Background: Observation and splenic artery embolotherapy (SAE) are nonoperative management (NOM) modalities for adult blunt splenic injury; however, they are quite different, inconsistently applied, and controversial. This meta-analysis compares the known outcomes data for observational management versus SAE by splenic injury grade cohort. Methods: Thirty-three blunt splenic injury outcomes articles, published between 1994 and 2009, comprising 24 unique data sets are identified. Of these, nine gave outcomes data by splenic injury grade for observational management and SAE separately. Failure rates were collected and analyzed using random effects estimates. Results: Overall, 68.4% of the 10,157 patients were managed nonoperatively. The overall failure rate estimate of NOM is 8.3% with a 95% confidence interval (CI) of 6.7% to 10.2%. The observational management failure rate estimate without SAE increases from 4.7% to 83.1% in splenic injury grade 1 to 5 patients. The overall failure rate estimate of SAE is 15.7% (95% CI, 10.4-23.2) and did not vary significantly from splenic injury grades 1 to 5 (p = 0.413). The failure rate of observational management without SAE is statistically higher than the failure rare estimate of SAE in splenic injury grade 4 and 5 injuries: 43.7% (95% CI, 25.5-63.8) versus 17.3% (95% CI, 7.8-34.1), p = 0.035 and 83.1% (95% CI, 45.2-96.7) versus 25.0% (95% CI, 8.7-53.8), p = 0.016, respectively. Conclusions: This meta-analysis synthesizes NOM outcomes data by modality and splenic injury grade. The failure rate of observational management increases with splenic injury grade, whereas the failure rate of SAE does not change significantly. SAE is associated with significantly higher splenic salvage rates in splenic injury grade 4 and 5 injuries.
