UNASSIGNED: Nonoperative management (NOM) is the standard treatment for hemodynamically stable blunt splenic injury (BSI). However, NOM failure is a significant source of morbidity and mortality. We developed a clinical risk scoring system for NOM failure in BSI.
UNASSIGNED: Data from the Japanese Trauma Data Bank from 2008 to 2018 were analyzed. Eligible patients were restricted to those who underwent NOM with high-grade BSI (Organ Injury Scale ≥3). The primary outcome was a predictive score for NOM failure based on risk estimation.
UNASSIGNED: There were 1651 patients included in this analysis, among whom 110 (6.7%) patients had NOM failure. Multivariate analysis identified seven variables associated with failed NOM: systolic blood pressure, Glasgow coma scale, Injury Severity Score, other concomitant abdominal injury, pelvic injury, high-grade BSI, and angioembolization. An eight-point predictive score was developed with a cut-off of greater than 5 points (specificity, 98.2%; sensitivity, 25.5%) with an area under the curve of 0.81.
UNASSIGNED: The clinical predictive score had good ability to predict NOM failure and may help surgeons to make better decisions for BSI.

Angioembolization (AE) is an adjunct to nonoperative management (NOM) of splenic injuries. We hypothesize that failure of AE is associated with blood transfusion, grade of injury, and technique of AE.
We performed a retrospective (2010-2017) multicenter study (nine Level I trauma centers) of adult trauma patients with splenic injuries who underwent splenic AE. Variables included patient physiology, injury grade, transfusion requirement, and embolization technique. The primary outcome was NOM failure requiring splenectomy. Secondary outcomes were mortality, complications, and length of stay.
A total of 409 patients met inclusion criteria; only 33 patients (8%) required delayed splenectomy. Patients who failed received more blood in the first 24 h (P = 0.009) and more often received massive transfusion (P = 0.01). There was no difference in failure rates for grade of injury, contrast blush on computed tomography, and branch embolized. After logistic regression, transfusion in the first 24 h was independently associated with failure of NOM (P = 0.02). Patients who failed NOM had more complications (P = 0.002) and spent more days in the intensive care unit (P < 0.0001), on the ventilator (P = 0.0001), and in the hospital (P < 0.0001). Patients who failed NOM had a higher mortality (15% versus 3%, P = 0.007), and delayed splenectomy was independently associated with mortality (odds ratio, 4.2; 95% confidence interval, 1.2-14.7; P = 0.03).
AE for splenic injury leads to effective NOM in 92% of patients. Transfusion in the first 24 h is independently associated with failure of NOM. Patients who required a delayed splenectomy suffered more complications and had higher hospital length of stay. Failure of NOM is independently associated with a fourfold increase in mortality.