222 Chapter 10 Adjusted mortality data that accounted for patient characteristics and treatment variables was available from 32 studies (95 469 patients) and revealed a similarly increased mortality risk in female relative to male patients (pooled adjusted OR [aOR], 1.18; 95% CI, 1.11-1.27) (Figure 3). An influence analysis revealed that exclusion of any single study did not significantly alter the findings from the overall cohort (eAppendix 5 in Supplement 1). A sensitivity analysis with only studies that had an OR that was either reported or could be directly calculated (ie, excluding 14 studies in which RR or HR were reported) similarly did not change the overall findings (eFigure 3 in Supplement 1). No funnel plot asymmetry was found (eFigure 4 in Supplement 1). Substantial heterogeneity was observed in this analysis of adjusted mortality data (Q = 66.98; P < .001; I2 = 51%). Meta-analyses on subsets of studies showed that variation in the geographic location of the study impacted heterogeneity. Meta-analyses of studies conducted in individual geographic regions all had lower observed heterogeneity than the overall cohort (overall I2 = 51%): Europe (19 studies; I2 = 41%), North America (5 studies; I2 = 12%), East Asia (4 studies; I2 = 0%), and Middle East (3 studies; I2 = 0%). The pooled aOR varied significantly based on geographic location of study and ranged from 0.96 (95% CI, 0.76-1.22) for studies conducted in East Asia to 1.57 (95% CI, 1.23-2.01) for studies conducted in North America. Stratification of studies by mortality time point or by methicillin resistance did not impact heterogeneity. Evaluation of the evidence Given that this systematic review contained observational studies that accounted for confounding through statistical adjustment (ie, the adjusted analysis), the baseline strength of evidence was moderate. The mortality effect estimate was downrated due to a serious risk of bias because studies without a sex-difference in a univariable analysis would likely not have included this variable in a multivariable analysis. We did not have serious concerns about inconsistency, indirectness, imprecision, or publication bias. Therefore, the overall strength of evidence for the association of female sex with increased mortality risk in patients with SAB was low (eTable 2 in Supplement 1).
RkJQdWJsaXNoZXIy MTk4NDMw