Is the mortality higher in the pulmonary vs the extrapulmonary ARDS? A meta analysis.

<bold>Background and Aim: </bold>ARDS can occur from the following two pathogenetic pathways: a direct pulmonary injury (ARDSp); and an indirect injury (ARDSexp). The predisposing clinical factor can influence the pathogenesis and clinical outcome of ARDS. This metaanalysis was aimed at evaluating whether there is any difference in mortality between the two groups.<bold>Methods: </bold>We searched the MEDLINE, EMBASE, and CINAHL databases for relevant studies published from 1987 to 2007, and included studies that have reported mortality in the two groups of ARDS. We calculated the odds ratio (OR) and 95% confidence interval (CI) to assess mortality in patients with ARDSp vs patients with ARDSexp and pooled the results using three different statistical models.<bold>Results: </bold>Our search yielded 34 studies. In all, the studies involved 4,311 patients with 2,330 patients in the ARDSp group and 1,981 patients in the ARDSexp group. The OR of mortality in ARDSp group compared to the ARDSexp group was 1.11 (95% CI, 0.88 to 1.39), as determined by the random-effects model; 1.04 (95% CI, 0.92 to 1.18), as determined by the fixed-effects model; and 1.04 (95% CI, 0.92 to 1.18), as determined by the exact method, indicating that mortality is similar in the two groups. The mortality was no different whether the studies were classified as prospective (OR, 1.15; 95% CI, 0.87 to 1.51) or retrospective (OR, 1.01; 95% CI, 0.61 to 1.69); small (OR, 1.11; 95% CI, 0.77 to 1.60) or large (OR, 1.1; 95% CI, 0.82 to 1.49); or observational (OR, 1.10; 95% CI, 0.82 to 1.49) or interventional (OR, 0.97; 95% CI, 0.79 to 1.19). There was methodological and statistical heterogeneity (I(2), 50.9%; 95% CI, 21.3 to 66.2%; chi(2) statistic, 67.22; p = 0.0004).<bold>Conclusions: </bold>The results of this study suggest that there is no difference in mortality between these two groups. Further studies should focus on specific etiologies within the subgroups rather than focusing on the broader division of ARDSp and ARDSexp. [ABSTRACT FROM AUTHOR]