalence rate of 6.10% (95% CI 3.15–9.04%) of COVID-19 in patients with any IBD. Heterogeneity was very high (I2: 100%); hence a random effect model was used.

Subgroup analysis indicated a prevalence of 9.43% (95% CI − 13.86 to 32.73%; n = 10 studies) of COVID-19 among the patients with CD26,27,28,34,35,45,46,73,76,84 and 8.58% (95% CI − 8.22 to 25.38; n = 10 studies) among those patients with UC27,28,34,35,45,46,55,73,76,84. This analysis indicated that, type of IBD was not a contributing factor to the heterogeneity as there was no change in the level of heterogeneity even after a subgroup analysis (Fig. 2).

Figure 2
figure 2

Prevalence of COVID-19 in patients with IBD.

Full size image

Visual inspection of funnel plot observed an obvious asymmetry (Supplementary Fig. S3A) indicating the chances of publication bias which was confirmed by Begg’s test (p = 0.014), but not with Egger’s test (p = 0.087). A sensitivity analysis by removing a study by Singh et al.99 indicated no much difference from the overall pooled estimate (6.07%; 95% CI 3.09–9.06%; 62 studies). The result is provided in Supplementary Fig. S4A.

Risk of COVID-19

A meta-analysis of 22 studies11,12,13,14,31,34,39,44,51,55,58,59,72,75,77,79,83,89,91,92,96,98,105 indicated a non-significant association between the IBD and COVID-19 (OR 1.15; 95% CI 0.97–1.37; p = 0.11) compared to non-IBD patients. A significant heterogeneity (I2: 90%) was observed, hence a random effect model was applied.

A Subgroup analysis by type of IBD revealed similar non-significant association with CD patients (OR 0.88; 95% CI 0.67–1.15; p = 0.33; n = 9 studies)14,31,39,55,59,79,92,96,98. Whereas, UC was significantly associated with a higher risk of COVID-19 (OR 1.37; 95% CI 1.07–1.74; p = 0.01; n = 9 studies)11,39,55,75,79,91,92,96,98 compared to the non-UC patients. However, the heterogeneity was not observed or non-significant when we performed a subgroup analysis based on the type of IBD such as CD (I2: 18%) and UC (I2: 0%). This indicates that the type of IBD might have contributed to the variation observed among the study findings (Fig. 3).

Figure 3
figure 3

Risk of COVID-19 among the IBD patients.

Full size image

Visual inspection of funnel plot does not show an obvious asymmetry which is suggestive no publication bias (Supplementary Fig. S3B) which was further confirmed by Egger’s (p = 0.999) and Begg’s test (p = 0.649). A sensitivity analysis by removing single study by Grunert et al.105 estimated no changes in the actual results (OR 1.16; 95% CI 0.97–1.38; 21 studies). The results are provided in Supplementary File S4B.

COVID-19 hospitalization among the patients with IBD

Prevalence of COVID-19 hospitalization

A total of 32 studies11,25,27,31,33,34,37,38,47,48,49,51,52,53,57,60,65,67,68,70,71,72,74,76,77,79,81,85,94,95,100,101 reported COVID-19-related hospitalizations among IBD patients and a pooled estimate identified that 10.63% (95% CI 6.67–14.60%) of IBD patients were admitted to the hospital. There was high heterogeneity (I2: 100%) observed among the studies and random effect model was applied.

Subgroup analysis observed that, only two studies57,76 recorded the COVID-19-related hospitalization rate in patients with CD and UC separately, which was 9.43% (95% CI − 7.90 to 26.75%) and 11.85% (95% CI − 9.25, 32.95%), respectively. No change was in heterogeneity was observed after subgroup analysis based on the type of IBD (Fig. 4).

Figure 4
figure 4

COVID-19 hospitalization among the patients with IBD.

Full size image

Visual inspection of funnel plot observed an obvious asymmetry suggestive of publication bias (Supplementary File S3C) which was not confirmed through Egger’s (p = 0.907) and Begg’s (p = 0.252) test. A sensitivity analysis by removing single study estimated no much changes in actual results (10.94; 95% CI 6.92, 14.96; 31 studies). The results are provided in Supplementary File S4C.

