CHEST Guidelines-Response_chest

Response_chest_9

Pdf Summary

The document discusses criticisms and alternative methods for a meta-analysis conducted on the safety of thoracentesis and tube thoracostomy in patients with uncorrected coagulopathy. The original study used the Freeman-Tukey double arcsine transformation to report a pooled incidence of zero percent for major bleeding and mortality, which was critiqued for methodological choices. <br /><br />Firstly, the study employed Miller’s implementation of the double arcsine transformation, criticized for its bias in rare events. The estimate was biased towards zero due to using a denominator of 100 by default, leading to rounding off of the incidence rate. Alternative calculations with a denominator of 1,000 suggested the incidence was 0.63%, indicating there were 6.3 events per 1,000 observations. <br /><br />Additionally, the study utilized funnel plots to assess publication bias in the meta-analysis, which are noted to be inaccurate for proportion studies. The Doi plot, a more suitable method, indicated asymmetries suggesting publication bias due to smaller studies reporting higher incidences of bleeding. The Egger test, used by the original authors, was criticized for being underpowered in these studies, and a Luis Furuya-Kanamori index provided different indications of publication bias.<br /><br />Lastly, the use of the DerSimonian-Laird random-effects model is questioned, suggesting that it may overestimate incidences in cases of publication bias, with the inverse variance heterogeneity model recommended as an alternative. <br /><br />The response from the original authors acknowledges these methodological critiques but maintains that even with the adjustments, the incidence rate remains low, reinforcing the study's conclusions about the safety of the procedures. They conclude by emphasizing the overall low complication rates, despite acknowledging the limitations of their analytic approach.

Keywords

thoracentesis

tube thoracostomy

uncorrected coagulopathy

meta-analysis

Freeman-Tukey transformation

publication bias

Doi plot

DerSimonian-Laird model

Egger test

safety procedures