Six RCTs (n=446) and 7 retrospective studies (n=545 from 5 studies; the sample size was not reported in the other 2 studies) were included.
Mortality (4 RCTs and 5 retrospective studies).
There were no statistically significant differences in mortality rates between ET and no ET (3 RCTs) or between ET and LT (1 RCT and 5 retrospective studies). There was no evidence of any survival benefit for ET in the meta-analysis of 4 RCTs (RR 1.12, 95% CI: 0.74, 1.70; no evidence of heterogeneity, p=0.36).
Pneumonia (6 RCTs and 3 retrospective studies). Pneumonia rates were statistically significantly lower in 1 RCT and 2 retrospective studies comparing ET with LT. The meta-analysis of 6 RCTs found no evidence of any benefit for ET (RR 0.94, 95% CI: 0.85, 1.05). There was statistically significant heterogeneity between the studies (p=0.02) which disappeared when the trial of ET versus LT was removed from the analysis (the other studies compared ET with no ET), but the overall conclusion remained the same.
Ventilator and ICU days (5 RCTs and 6 retrospective studies). One RCT and 6 retrospective studies reported statistically significant reductions in the number of ventilator or ICU days for ET compared with LT. The meta-analysis of 5 RCTs showed a benefit for ET (p=0.0002) although there was significant heterogeneity (p=0.0008) between the study results. The heterogeneity was no longer significant when the only trial of ET versus LT was removed (p=0.63), although there was no longer any evidence of a benefit for ET from the meta-analysis results when excluding this study (p=0.27).
The subgroup analysis found that ET was associated with a non- statistically significant reduction in ventilator or ICU days compared with no ET in brain injured patients (p=0.06; based on 2 RCTs, n=129), but there was no difference between ET and no ET in non brain-injured patients (p=0.79; based on 3 RCTs).
Laryngotracheal pathology.
Five studies (3 RCTs of ET versus no ET and 2 retrospective studies of ET versus LT) reported laryngotracheal pathology rates, with the retrospective studies finding no tracheal stenosis in any patients and the RCTs finding no significant differences in pathology rates.