Premium
The Predictive Value of Preendoscopic Risk Scores to Predict Adverse Outcomes in Emergency Department Patients With Upper Gastrointestinal Bleeding: A Systematic Review
Author(s) -
Ramaekers Rosa,
Mukarram Muhammad,
Smith Christine A. M.,
Thiruganasambandamoorthy Venkatesh
Publication year - 2016
Publication title -
academic emergency medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.221
H-Index - 124
eISSN - 1553-2712
pISSN - 1069-6563
DOI - 10.1111/acem.13101
Subject(s) - medicine , emergency department , confidence interval , medline , meta analysis , cutoff , risk assessment , upper gastrointestinal bleeding , risk stratification , predictive value of tests , emergency medicine , endoscopy , physics , computer security , quantum mechanics , psychiatry , political science , computer science , law
Objectives Risk stratification of emergency department ( ED ) patients with upper gastrointestinal bleeding ( UGIB ) using preendoscopic risk scores can aid ED physicians in disposition decision‐making. We conducted a systematic review to assess the predictive value of preendoscopic risk scores for 30‐day serious adverse events. Methods We searched MEDLINE , PubMed, Embase, and the Cochrane Database of Systematic Reviews from inception to March 2015. We included studies involving adult ED UGIB patients evaluating preendoscopic risk scores and excluded reviews, case reports, and animal studies. The composite outcome included 30‐day mortality, recurrent bleeding, and need for intervention. In two phases (screening and full review), two reviewers independently screened articles for inclusion and extracted patient‐level data. The consensus data were used for analysis. We reported sensitivity, specificity, positive and negative predictive value, and positive and negative likelihood ratios with 95% confidence intervals. Results We identified 3,173 articles, of which 16 were included: three studied Glasgow Blatchford score ( GBS ); one studied clinical Rockall score ( cR ockall); two studied AIMS 65; six compared GBS and cR ockall; three compared GBS , a modification of the GBS , and cR ockall; and one compared the GBS and AIMS 65. Overall, the sensitivity and specificity of the GBS were 0.98 and 0.16, respectively; for the cR ockall they were 0.93 and 0.24, respectively; and for the AIMS 65 they were 0.79 and 0.61, respectively. The GBS with a cutoff point of 0 had a sensitivity of 0.99 and a specificity of 0.08. Conclusion The GBS with a cutoff point of 0 was superior over other cutoff points and risk scores for identifying low‐risk patients but had a very low specificity. None of the risk scores identified by our systematic review were robust and, hence, cannot be recommended for use in clinical practice. Future prospective studies are needed to develop robust new scores for use in ED patients with UGIB .