Clinical determinants of the severity of Middle East respiratory syndrome (MERS): a systematic review and meta-analysis

Abstract

Background: While the risk of severe complications of Middle East respiratory syndrome (MERS) and its determinants have been explored in previous studies, a systematic analysis of published articles with different designs and populations has yet to be conducted. The present study aimed to systematically review the risk of death associated with MERS as well as risk factors for associated complications.

Methods: PubMed and Web of Science databases were searched for clinical and epidemiological studies on confirmed cases of MERS. Eligible articles reported clinical outcomes, especially severe complications or death associated with MERS. Risks of admission to intensive care unit (ICU), mechanical ventilation and death were estimated. Subsequently, potential associations between MERS-associated death and age, sex, underlying medical conditions and study design were explored.

Results: A total of 25 eligible articles were identified. The case fatality risk ranged from 14.5 to 100%, with the pooled estimate at 39.1%. The risks of ICU admission and mechanical ventilation ranged from 44.4 to 100% and from 25.0 to 100%, with pooled estimates at 78.2 and 73.0%, respectively. These risks showed a substantial heterogeneity among the identified studies, and appeared to be the highest in case studies focusing on ICU cases. We identified older age, male sex and underlying medical conditions, including diabetes mellitus, renal disease, respiratory disease, heart disease and hypertension, as clinical predictors of death associated with MERS. In ICU case studies, the expected odds ratios (OR) of death among patients with underlying heart disease or renal disease to patients without such comorbidities were 0.6 (95% Confidence Interval (CI): 0.1, 4.3) and 0.6 (95% CI: 0.0, 2.1), respectively, while the ORs were 3.8 (95% CI: 3.4, 4.2) and 2.4 (95% CI: 2.0, 2.9), respectively, in studies with other types of designs.

Conclusions: The heterogeneity for the risk of death and severe manifestations was substantially high among the studies, and varying study designs was one of the underlying reasons for this heterogeneity. A statistical estimation of the risk of MERS death and identification of risk factors must be conducted, particularly considering the study design and potential biases associated with case detection and diagnosis.

Keywords: Ascertainment bias; Case fatality ratio; Comorbidity; Middle East respiratory syndrome.

Fig. 1

Flow diagram of study selection

Fig. 2

Estimated risks associated with Middle East respiratory syndrome (MERS) by published study. Panels show the risk estimates by study outcome: (a) risk of death, (b) risk of admission to the intensive care unit (ICU) and (c) risk of requiring mechanical ventilation. CFR represents the case fatality risk. The size of the diamonds reflects the sample size, and the whiskers extend to the lower and upper values of the 95% confidence interval (CI). The diamond without fill represents the pooled estimate using the inverse variance of the risk of death. I² measures the extent of the heterogeneity, representing the proportion of variance in a meta-analysis that is attributable to study heterogeneity. Khalid et al., 2015a is [16] while Khalid et al., 2015b is [19]

Fig. 3

Age and sex distributions related to MERS-associated death by published study. a Ages of patients that survived and died of Middle East respiratory syndrome (MERS) are compared. The range represents the minimum and maximum age. b Odds ratio (OR) of MERS death among men compared with women. The size of the diamonds reflects the sample size, and the whiskers extend to the lower and upper values of the 95% confidence interval (CI). The vertical dashed line shows the threshold value of OR = 1. The diamond without fill represents the pooled estimate using the inverse variance of OR. Khalid et al., 2015a is [16]. ICU represents intensive care unit

Fig. 4

Estimated risks associated with Middle East respiratory syndrome (MERS) by study design. Panels show the risk estimates by study outcome: (a) risk of death, (b) risk of admission to intensive care unit (ICU) and (c) risk of mechanical ventilation. CFR represents the case fatality risk. The estimate for each study design represents the pooled risk of death calculated using the inverse variance of the risk of death in each published study. The size of the diamonds reflects the sample size, and the whiskers extend to the lower and upper values of the 95% confidence interval (CI). The diamond without fill represents the pooled estimate using the inverse variance of the risk of death. I² measures the extent of the heterogeneity, representing the proportion of variance in a meta-analysis that is attributable to study heterogeneity

Fig. 5

Risk of death by Middle East respiratory syndrome (MERS) by underlying medical condition. Panels a, b, c, d, and e show the risk estimates by underlying medical condition. The odds ratio (OR) for MERS death to compare those with underlying conditions against those without underlying conditions was calculated. The size of the diamonds reflects the sample size, and the whiskers extend to the lower and upper values of the 95% confidence interval (CI). The vertical dashed line shows the threshold value of OR = 1. The diamond without fill represents the pooled estimate using the inverse variance of the OR. I² measures the extent of heterogeneity, representing the proportion of variance in a meta-analysis that is attributable to study heterogeneity

Fig. 6

Risk factors of death by Middle East respiratory syndrome (MERS) by study design. Panels show the risk estimates by covariate: (a) sex, (b): heart disease and (c): renal disease). Odds ratio (OR) represents the odds ratio of death among men with underlying medical condition compared with women without comorbidities, respectively. The size of the diamonds reflects the sample size, and the whiskers extend to the lower and upper values of the 95% confidence interval (CI). The diamond without fill represents the pooled estimate using the inverse variance of the risk of death. I² measures the extent of heterogeneity, representing the proportion of variance in a meta-analysis that is attributable to study heterogeneity