Penuelas et al (the Ventila group) report on 2,714 patients who were successfully extubated after at least 12 hours on mechanical ventilation, in 349 ICUs in 23 countries. (Enrollment was only for one month.)
About half were extubated within one day, and almost all (2,560 or 94%) were extubated within a week. They had a 28-day mortality of ~7%, and those who took 6-7 days had the same rough mortality as those extubated within 24 hours.
Those few (154 or 6%) who took longer than 7 days to be extubated had a mortality of 12%, an odds ratio of 1.97. AJRCCM 2011;184:430-437.
But let’s keep the authors’ conclusion, “Only patients who need more than 7 days for weaning have an increased mortality,” in perspective. These patients were young – only ~58 years old — and the authors only followed them for 28 days. (Most significantly, there is a built-in positive spin here: they only report on people who successfully escaped the ventilator.)
In fact, there is dramatic heterogeneity in reported outcomes among different patient populations undergoing mechanical ventilation. For example, in one of the surprisingly few extant observational population analyses incorporating real-world longitudinal data, Amber Barnato et al (AJRCCM 2011) showed that among elderly (~76 years) Medicare beneficiaries in the U.S. undergoing mechanical ventilation, almost 75% are dead within one year. Another 12% are disabled in nursing homes. Yikes.
These kinds of differences are so marked that I wonder if studying mechanical ventilation per se, as if it retains something “studyable” that’s independent of those being ventilated, isn’t quixotic or even misleading (by promoting improper extrapolation to unrelated clinical situations, and just perpetuating our general confusion and inaccuracy in prognosis). For a start, a follow-up of at least 90 days in these kinds of studies should be mandatory.