Does comorbidity increase the risk of patients with COVID-19: evidence from meta-analysis

Literature search and screening

The database searches identified a total of 324 potentially relevant articles. After the exclusion of duplicate references,243 articles were considered for the meta-analysis. Of these,208 studies were excluded after screening the title and abstract. After careful review of the full texts, 29 articles were excluded because they were reviews, cases, and insufficient data. Six studies qualified for inclusion [1, 4–8]. The flow diagram (Figure 1) showed the detailed literature search steps.

Figure 1. Flow diagram of the literature search and selection process in the meta-analysis.

Characteristics and quality of studies

A total of 1558 samples from 6 retrospective studies were enrolled in this meta-analysis [1, 4–8]. All six studies were performed in China. Six studies [1, 4–8] reported that hypertension, diabetes, and COPD, five covered liver disease [1, 4–7], four investigated malignancy [1, 4, 5, 8], renal disease [4–7], and cardiovascular disease [4–7], and three [4, 5, 7] researched cerebrovascular disease. Two studies [1, 5] used whether patients experienced ICU care to judge the severity of the disease, and the other four studies used clinical symptoms to judge the severity of the disease. All articles are of high quality because of NOS score no less than 6. Detailed descriptions of the studies included are shown in Table 1.

Hypertension, diabetes, and COPD

Six studies, including 324 severe group cases and 1234 non-severe group cases, provided the data in terms of hypertension, diabetes, and COPD [1, 4–8]. The heterogeneity test showed low heterogeneity among these studies, and a fixed-effects model was used for the meta-analysis. The results find that COVID-19 patients with hypertension (OR: 2.29, 95% CI: 1.69-3.10, P<0.001) (Figure 2A), diabetes (OR: 2.47, 95% CI: 1.67-3.66, P<0.001) (Figure 2B), or COPD (OR: 5.97, 95% CI: 2.49-14.29, P<0.001) (Figure 2C) had a higher risk of exacerbation.

Figure 2. Relationship between comorbidity and patients with COVID-19. (A) Hypertension; (B) Diabetes; (C) COPD; (D) Liver Disease.
(E) Malignancy; (F) Renal disease; (G) Cardiovascular disease; (H) Cerebrovascular disease.

Cardiovascular disease and cerebrovascular disease

Four included studies reported the relationship between cardiovascular disease and patients with severe COVID-19 [1, 4–6]. No significant heterogeneity was found (I² =0, P=0.989) among these trials, so a fixed effect pattern was selected. The results showed that cardiovascular disease is a risk factor for patients with COVID-19 (OR:2.93, 95% CI: 1.73-4.96, P<0.001) (Figure 2G).

Three studies provided the data in terms of cerebrovascular disease [4, 5, 7]. A fixed-effects model was used since the heterogeneity test suggested that there was no significant heterogeneity (I² =44.8%, P=0.163). The meta-analysis shows a significant relationship between patients with severe COVID-19 and cerebrovascular disease (OR:3.89, 95% CI: 1.64-9.22, P=0.002) (Figure 2H).

Liver disease, malignancy, and renal disease

Five studies comprising 313 severe group cases and 1167 non-severe group cases evaluated the role of liver disease in patients with COVID-19 [1, 4–7]. The meta-analysis showed that patients with the previous liver disease did not increase the risk of disease progression (OR:0.67, 95% CI: 0.30-1.49, P=0.326) (Figure 2D).

The relative risk assessments associated with malignancy and kidney disease are presented in Figure 2E and 2F, respectively. The meta-analysis suggested that there was no correlation between malignant tumor (the 95% confidence interval includes 1) or kidney disease (P=0.070) and COVID-19 patients' aggravation.

Subgroup analysis

To further verify the correlation of comorbidity and COVID-19 patients' aggravation, subgroup analysis was conducted. The results of the subgroup analysis are presented in Table 2. The Subgroup analysis results further support the results of hypertension, COPD, liver disease, and renal disease. In the clinical symptom group, we further observed that hypertension, diabetes, COPD, malignancy, and cardiovascular disease were a risk factor in COVID-19 patients.

Publication bias

The risk of publication bias was analyzed in the following comorbidities: hypertension, diabetes, COPD, and liver disease. Figure 3 shows the results of publication bias, which were evaluated by funnel plots and Eggers test. Begg’s test (All Pr>0.05) and Egger’s regression test (All P >0.05) suggest no significant publication bias.

Figure 3. Publication bias assessment. (A) Hypertension; (B) Diabetes; (C) COPD; (D) Liver Disease.

Search strategy and study selection

The Meta-analysis was performed according to the Preferred Reporting Items for Systematic reviews and Meta-analysis (PRISMA) statement [12]. Relevant literature was extracted by systematic retrieval of PubMed (Medline), EMBASE, Springer, Web of Science, and Cochrane Library databases up to date to March 1, 2020. Our search strategy included terms for “2019-nCoV” or “Coronavirus” or “COVID-19” or “SARS-CoV-2” or “2019-nCoV” or “Wuhan Coronavirus.” Besides, we manually screened out the relevant potential article in the references selected. The above process was performed independently by two participants.

Inclusion criteria are as follows: (1) Types of Studies: published studies reported the relationship between comorbidity and patients with COVID-19; (2) Subjects: diagnosed patients with COVID-19; (3) Exposure intervention: COVID-19 patients with comorbidity included: hypertension, diabetes, chronic obstructive pulmonary disease (COPD), liver disease, malignancy, renal disease, cardiovascular disease, cerebrovascular disease; (4) Outcome indicator: the odds ratios (OR) with 95% confidence intervals (CI) for each comorbidity.

The exclusion criteria: (1) Case reports, reviews, summaries of discussions, (2) Insufficient data information provided; (3) Patients were not stratified for the degree of severity.

Data extraction and quality assessment

Two participants separately conducted literature screening, data extraction, and literature quality evaluation, and any differences could be resolved through discussion or a third analyst. Information extracted from the included literature: first author surname, year of publication, country of the population, sample size, relevant data on comorbidity of severe and non-severe patients, etc.

The Newcastle-Ottawa scale (NOS) was adopted to evaluate the process in terms of queue selection, comparability of queues, and evaluation of results [13]. The quality of the included studies was assessed independently by two participants. NOS scores of at least six were considered high-quality literature. Higher NOS scores showed higher literature quality.

Statistical analysis

All data analysis was performed using Stata12.0 software (Stata Corp, College Station, Texas). The OR and relevant 95% CI were used to estimate pooled results from studies. After that, the heterogeneity test was conducted. When P≥0.05 or I²<50% was performed, it indicated that there was no obvious heterogeneity, and the fixed-effect model should be applied for a merger. Otherwise, the random-effect model was applied. Results were considered significant statistically when the p-value less than 0.05.

Studies were grouped according to the type of disease severity judgment basis. One subgroup is based on the clinical symptoms of patients, and the other subgroup is based on whether patients experience ICU care or not. Subgroup sensitivity analyses were conducted to explore potential sources of heterogeneity.

Publication bias was assessed using Begg funnel plot and Egger test linear regression test (where at least five studies were available). If P < 0.05 indicates obvious publication bias.