No abstract

BACKGROUND: Acute traumatic spinal cord injuries (SCIs) often result in impairments in respiration that may lead to a sequelae of pulmonary dysfunction, increased risk of infection, and death. The optimal timing for tracheostomy in patients with acute SCI is currently unknown. This systematic review and meta-analysis aims to assess the optimal timing of tracheostomy in SCI patients and evaluate the potential benefits of early versus late tracheostomy. METHODS: We searched Medline, PubMed, Embase, Cochrane Central, Cochrane Database of Systematic Reviews, and PsycINFO for published studies. We included studies on adults with SCI who underwent early or late tracheostomy and compared outcomes. In addition, studies that reported a concomitant traumatic brain injury were excluded. Data were extracted independently by 2 reviewers and copied into R software for analysis. A random-effects meta-analysis was performed to estimate the pooled odds ratio (OR) or mean difference (MD). RESULTS: Eight studies with a total of 1220 patients met our inclusion criteria. The mean age and gender between early and late tracheostomy groups were similar. The majority of the studies performed an early tracheostomy within 7 days from either time of injury or tracheal intubation. Patients with a cervical SCI were twice as likely to undergo an early tracheostomy (OR = 2.13; 95% confidence interval [CI], 1.24–3.64; P = .006) compared to patients with a thoracic SCI. Early tracheostomy reduced the mean intensive care unit (ICU) length of stay by 13 days (95% CI, −19.18 to −7.00; P = .001) and the mean duration of mechanical ventilation by 18.30 days (95% CI, −24.33 to −12.28; P = .001). Although the pooled risk of in-hospital mortality was lower with early tracheostomy compared to late tracheostomy, the results were not significant (OR = 0.56; 95% CI, 0.32–1.01; P = .054). In the subgroup analysis, mortality was significantly lower in the early tracheostomy group (OR = 0.27; P = .006). Finally, no differences in pneumonia between early and late tracheostomy groups were noted. CONCLUSIONS: Based on the available data, patients with early tracheostomy within the first 7 days of injury or tracheal intubation had higher cervical SCI, shorter ICU length of stay, and shorter duration of mechanical ventilation compared to late tracheostomy. The risk of in-hospital mortality may be lower following an early tracheostomy. However, due to the quality of studies and insufficient clinical data available, it is challenging to make conclusive interpretations. Future prospective trials with a larger patient population are needed to fully assess short- and long-term outcomes of tracheostomy timing following acute SCI.

Karn Wijarnpreecha and Fang Li shared co-first authorship.

In this study, we discuss a decision theoretic or fully Bayesian approach to the sample size question in clinical trials with binary responses. Data are assumed to come from two binomial distributions. A Dirichlet distribution is assumed to describe prior knowledge of the two success probabilities p1 and p2. The parameter of interest is p = p1 - p2. The optimal size of the trial is obtained by maximising the expected net benefit function. The methodology presented in this article extends previous work by the assumption of dependent prior distributions for p1 and p2.

Currently, methods for conducting multiple treatment propensity scoring in the presence of high‐dimensional covariate spaces that result from “big data” are lacking—the most prominent method relies on inverse probability treatment weighting (IPTW). However, IPTW only utilizes one element of the generalized propensity score (GPS) vector, which can lead to a loss of information and inadequate covariate balance in the presence of multiple treatments. This limitation motivates the development of a novel propensity score method that uses the entire GPS vector to establish a scalar balancing score that, when adjusted for, achieves covariate balance in the presence of potentially high‐dimensional covariates. Specifically, the generalized propensity score cumulative distribution function (GPS‐CDF) method is introduced. A one‐parameter power function fits the CDF of the GPS vector and a resulting scalar balancing score is used for matching and/or stratification. Simulation results show superior performance of the new method compared to IPTW both in achieving covariate balance and estimating average treatment effects in the presence of multiple treatments. The proposed approach is applied to a study derived from electronic medical records to determine the causal relationship between three different vasopressors and mortality in patients with non‐traumatic aneurysmal subarachnoid hemorrhage. Results suggest that the GPS‐CDF method performs well when applied to large observational studies with multiple treatments that have large covariate spaces.