Risk of COVID-19 hospitalization

A pooled estimate of 13 studies12,13,27,48,49,50,57,63,72,77,94,99,101 indicated that IBD (OR 1.08; 95% CI 0.87–1.33; p = 0.50) was not significantly associated with risk of COVID-19 associated hospitalization. A random-effect model was used for the analysis as there was significant heterogeneity (I2: 87%).

The subgroup analysis recorded a significantly higher risk of COVID-19-related hospitalization in UC patients (OR 1.28; 95% CI 1.19–1.38; p < 0.00001; n = 4 studies) compared to non-UC patients27,50,57,94. However, CD (OR 0.94; 95% CI 0.84–1.06; p = 0.32; n = 2 studies)48,50 and MC (OR 1.28; 95% CI 0.95–1.72; p = 0.11; 1 study)60 was not significantly associated with risk of COVID-19 associated hospitalization. There was a non-significant heterogeneity after the subgroup analysis based on the type of IBD such as CD and UC (I2: 0%). This indicates that type of IBD might be a significant factor that contributed to the variation observed among the study findings (Fig. 5).

Figure 5
figure 5

Risk of COVID-19 hospitalization among the patients with IBD.

Full size image

The visual inspection of funnel plot observed an obvious asymmetry (Supplementary File S3D) suggestive of publication bias which was not confirmed by Egger’s (p = 0.325) and Begg’s (p = 0.228). A sensitivity analysis by removing Maconi et al.13 observed no changes in the actual findings (OR 1.10; 95% CI 0.89–1.36; 12 studies). The result is provided in Supplementary File S4D.

Severity of COVID-19 in patients with IBD

The severity of COVID-19 among IBD patients was reported in 19 studies11,35,46,49,53,65,66,70,74,77,78,79,80,81,82,87,89,95,101. The meta-analysis of these studies recorded that 7.95% (95% CI 1.10–57.26%; n = 9 studies) had mild disease35,46,53,66,78,79,80,81,82, 2.86% (95% CI 0.92–8.91%, n = 1 study) had moderate disease35 and 40.43% (95% CI 0.05–31,869.21%; n = 14 studies) had severe COVID-1911,35,46,49,65,70,74,77,79,82,87,89,95,101 in the IBD population. A significant level of heterogeneity (I2: 100%) was observed among the studies, hence a random effect model was used for the analysis (Fig. 6).

Figure 6
figure 6

Severity of COVID-19 among IBD patients.

Full size image

The visual inspection of funnel plot observed an obvious asymmetry (Supplementary File S3E) suggestive of publication bias which was confirmed by Egger’s (p = 0.025) and Begg’s (p = 0.0002). The sensitivity analysis was not performed for this analysis.

Risk of severe COVID-19 in patients with IBD

A summary estimate of 9 studies30,31,63,77,79,83,94,99,101 indicated no association between IBD and severe COVID-19 (OR 1.05; 95% CI 0.73–1.49; p = 0.80) compared to non-IBD patients. A substantial level of heterogeneity (I2: 66%) was observed, hence random effect model was applied.

The subgroup analysis indicated that CD patients30,79 had a significantly lesser risk of severe COVID-19 (OR 0.48; 95% CI 0.26–0.89; p = 0.02; 2 studies) while UC patients30,79,94 had a significantly higher risk of severe COVID-19 (OR 2.45; 95% CI 1.46–4.11; p < 0.0007; 3 studies). There was non-significant risk of severe COVID-19 with MC (OR 1.39; 95% CI 0.95–2.03; p = 0.09; 1 study)63 and IBD-unclassified (IBD-U) (OR 0.62; 95% CI 0.17–2.25; p = 0.47; 1 study)30. The heterogeneity observed in the overall analysis was not observed in subgroup analysis based on the type of IBD such as CD and UC (I2: 0%). This indicates that type of IBD might be a significant factor that contributed to the variation observed among the study findings (Fig. 7).

Figure 7
figure 7

Risk of severe COVID-19 among the patients with IBD.

Full size image

The visual inspection of funnel plot observed no obvious asymmetry (Supplementary File S3F) which is suggestive of no publication bias which was confirmed by Egger’s (p = 0.159) and Begg’s (p = 0.272). The sensitivity analysis by removing Burke et al.79 indicated no changes in overall results (OR 1.02; 95% CI 0.71–1.47). The result is provided in Supplementary File S4E.

