Prognostic factors for failure of nonoperative management in adults with blunt splenic injury: a systematic review.

BACKGROUND Contradictory findings are reported in the literature concerning prognostic factors for failure of nonoperative management (NOM) in the treatment of adults with blunt splenic injury. The objective of this systematic review was to identify prognostic factors for failure of NOM, with or without angiography and embolization. METHODS MEDLINE, Embase, and the Cochrane Library databases were searched. Prospective or retrospective cohort studies addressing failure of nonoperative treatment, with and/or without angiography and embolization, of blunt abdominal injuries were included. Methodological quality of the studies was assessed. RESULTS A total of 335 titles and abstracts were screened, of which 31 fulfilled the inclusion criteria. No randomized controlled trials were found. Ten articles were qualified as high-quality articles and used for data extraction (best-evidence synthesis). A total of 25 prognostic factors were investigated, of which 14 were statistically significant in one or more studies. Strong evidence exists that age of 40 years or above, Injury Severity Score (ISS) of 25 or greater, and splenic injury grade of 3 or greater are prognostic factors for failure of NOM. Moderate evidence was found for a splenic Abbreviated Injury Scale score of 3 or greater, trauma and ISS of less than 0.80, the presence of an intraparenchymal contrast blush, as well as transfusion of 1 unit of packed red blood cells or more. Limited evidence was found for large hemoperitoneum, lower Revised Trauma Score, lower Glasgow Coma Scale score, lower systolic blood pressure, male sex, the presence of traumatic brain injury, and splenic embolization as protective factor for failure of NOM. CONCLUSION Awareness for failure of NOM is required in patients aged 40 years or older, in patients with an ISS of 25 or higher or those with splenic injury grade 3 or higher. The prognostic factors for failure that we identified should be confirmed in future prospective cohort studies or meta-analyses using individual patient data. LEVEL OF EVIDENCE Systematic review, level III.