Email updates

Keep up to date with the latest news and content from BMC Gastroenterology and BioMed Central.

Open Access Research article

Risk score to predict gastrointestinal bleeding after acute ischemic stroke

Ruijun Ji1, Haipeng Shen2, Yuesong Pan1, Penglian Wang1, Gaifen Liu1, Yilong Wang1, Hao Li1, Aneesh B Singhal3, Yongjun Wang1* and On behalf of China National Stroke Registry (CNSR) investigators

Author Affiliations

1 Tiantan Comprehensive Stroke Center, Tiantan Hospital, Capital Medical University, No.6 Tiantanxili, Dongcheng District, Beijing 100050, China

2 Department of Statistics and Operation Research, University of North Carolina, Chapel Hill, NC, USA

3 Department of Neurology, Massachusetts General Hospital, Boston, MA, USA

For all author emails, please log on.

BMC Gastroenterology 2014, 14:130  doi:10.1186/1471-230X-14-130

Published: 25 July 2014

Abstract

Background

Gastrointestinal bleeding (GIB) is a common and often serious complication after stroke. Although several risk factors for post-stroke GIB have been identified, no reliable or validated scoring system is currently available to predict GIB after acute stroke in routine clinical practice or clinical trials. In the present study, we aimed to develop and validate a risk model (acute ischemic stroke associated gastrointestinal bleeding score, the AIS-GIB score) to predict in-hospital GIB after acute ischemic stroke.

Methods

The AIS-GIB score was developed from data in the China National Stroke Registry (CNSR). Eligible patients in the CNSR were randomly divided into derivation (60%) and internal validation (40%) cohorts. External validation was performed using data from the prospective Chinese Intracranial Atherosclerosis Study (CICAS). Independent predictors of in-hospital GIB were obtained using multivariable logistic regression in the derivation cohort, and β-coefficients were used to generate point scoring system for the AIS-GIB. The area under the receiver operating characteristic curve (AUROC) and the Hosmer-Lemeshow goodness-of-fit test were used to assess model discrimination and calibration, respectively.

Results

A total of 8,820, 5,882, and 2,938 patients were enrolled in the derivation, internal validation and external validation cohorts. The overall in-hospital GIB after AIS was 2.6%, 2.3%, and 1.5% in the derivation, internal, and external validation cohort, respectively. An 18-point AIS-GIB score was developed from the set of independent predictors of GIB including age, gender, history of hypertension, hepatic cirrhosis, peptic ulcer or previous GIB, pre-stroke dependence, admission National Institutes of Health stroke scale score, Glasgow Coma Scale score and stroke subtype (Oxfordshire). The AIS-GIB score showed good discrimination in the derivation (0.79; 95% CI, 0.764-0.825), internal (0.78; 95% CI, 0.74-0.82) and external (0.76; 95% CI, 0.71-0.82) validation cohorts. The AIS-GIB score was well calibrated in the derivation (P = 0.42), internal (P = 0.45) and external (P = 0.86) validation cohorts.

Conclusion

The AIS-GIB score is a valid clinical grading scale to predict in-hospital GIB after AIS. Further studies on the effect of the AIS-GIB score on reducing GIB and improving outcome after AIS are warranted.