Mortality in COVID-19 IBD patients

A pooled analysis of 24 studies25,26,27,33,34,38,47,48,57,60,65,67,72,74,76,77,79,82,86,89,94,95,101 estimated an overall mortality rate of 1.94% (95% CI 1.29–2.59%) in COVID-19 patients with IBD. A significant level of heterogeneity (I2: 98%) was observed among the studies.

Subgroup analysis indicates that, a single study76 reported mortality rate of 0.14% (95% CI − 1.82 to 2.10%) in CD patients and an estimate of 3 studies26,57,76 indicated a mortality rate of 2.79% (95% CI 0.60–4.99%) in patients with UC. The subgroup analysis based on the type of IBD did not alter the level of heterogeneity indicative of its non-contribution in heterogeneity (Fig. 8).

Figure 8
figure 8

Mortality rate in COVID-19 IBD patients.

Full size image

The visual inspection of funnel plot observed no obvious asymmetry (Supplementary File S3F) which is suggestive of no publication bias which was confirmed by Egger’s (p = 0.348) and Begg’s (p = 0.881). A sensitivity analysis by removing the study by Axelrad et al.67 indicated no changes in overall results (OR 1.90; 95% CI 1.30–2.62). The result is provided in Supplementary File S4F.

Risk of death or mortality

A meta-analysis of 11 studies11,12,13,27,48,50,51,57,62,77,101 observed that IBD was not significantly associated with COVID-19-related mortality (OR 2.31; 95% CI 0.78–6.81; p = 0.13) compared to non-IBD patients. A random-effect model was used as there was a significant level of heterogeneity (p < 0.10; I2: 98%).

The subgroup analysis also observed the similar non-significant association with CD (OR 6.28; 95% CI 0.55, 72.07; p = 0.14; 2 studies) and UC (OR 4.51; 95% CI 0.78–26.15; p = 0.09; 4 studies) compared to non-CD48,50 and non-UC participants11,27,50,57, respectively. There was no change in the level of heterogeneity (Fig. 9).

Figure 9
figure 9

Risk of COVID-19 mortality among the IBD population.

Full size image

The visual inspection of funnel plot observed no obvious asymmetry (Supplementary File S3H) which is suggestive of no publication bias which was confirmed by Egger’s (p = 0.849) and Begg’s (p = 0.881) (Fig. 10). A sensitivity analysis by removing the study by Maconi et al.13 indicated no changes in overall results (OR 2.29; 95% CI 0.75–6.94). The result is provided in Supplementary File S4G.

Figure 10
figure 10

Funnel plot for the risk of COVID-19 outcomes in IBD patients.

Full size image

Discussion

As there is conflicting evidence with respect to the incidence of COVID-19 in patients with IBD, the currently available guidelines for IBD management support the continuation of the use of biologics such as tofacitinib, ustekinumab, and vedolizumab108. Moreover, existing evidence fails to establish a positive association between the use of biologics or immunosuppressives with the risk of COVID-1979. Additionally, biologics use was associated with a lower risk of COVID-19 hospitalization, intensive care unit admission, and mortality among IBD patients107. Most of the studies were from the USA, Italy, and Spain, and the remaining countries were observed to have a lesser number of studies from an IBD population. This is an indication of underreporting, which might be due to a lack of manpower or test kits, and other barriers to access the data and patients107.

