Table 2

Number needed to treat with steroids to reduce reintubation and stridor in unselected, selected and overall populations

Unselected

Selected

Overall

(unselected+selected)

NNT to prevent one reintubation episode

44 (95% CI ≥ 26 to ∞)

9 (95% CI = 7 to 19)

28 (95% CI = 20 to 61)

NNT to prevent one stridor episode

Not calculated

5 (95% CI = 4 to 8)

11 (95% CI = 8 to 42)

Selected population is defined as patients at high risk of developing post-extubation stridor and reintubation in which the cuff-leak test showed absence or a low level of leak (less than 110 to 140 ml in absolute value or less than 12% to 25% in relative value).

Unselected population is defined as patients included in trials that did not use the cuff-leak test to select patients.

Overall population is defined as patients included in both trials that did use and did not use the cuff-leak test to selected patients (unselected+selected).

The NNT was calculated only when a significant result was observed.

CI = confidence interval; NNT = number needed to treat.

Jaber et al. Critical Care 2009 13:R49   doi:10.1186/cc7772