Diabetes mellitus increases the risk of hospital mortality in patients with Covid-19: Systematic review with meta-analysis

Abstract

Background: The mortality rate associated with Covid-19 varies considerably among studies and determinants of this variability are not well characterized.

Methods: A systematic review of peer-reviewed literature published through March 31, 2020 was performed to estimate the mortality rate among hospitalized patients in China with a confirmed diagnosis of Covid-19. Hospital mortality rates were estimated using an inverse variance-weighted random-effects meta-analysis model. Funnel plot symmetry was evaluated for small-study effects, a one-study removed sensitivity analysis assessed the influence of individual studies on the pooled mortality rate, and metaregression assessed the association of potential confounding variables with mortality rates.

Results: The review included 16 observational studies involving 1832 hospitalized patients with a diagnosis of Covid-19. The surveillance period among studies ranged from December 16, 2019 to February 23, 2020. The median patient age was 53 years and 53% were males. A total of 38.5% of patients presented with at least 1 comorbidity, most commonly hypertension (24.0%), cardiac disease (15.1%), and diabetes mellitus (14.4%). Fever and cough, reported in 84.8% and 61.7% of patients respectively, were the most common patient symptoms. The pooled mortality rate was 9.9% (95% confidence interval 6.1% to 14.5%). Funnel plot asymmetry was not observed and the meta-analysis results were not substantially influenced by any single study since the pooled mortality rate ranged from 8.9% to 11.1% following iterative removal of one study at a time. Substantial heterogeneity in the mortality rate was identified among studies (I = 87%; P < .001). In a metaregression that included demographics, patient risk factors, and presenting symptoms, only a higher prevalence of diabetes mellitus was associated with a higher mortality rate (P = .03).

Conclusions: In a meta-analysis of hospitalized patients in China with a diagnosis of Covid-19, the mortality rate was 9.9% and a higher diabetes mellitus prevalence was independently associated with a worse prognosis. The independent influence of diabetes mellitus with Covid-19 mortality should be viewed as hypothesis-generating and warrants further study.

Figure 1

PRISMA Study Flow Diagram. PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-analyses.

Figure 2

Prevalence of patient risk factors, presenting symptoms, and administered treatments in observational Covid-19 studies in China. Plotted values represent the weighted frequency and 95% confidence interval among included studies.

Figure 3

Forest plot of mortality in hospitalized patients in China diagnosed with Covid-19. The mortality rate and 95% confidence interval are plotted for each study. The pooled mortality rate (diamond apex) and 95% confidence interval (diamond width) is calculated using a random effects model. Pooled mortality = 9.9%. Heterogeneity: I ² = 87%; P < .001.

Figure 4

Funnel plot of mortality in hospitalized patients in China diagnosed with Covid-19. Funnel plot asymmetry due to small-study effects was not apparent by visual inspection.

Figure 5

Metaregression of the association between diabetes mellitus and mortality in hospitalized patients in China diagnosed with Covid-19. Open circles represent values of individual studies where circle size is proportional to the study weight in the random-effects model. The red line represents the regression line of best fit. P = .03.