Spontaneous Retroperitoneal Hematoma: Clinical Profile and Predictors of Mortality.

Abstract

Background: Spontaneous retroperitoneal hematoma (SRPHs) is a serious disorder infrequently reported in the literature. Our aim was to analyze the clinical profile and management of a series of cases and to determine mortality predictors.

Design: Retrospective cohort study.

Methods: We studied the consecutive series of patients with a SRPH at a university hospital from 2008 to 2023. Collected variables included clinical, physiological, and analytical characteristics as well as treatments provided. A modified Acute Physiology and Chronic Health Disease Classification System II (m-APACHE II) and Charlson scores were calculated. The association of these factors with mortality during hospital admission and follow-up were evaluated with logistic and Cox regression, respectively.

Results: Eighty-five patients with SRPH (mean age 75 years, 62.4% males) were identified. Of these, 56 (65.9%) were admitted for reasons other than SRPH, 67 (78.8%) received anticoagulants, and 53 (62.4%) had active bleeding on a CT scan. Seven patients were treated with palliative care. In the remaining 78 patients, management included transfusion (71 cases, 91%), anticoagulation reversal (25 cases, 32.1%), vasoactive drugs (21 cases, 26.9%), and admission to intermediate or critical care units (40 cases, 51.3%). Angiography was indicated in 34 (43.6%) patients, including embolization in 30 of these cases. Surgical drainage was required in two cases. Nineteen patients (24.5%) died during admission. m-APACHE II score (OR = 1.21; 1.08-1.36) was a predictive factor. One- and 5-year survival rates among hospital survivors were 68.4% and 29.2%, respectively. Survival was independently associated with the Charlson Comorbidity Index score (HR = 1.36; 1.14-1.63) and polypharmacy (HR = 1.13; 1.02-1.24).

Conclusion: SRPH is a serious disorder that requires complex therapeutic measures (critical care and angiography) in about half of patients. Several well-known scores are good predictors of mortality and could be useful in clinical decision-making.