Eleven studies (n=1,299, range 11 to 502) were included. Eight studies were prospective, two were retrospective and the design of the remaining study was unclear. Six studies enrolled consecutive patients. Results of the QUADAS assessment were not reported.
There was significant heterogeneity in estimates of sensitivity across studies (p<0.01, I2=74%). Sensitivity ranged from 69% to 100% and was above 84% in all but one study. Summary sensitivity was 93% (95% CI 91% to 94%). Sensitivity was higher in the eight studies that enrolled patients on the basis of a positive CT or PET result (94%, 95% CI 93% to 96%) compared with those that enrolled patients regardless of result (76%, 95% CI 65% to 85%) (p=0.02). There was no difference in estimates of sensitivity for any of the other subgroups investigated.
Specificity was found to be 100% in all studies that gave a summary specificity of 100% (95% CI 99% to 100%).
One patient had a confirmed pneumothorax that required drainage following the procedure. One patient suffered from hypoxemia during the procedure and recovered soon after. No other complications were reported.