Separation from mechanical ventilation and survival after spinal cord injury: a systematic review and meta-analysis

Abstract

Background: Prolonged need for mechanical ventilation greatly impacts life expectancy of patients after spinal cord injury (SCI). Weaning outcomes have never been systematically assessed. In this systematic review and meta-analysis, we aimed to investigate the probability of weaning success, duration of mechanical ventilation, mortality, and their predictors in mechanically ventilated patients with SCI.

Methods: We searched six databases from inception until August 2021 for randomized-controlled trials and observational studies enrolling adult patients (≥ 16 years) with SCI from any cause requiring mechanical ventilation. Titles and abstracts were screened independently by two reviewers. Full texts of the identified articles were then assessed for eligibility. Data were extracted independently and in duplicate by pairs of authors, using a standardized data collection form. Synthetic results are reported as meta-analytic means and proportions, based on random effects models.

Results: Thirty-nine studies (14,637 patients, mean age 43) were selected. Cervical lesions were predominant (12,717 patients had cervical lesions only, 1843 in association with other levels' lesions). Twenty-five studies were conducted in intensive care units (ICUs), 14 in rehabilitative settings. In ICU, the mean time from injury to hospitalization was 8 h [95% CI 7-9], mean duration of mechanical ventilation 27 days [20-34], probability of weaning success 63% [45-78] and mortality 8% [5-11]. Patients hospitalized in rehabilitation centres had a greater number of high-level lesions (C3 or above), were at 40 days [29-51] from injury and were ventilated for a mean of 97 days [65-128]; 82% [70-90] of them were successfully weaned, while mortality was 1% [0-19].

Conclusions: Although our study highlights the lack of uniform definition of weaning success, of clear factors associated with weaning outcomes, and of high-level evidence to guide optimal weaning in patients with SCI, it shows that around two-thirds of mechanically ventilated patients can be weaned in ICU after SCI. A substantial gain in weaning success can be obtained during rehabilitation, with additional duration of stay but minimal increase in mortality. The study is registered with PROSPERO (CRD42020156788).

Keywords: Intensive care unit; Mechanical ventilation; Rehabilitation; Respiratory failure; Spinal cord injury; Weaning.

Fig. 1

Flowchart of the study selection process according to PRISMA (www.prisma-statement.org)

Fig. 2

Forest plots for the outcome of complete liberation from the ventilator (left panel) and for the outcome of partial or complete weaning after rehabilitation (right panel). Studies are presented according to setting classification (intensive care units vs rehabilitation units): both overall and subgroup estimates are reported. Number of events and total number of patients at risk are reported for each study. The computation process has been described in the Methods. CI: confidence interval

Fig. 3

Forest plot for the outcome of duration of mechanical ventilation in intensive care units and rehabilitation units (upper panel). Studies are presented according to setting classification (intensive care units vs rehabilitation units): both overall and subgroup estimates are reported. Forest plots for the outcome of duration of mechanical ventilation for rehabilitation units (including the time to admission to rehabilitation) (lower panels). Weight refers to the relative contribution of each study to the meta-analytic estimate and is generated using the inverse variance method. CI:  confidence interval

Fig. 4

Forest plot for mortality. Studies are presented according to setting classification (intensive care units vs rehabilitation units): both overall and subgroup estimates are reported. CI: confidence interval