Diabetes and deaths of COVID-19 patients: Systematic review of meta-analyses

Background COVID-19 is affecting all kinds of patients including diabetics. This article provides an overview of conducted meta-analyses regarding the effect of diabetes on the deaths of COVID-19 patients. Methodology The study was conducted as per preferred reporting items for systematic reviews and meta-analysis (PRISMA) statement. Data sources The relevant meta-analyses were searched on PubMed till April 2021 and data was extracted from 24 relevant meta-analyses. The overall estimate was calculated in terms of odds ratio or relative risk with a 95% confidence interval. Results A total of 09 meta-analyses showed the association of diabetes with the death of COVID-19 patients and 15 meta-analyses have reported the association of diabetes with other comorbidities in the death of COVID-19 patients. The pooled odds ratio or relative risk has shown a significant association of diabetes alone or its associated comorbidities with deaths of COVID-19 patients Conclusion Patients with diabetes and its associated comorbidities need more monitoring if get SARS-Cov-2 infection to reduce deaths.


Introduction
Coronaviruses (CoVs) are a large group of viruses with positivesense, single-stranded RNA causing diseases in animals and humans. The first human coronavirus was found in 1962 and these are found to be associated with upper and lower respiratory tract infections. Over the last two decades, severe acute respiratory syndrome (SARS-CoV) and Middle East Respiratory Syndrome (MERS) emerged in different parts of the world [1] , however, not as much lethal as SARS-CoV-2 is. The first case of SARS-CoV-2 was reported in China [2] . It is believed that the SARS-CoV-2 infection is correlated to the Huanan seafood market but the exact source is still not verified. The infection caused by SARS-CoV-2 is already stated as a pandemic by world health organization (WHO) [3] . The symptoms of SARS-CoV-2 infection such as low-and high-grade fever, dry cough, difficulty in breathing, severe headache, fatigue, loss of taste, smell, etc. vary from person to person. The SARS-CoV-2 infection results in a series of immune responses causing cytokine It has been observed that diabetic patients with SARS-Cov-2 infection are more susceptible to enter into serious conditions [10] . You and colleagues [11] have reported a higher risk of death of diabetic COVID-19 patients as compared to non-diabetic patients in Korea. In a retrospective cohort study, it was observed that diabetes comorbidity was related with an increased risk of ICU admission and/or mortality [12] . Hui et al. [ 13 ] have reported a high risk of death in COVID-19 patients with diabetes as compared to non-diabetic patients. Studies conducted in New York city (NYC), also reported that the patients with diabetes were more expected to need mechanical ventilation or ICU admission [ 14,15 ]. Another large population-based study in England including 264,390 patients with type 1 DM and 2874,020 patients with type 2 DM also reported higher death in COVID-19 patients with DM, renal complications, cardiovascular diseases, obesity, etc. [16] . Recently, many meta-analyses have been published regarding the association of diabetes and COVID-19. However, still there is confusion among physicians and healthcare professionals regarding the same. Thus, we have extracted the information from a published meta-analysis and discussed it in this article.

Methodology
The study was conducted as per the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [17] .

Data sources
The published meta-analysis was searched in PubMed with the followings search terms (COVID 19) OR (COVID-19 virus infection) OR (COVID-19 infection) OR (2019 novel coronavirus infection) OR (2019-nCoV infection) OR (2019-nCoV disease) AND (diabetes mellitus) OR (type 2 DM)) OR (Type 1 DM)) OR (DM) AND (meta-analysis). AG and MKP both cross-check the studies. The conflict among studies was resolved after discussion with third author (AK).

Study selection
All meta-analyses on the impact of diabetes on the death of COVID-19 patients were considered which were published until April 2021. The death should be one of the outcomes of the metanalysis.

Data extraction
The data was extracted from a published meta-analysis in an Excel sheet which contained the following titles: references, year of publication, place of conduct of the study, number of included studies, pooled analysis results, limitations, etc. The demographic and clinical characteristics of included meta-analyses are presented in Table 1 and Table 2 .

Methodological assessment
The methodological assessment of the included meta-analyses was evaluated using the checklist [18] having 21 items and is listed in Table 3 . Twenty-four meta-analyses were included in the overview. The selection of meta-analyses is presented in Fig. 1 .

Results and discussion
The results of individual metanalysis were compiled in Tables 1 &  2 and are discussed below.