The findings from the current meta-analysis indicate an overall prevalence of 6.10% (95% CI 3.15–9.04%) of COVID-19 in patients with IBD. Moreover, a significant association could not be identified from the risk estimate (OR 1.15; 95% CI 0.97–1.37; p = 0.11). Similarly, a previous meta-analysis by Singh et al., suggested no difference in risk of COVID-19 in IBD patients when compared to the general population107. However, their findings with regards to the association between the risk of COVID-19 and type of IBD differed from our observation. They recorded a non-significant risk in both CD and UC patients, whereas our analysis showed a significantly higher risk in UC patients (OR 1.37; 95% CI 1.07–1.74; p = 0.01). This might be due to the single comparison group that was used by Singh et al., which is the general population. Interestingly, the meta-analysis of 14 studies performed by Tripathi et al., recorded a similar observation in which they recorded a very low incidence (1.01%) of COVID-infection in IBD cohort. Their therapy based analysis revealed a significantly poorer outcomes among those on corticosteroids or mesalamine, though anti-TNFs group had a better outcomes108. These findings were strengthened by another meta-analysis conducted by Alrashed et al.109,110. They also posed a higher risk with other management such as 5-aminosalicylic acid. However, use of vedolizumab, tofacitinib, and immunomodulators alone or in combination with anti-TNF were not associated with severe disease, rather anti-TNFs, and ustekinumab had a better outcomes. The reported COVID-19 hospitalization rate was 10.63% (95% CI 6.67–14.60%) in patients with IBD, which was lesser in the CD (9.43%; 95% CI − 7.90 to 26.75%) and higher in UC (11.85%; 95% CI − 9.25, 32.95%) subtypes, respectively. A similar trend was observed with the risk of COVID-19 hospitalization, where a non-significant association was found in the overall IBD population (p = 0.50), CD (p = 0.32), and MC (p = 0.11) patients. Moreover, the risk of COVID-19 hospitalization was significantly higher among patients with UC (p < 0.00001). The severe nature of disease, high level of immunosuppression, and higher hospitalization rate might have contributed to a significantly higher rate of COVID-associated hospitalization and severity in patients with UC than CD107,111.

We could observe through our meta-analysis that 7.95% and 2.86% of IBD patients had mild and moderate disease, though a higher percentage (40.43%; 95% CI 0.05–31,869.21%) had severe COVID-19. A non-significant association was observed between severe COVID-19 and IBD (p = 0.80), MC (p = 0.09) and IBD-U (p = 0.47). In contrast, a significantly lesser risk was observed in CD patients (OR 0.48; p = 0.02) and significantly higher risk in UC patients (OR 2.45; p < 0.0007). Along with the nature of the disease, factors such as advanced age of ≥ 65 years72,79, unvaccinated status89, CC subtype, use of oral steroids and proton pump inhibitors, rs13071258 A variant63, female gender, obesity, and concomitant diseases such as diabetes, hypertension, and asthma79 were associated with a higher risk of severe COVID-19 in IBD patients.

The pooled mortality rate was found to be 1.94%, 0.14%, and 2.79% in IBD, CD, and UC patients respectively. A non-significant association was observed between the COVID-19 mortality and IBD (p = 0.13) and subtypes such as CD (p = 0.14) and UC (p = 0.09). Comparatively, studies by Bezzio et al. (OR 8.45)11, and Ludvigsson et al. (OR 1.92)77 recorded significantly higher mortality in IBD patients. Similarly, Xu et al.50 and Bezzio et al.11 recorded significantly higher mortality in CD (OR 19.93) and UC (OR 22.65) patients, respectively. The evidence indicates that many other factors, such as the use of biologics12, advanced age11,42,95, active IBD status, higher Charlson comorbidity index score11, comorbidities, use of corticosteroids48,101,112 and thiopurines101 were significantly associated with COVID-19 mortality in the IBD population. Very recent evidence also indicates a non-significant effect of corticosteroids in mortality among patients with acute respiratory distress syndrome (ARDS), although positive evidence is reported in more recent randomized clinical trials113. Hence, the use of corticosteroids needs to be monitored in the general population as well as in IBD patients with COVID-19 or ARDS.

There was a significant level of heterogeneity observed in the pooled analysis of all outcomes such as the risk of COVID-19, COVID-19 hospitalization, and severe COVID-19, except for the mortality analysis. Through the subgroup analysis, we found that the type of IBD might have contributed to the heterogeneity as the heterogeneity decreased or became non-significant following the subgroup analysis based on the type of IBD. The risk of bias was observed to be lesser in our included studies which indicates a good quality of the studies. Our sensitivity analysis, which was done by removing the studies with the lowest weight, revealed the robustness of our findings by yielding a non-differing result from the original results.

Conclusions

The current evidence indicates that UC is significantly associated with a higher risk of COVID-19, COVID-19 hospitalization, and severe COVID-19 compared to non-UC participants. Additionally, CD patients had a significantly lesser risk of severe COVID-19 compared to non-CD patients. However, no significant association was observed between higher risk of COVID-19, COVID-19 hospitalization, severe COVID-19, and COVID-19 mortality among those who had been diagnosed non-specifically with IBD compared to non-IBD patients.