Patients Requiring Invasive Mechanical Ventilation outside the Intensive or High-care Units

This abstract has open access
Abstract Description

In resource limited regions, many critically ill patients receive invasive mechanical ventilation in a non-ICU/designated high-care environment. In Hong Kong there are different models-of-care provided for this group of patients in general wards: unstructured care in general wards, or a designated ward with either a designated ventilation team, or a supporting team from ICU. We conducted a prospective observational cohort study to evaluate outcomes, and whether different models-of-care are associated with mortality.

Data from 7 hospitals, from January to April 2016, was recorded. Hospital mortality, and time from study recruitment to death, or 90 days, was recorded. Standardized mortality ratio (SMR) using the Mortality Probability Model (MPM III) was calculated. Cox regression was used to estimate the hazard ratio (HR, with 95% CI) for comparing mortality between models-of-care, taking hospitals clustered within models-of-care into account.

We excluded 185 patients either undergoing limitation-of- life-support within 24 hours, or being cared in one hospital adopting a different model-of-care (only 15 eligible patients), the analysis was based on 285 patients, with 3 different models-of-care:

Model A: Designated ward/no designated ventilation team/supporting team from ICU (1 hospital)

Model B: Designated ward/designated ventilation team/no supporting team from ICU (2 hospitals)

Model C: No designated ward/no designated team/no supporting team from ICU (3 hospitals)

Of 285 patients, 173 died (61%, 95% CI: 55%-66%) in hospital, and 187 (66%, 95% CI: 60%-71%) had died within 90 days after intubation. Overall SMR was 1.82 (95% CI:1.56-2.11). In the cox regression model, stratified by mechanical ventilation duration (< 48h vs ≥48h), and adjusted for MPM III score and causes for respiratory failure, there was a significant difference between models-of-care (P< 0.001). Discrimination was acceptable (c-statistic=0.71). A designated ward, and a ventilation team or supporting team from ICU may improve survival.

Abstract ID :
HAC1374
Submission Type
914 visits