A systematic review and meta-regression analysis of nonoperative management of blunt traumatic thoracic aortic injury in 2897 patients.

BACKGROUND

Thoracic endovascular aortic repair has transformed the management of blunt traumatic thoracic aortic injuries (BTTAI). Recent studies have suggested that the nonoperative management (NOM) of BTTAI may be a viable alternative. We investigated the NOM of BTTAI by conducting a systematic review and meta-analysis of the mortality proportions and incidence of complications.

METHODS

We searched PubMed through June 22, 2017, and referenced lists of included studies without language restriction, with the assistance of a trained librarian. We included studies that reported the NOM of BTTAI (≥5 participants). Two authors independently screened titles, abstracts, and performed data extraction. Pooled prevalence of mortality (aortic related, in hospital) were obtained based on binomial distribution with Freeman-Tukey double-arcsine transformation and continuity correction. The random-effects model was used for all analyses to account for variation between studies. Meta-regression was performed to explore sources of heterogeneity, including Injury Severity Score, age, and gender.

RESULTS

We included 35 studies comprising 2897 participants. The pooled prevalence of all-cause in-patient mortality in the overall, grade I, grade II, grade III, and grade IV populations are as follows: 29.0% (95% confidence interval [CI], 19.3%-39.6%; I = 95%; P < .01), 6.8% (95% CI, 0.6%-19.3%; I = 52%; P = .03), 0% (95% CI, 0%-2.0%; I = 0%; P = .81), 29.2% (95% CI, 17%-42.5%; I = 3%; P = .41), and 87.4% (95% CI, 16.4%-100%; I = 48%; P = .14), respectively. The combined incidence of aortic-related in-patient mortality in the overall, grade I, grade II, and grade III populations are: 2.4% (95% CI, 0.4%-5.5%; I = 60%; P < .01), 0.93% (95% CI, 0%-14.2%; I = 65%; P < .01), 0% (95% CI, 0%-1.8%; I = 0%; P = .99), and 0.13% (95% CI, 0%-6.4%; I = 14%; P = .33), respectively. The total proportion of postdischarge aortic-related mortality is 0% (95% CI, 0%-0.5%; I = 0%; P = .91). Meta-regression showed a decreased risk of in-hospital mortality as age increases (β = .99; 95% CI, 0.98-1.00), an increased risk of in-hospital mortality with a higher Injury Severity Score (β = 1.02; 95% CI, 1.00-1.04), and a decreased risk of in-hospital mortality among male patients (β = .54; 95% CI, 0.3-0.90).

CONCLUSIONS

This study provides, to our knowledge, the most up-to-date pooled estimate of mortality rates after the NOM of BTTAI. However, its interpretation is limited by the paucity of data and substantial quantitative heterogeneity. If patients are to be managed nonoperatively, we would recommend the judicious use of active surveillance in a select group of patients in the short, mid, and long term.