In-hospital mortality of patients requiring unplanned postoperative ventilatory support: a multicenter observational study

BACKGROUND: Most patients who are admitted non-intubated to surgery are extubated at surgery conclusion. Yet, 1–2% require unplanned postoperative ventilatory support. The outcome of these patients has not been thoroughly evaluated to date and is the focus of the present study. METHODS: Two-center observational study assessing characteristics and outcomes of surgical patients with unplanned mechanical ventilation during PACU stay between 2017 and 2019. Patients who arrived intubated to the operating room or were transferred directly to the intensive care unit (ICU) were excluded. The co-primary aims were to assess overall in-hospital mortality and to compare mortality between patients who were extubated in PACU and those who were discharged from PACU still intubated. The secondary aims were to compare postoperative respiratory infection and unplanned admissions to the ICU. Multivariate logistic regression was used to compare the groups and adjust for potential confounding variables. RESULTS: Overall, 698 patients were included. Of these, 135 died during hospital stay (mortality rate 19.3%, compared with 1.0% overall postoperative in-hospital mortality). Patients who still required ventilatory support at PACU discharge were significantly sicker, majority needed emergency surgery, and had more complicated surgical course compared to those who were extubated in PACU. In addition, their mortality rate [36% vs. 9%, adjusted OR (95% CI) 5.8 (3.8–8.8), p < 0.001], postoperative respiratory infection, and unplanned admission to ICU rates were also significantly higher. CONCLUSION: Unplanned postoperative mechanical ventilation is associated with noteworthy morbidity and mortality, with significantly higher rates in those of need for protracted (vs. short) mechanical ventilation. The remarkable mortality rate in patients extubated shortly after arriving to the PACU emphasizes the need for further studies to explore prompting factors and whether we can intervene to improve patients’ outcome.