Seventeen RCTs (n=1,212) were included.
Study quality was limited. Only 7 studies clearly reported allocation concealment and outcome assessment was blinded in only two. None had a loss to follow-up of more than 5% for short-term outcomes, but for longer-term outcomes only 18 to 60% of the participants were followed up. Fourteen studies had clearly defined outcomes and, apart for one (in each case), all had balanced groups at baseline and analysed the data on an intention-to-treat basis. The studies generally excluded patients with complications such as difficult anatomy or coagulation problems, which limits the generalisability of findings. Moreover, the effect on the outcomes of operator experience is unknown.
Wound infection (11 RCTs, n=870): there were significantly fewer clinically important wound infections in the PDT group than in the ST group (OR 0.28, 95% CI: 0.16, 0.49, p<0.0005).
Other short-term outcomes: there was no statistically significant difference between the groups for hospital mortality (12 RCTs, n=914), clinically relevant bleeding (10 RCTs, n=861), other major complications (8 RCTs, n=574), or duration of translaryngeal intubation prior to tracheostomy. There were few data on duration of mechanical ventilation and length of ICU stay, thus findings were not reported for these outcomes.
There was no evidence of statistically significant heterogeneity or evidence of publication bias for any outcome. Nor was there any evidence from sensitivity analyses that the results differed when RCTs with adequate evidence of allocation concealment were pooled.
Subgroup analyses: when studies were subgrouped by setting for performing the tracheostomy, in studies in which ST was performed in the operating theatre there was a statistically significant reduction in the bleeding rate in the PDT group (OR 0.29, 95% CI: 0.12, 0.75, p=0.01) and a reduction in the mortality rate in the PDT group that was of borderline statistical significance (OR 0.71, 95% CI: 0.50, 1.0. p=0.05). The method of PDT used and the use of a bronchoscope to guide placement did not significantly affect major outcomes.
Long-term outcomes (8 RCTs, n=800): the most commonly reported long-term complications were delayed closure of the stoma, airway symptoms and poor scar aesthetics. The low rate of follow-up precluded any conclusions being drawn for long-term outcomes.