WE7.6 Failure of intensive monitoring in high grade blunt splenic injury: an important consideration beyond NOM failure. (9th August 2022)

• Record Type: Journal Article
• Authors: Lee, Sheah Lin; Long, Joe; Rizkallah, George; Modi, Sachin; Elberm, Hassan
• Is Part Of: British journal of surgery. Volume 109(2022)Supplement 5
• Publication Date: 2022-08-09
• Subjects: Surgery -- Periodicals; 617.005
• DOI: 10.1093/bjs/znac248.159
• Languages: English
• ISSNs: 0007-1323

Abstract

Abstract: Aims: Non-operative management (NOM), including intensive monitoring (IM) and angioembolisation (AE), is the standard of care for haemodynamically stable patients with blunt splenic injury to preserve spleen and it's function. Internal review of local Level 1 Trauma Center revealed a NOM failure rate of 4.4%, but an IM failure rate of 16.2% in high grade blunt splenic injury (HGBSI). Age, Injury Severity Score (ISS) and moderate to severe injuries are predictors for NOM failure. However, predictors of IM failure, which can be successfully treated with AE, is not well understood, motivating the current study. Methods: Patients with BSI are identified from Trauma Audit and Research Network (TARN) databse from 2015–2020. Clinical, biochemical and radiological data were collated from TARN database and electronic medical records. HGBSI is defined as AAST grade III to V. Results: Sixty-eight patients with HGBSI were identified. Seventy-eight percent (n=53) were initially managed with IM, with 21% (n=11) requiring subsequent interventions. Majority who failed IM were involved in road traffic accident compared to other causes (p-value=0.035). In addition, a higher proportion had systole blood pressure<90mmHg (p-value=0.033) and pH<7.2 (p-value=0.038) at presentation. AAST grade, ISS or age did not influence requirement for further intervention. Nine of the eleven patients were treated successfully with AE. Conclusion: Understanding predictors of IM failure will enable better selection of patients for AE at initial presentation, leading to potential reduction in duration of IM, length of stay and blood transfusion. Preliminary work using machine learning algorithm to predict IM failure with larger dataset is on-going.