Meta-analysis of studies containing COVID-19 patients with diabetes
Nine meta-analyses showed that diabetes was associated with the death of COVID-19 patients. Shang et al. [19] have done a meta-analysis of 28 studies and the overall estimate was measured in terms of odd ratio (OR) with 95% CI. The pooled OR was reported 2.21 with 95% CI (1.83-2.66), and p < 0.001 which indicates the significant association of diabetes with death of COVID-19 patients. The heterogeneity among studies was calculated by I 2 statistics and reported as 50%. Aggarwal et al. [20] have also shown an association of diabetes with death of COVID-19 patients [OR: 2.03, 95% CI: 1.29-3.20]. There is no heterogeneity among studies (I 2 = 0%, Cochran's Q = 2.63, P = 0.45). However, the number of included studies for statistical analysis was less ( n = 4). The result of meta-analysis done by Varikasuvu et al. [21] showed a significant increase in death of COVID-19 patients with diabetes [OR = 2.52, (1.93-3.30), p < 0.00001]. Mantovani, et al. [22] have done a metaanalysis of 15 studies using the random-effect model and reported pooled OR 2.68 with 95% CI (2.09-3.44) which indicates a greater risk of in-hospital mortality associated with COVID-19 diabetic patients. The I 2 was reported to be 46.7% which indicates heterogeneity among studies. Guo et al. [23] have reported pooled relative risk (RR) 2.96 (95% CI: 2.31-3.79; p < 0.001) which indicates diabetes is associated with disease severity or death of COVID-19 patients. Kumar et al. [24] have done a metanalysis to find out the association of diabetes with the severity and mortality of COVID-19 patients. The pooled OR 2.75 [95% CI: 2.09-3.62; p < 0.01 indicates a significant association of diabetes with the severity of COVID-19 patients. The pooled OR with mortality was reported 1.90 [95% CI: 1.37-2.64; p < 0.01]) which also indicates a significant association of diabetes with mortality of COVID-19 patients. Hussain, et al. [25] have also reported an association of diabetes with mortality in COVID-19 patients with a pooled risk ratio of 1.61 (95% CI: 1.16-2.25%), p = 0.005. Wu et al. [26] have reported pooled OR of 1.75 (95% CI 1.31-2.36; p = 0.0002) which indicates a significant association of diabetes with mortality of COVID-19 patients. Huang et al. [27] reported RR 2.12 [1.44, 3.11], p < 0.001 which also indicates a significant correlation. The I 2 was found to be 72% which indicates heterogeneity among studies. All these meta-analyses of studies containing COVID-19 patients with diabetes as one of the comorbidities are compiled in Table 1 .

Meta-analysis of studies containing COVID-19 patients with diabetes as one of the comorbidities
In most cases, diabetes is associated with other medical conditions such as hypertension, obesity, hyperlipidemia, etc. The search of the current study has found 15 meta-analyses which are mentioned as follows. Tian et al. [28] have reported high mortality of patients with hypertension (56.8%) followed by diabetes (31.2%) and coronary heart disease/cardiovascular disease (CHD/CVD) (21.5%). The pooled OR in diabetic patients was reported 2.0 with 95% CI, 1.7 -2.3; P < .00001 which indicates a significant correlation. Nandy et al. [29] have reported the effect of various comorbidities as well as diabetes on COVID-19 patients. The pooled OR with diabetes was reported with OR 2.28 (1.40 to 5.55) which indicates a significant correlation. Qiu, et al. [30] have also reported various risk factors such as hypertension, chronic cardiovascular and cerebrovascular disease, diabetes in COVID-19 patients. The risk of complications associated with COVID-19 diabetic patients was reported 22.2% (95% CI 19.30 ∼25.10%). Lu et al. [31] have conducted a meta-analysis of studies conducted in Wuhan and reported diabetes, chronic lung disease, and hypertension as one of the mortality risks factors. The pooled OR in cases of diabetes was reported with OR = 3.73, 95% CI 2.35-5.90. Mesas et al. [32] have reported pooled odd ratio for diabetes 2.12 (1.79, 2.52), p-value = < 0.001, I 2 = 77.9. Dorjee et al. [33] have found a significant risk of death in diabetic patients. The relative risk was found to be RR: 1.50 [95% CI: 1.35-1.66]. However, the heterogeneity among studies was found to be higher. Ng et al. [34] have done the metanalysis to find out the effect of various co-morbid conditions including diabetes in COVID-19 patients. The pooled hazards ratio was found to be 1.94 (1.54, 2.46) which indicates a significant risk of mortality in COVID-19 patients with diabetes. Ssentongo et al. [35] have also reported significant risk [pooled relative risk: 1.48 (1.02    Table 2 .

Limitations
One of limitation of the current study is moderate to high heterogeneity in terms of types of the population, sample size, and age of included subjects among studies. In most of the ameta-analysis, authors have not done sub-group analysis due to a limited number of studies. Thus, further studies with large sample sizes are required to do subgroup analysis. We have searched only PubMed and meta-analysis published in the English language is only considered.

Future perspectives
Diabetes and its associated comorbidities are significantly associated with the death of COVID-19 patients in all of the published metaanalyses. Therefore, COVID-19 patients with diabetes and its associated comorbidities need much more monitoring to reduce deaths.
Clinically relevant and focused study question included Terms used for electronic literature search provided Manual search conducted through references of articles, abstracts Characteristics of studies listed (ie, sample size, patient demographics) Inclusion and exclusion criteria provided for studies

Evaluation of results
Studies were combinable Appropriate statistical methods used to combine results

Evaluation for publication bias
Publication bias addressed through evaluation methods such as funnel plot or sensitivity analysis

Applicability of results
Results were generalizable

Funding source
Funding source(s) stated Fig. 1. The scheme of selection of meta-analysis.

Declaration of Competing Interest
The authors declare that they have no competing interests.