To the Editor Dr Subirà and colleagues evaluated 2 types of SBTs to evaluate readiness for extubation, a short-duration (30-minute) PSV strategy compared with a long-duration (2-hour) T-piece strategy, and found higher extubation rates with the PSV strategy (82.3% vs 74.0% at 72 hours; 92.5% vs 84.1% after the first SBT), with no difference in reintubation rates within 72 hours of extubation (11.1% vs 11.9%).1 Surprisingly, patients underwent their first SBT when the pretest probability of successful separation from the ventilator was already extremely high (as indicated by the success rate of the trial). When the pretest probability is high, any test or strategy having a risk of false-negative results (ie, predicting failure in a patient who could be extubated) will compromise predictive performance, whereas the risk of false-positive results (ie, predicting success in a patient who cannot be extubated) will be almost nonexistent.2 The design of this study put the T-piece group at higher risk of false-negative results than the PSV group.3,4 The results were therefore expected, as predicted by Bayes’ theorem.2 The relevant question for clinical practice is why the authors included patients at a late stage when the pretest probability of extubation was already high. Moreover, the absence of differences in length of stay in the intensive care unit casts doubt on any inference that the 30-minute PSV strategy meaningfully accelerated liberation from mechanical ventilation.
Mancebo J, Goligher E, Brochard L. Spontaneous Breathing Trials and Successful Extubation. JAMA. 2019;322(17):1716–1717. doi:https://doi.org/10.1001/jama.2019.14226
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: