Introduction The aim of this study was to evaluate the usefulness of the Shock Index (heart rate/systolic blood pressure), a widely used measure of haemodynamic stability in trauma and sepsis, in triaging patients with suspected upper gastrointestinal (UGI) bleeds. Compliance with assigning Blatchford and Pre-Rockall scores, the current risk assessment scores for UGI bleeds, is poor despite being mandated by national guidelines. We therefore wished to assess whether the Shock Index, a quick and easy measure, was as accurate at triaging these patients in determining the severity of the bleed.
Methods Data from 46 patients who underwent oesophago-gastro-duodenoscopies (OGD) between January and October 2015 was retrospectively analysed and the Blatchford, Pre-Rockall and Shock Index scores calculated. In our cohort, 19 patients were admitted via A&E and there was only one incidence of variceal bleeding and an additional case of cirrhosis. Sensitivity and specificity were calculated and Fischer’s Exact Test was performed to ascertain which of the scores was most correlated with adverse outcome (defined as needing transfusion). The cutoff point for an abnormal Shock index score was defined as >0.7, as used in the 2015 NCEPOD report.
Results Of the 46 patients reviewed, 17 had adverse outcomes indicating a severe UGI bleed. The Shock Index was raised in 8/17 of these patients (sensitivity 47.1%) whereas the Blatchford score was elevated (>0) in 16/17 (sensitivity 94.1%) and the Pre-Rockall score was raised (>0) in 15/17 (sensitivity 88.4%). However, of the 29 patients without a severe UGI bleed, the Shock index was <0.7 in 16/29 (specificity 55.2%) while 19/29 had a raised Blatchford score in the absence of a severe bleed (specificity 34.2%) and 17/29 had a raised Pre-Rockall (specificity 41.3%). Fischer’s Exact Test showed that the Blatchford and Rockall scores were correlated with severe outcomes with p values of 0.0356 and 0.0487 respectively, the Shock Index had a p value of 0.8834 demonstrating a non significant correlation with bleed severity.
Conclusion The Shock Index, a novel tool to assess the severity of UGI bleeds, lacks the sensitivity of the Blatchford and Pre-Rockall scores but is a better differentiator of those with mild UGI bleeds, as reflected in its greater specificity. The 2015 NCEPOD study demonstrated a sensitivity of 64.2% for the shock index when looking at patients requiring 4 or more units of blood, suggesting that it is better at identifying more serious bleeds. The role of the shock index could be as a useful adjunct to the more widely used Blatchford and Pre-Rockall scores and have a role in stratifying patients in which these scores are raised.
Disclosure of Interest None Declared