Pulmonary Vascular Disease |

Accuracy of the IMPROVE Bleeding Risk Score for Hospitalized Medical Patients FREE TO VIEW

Brian Foster, DO; Elizabeth Marx, BS; Sarah Petteys, MD; Paul Clark, DO; Jordanna Hostler, MD; Joshua Mitchell, MD; Jacob Collen, MD; Aaron Holley, MD
Author and Funding Information

Walter Reed National Military Medical Center, Bethesda, MD

Chest. 2014;146(4_MeetingAbstracts):823A. doi:10.1378/chest.1993563
Text Size: A A A
Published online


SESSION TITLE: DVT/PE/Pulmonary Hypertension Posters I

SESSION TYPE: Original Investigation Poster

PRESENTED ON: Wednesday, October 29, 2014 at 01:30 PM - 02:30 PM

PURPOSE: Recent guidelines from the ACCP recommend all medical inpatients be assessed for venous thromboembolism (VTE) and bleeding risk before chemical VTE prophylaxis is ordered. Although objective scoring systems have been constructed few have been externally validated. Using data from patients admitted to medicine services over an 18 month period we provide external validation and assessment of accuracy for the IMPROVE Bleeding Risk Score.

METHODS: Data was collected as part of a large quality improvement project. Patients who met the eligibility criteria for the original IMPROVE study were selected and bleeding rates were analyzed. The IMPROVE bleeding risk score was calculated to test accuracy for predicting bleeding during hospitalization. We also abstracted data on all chemical prophylaxis administered, including dosage and duration.

RESULTS: Over the 18 month time period 1294 (77.6%) patients were admitted to a medicine ward and 374 (22.4%) to an intensive care unit (ICU). 981 (79.4%) of patients had an IMPROVE score < 7.0 and 254 (20.6%) had a score ≥ 7.0. Patients with an IMPROVE score ≥ 7.0 had significantly greater rates of major (3.9% vs 1.2%; p=0.01) and major plus clinically relevant non-major (5.1% vs 2.1%; p=0.02) bleeding at 14 days. Neither chemical prophylaxis in general nor specific agents affected the performance of the IMPROVE score for predicting major (OR 2.6, 95% CI: 1.1-6.1; p=0.03) or major plus clinically relevant non-major bleeding (OR 2.0, 95% CI: 1.0-4.0; p=0.06) at 14 days when assessed by cox-regression. Area and the curve for receiver operator characteristic (ROC) curve was 0.67 (95% CI: 0.57-0.77; p=0.01) and 0.64 (95% CI: 0.55-0.73) for IMPROVE predicting major and major plus clinically relevant non-major bleeding. For our data, a score of 7.0 provided a sensitivity of 43% and specificity of 81% for predicting a major bleed at 14 days, but a threshold of 5.5 showed maximum accuracy (57% sensitivity and 65% specificity).

CONCLUSIONS: We have provided external validation for the IMPROVE Bleeding Risk Score. Accuracy is good and thresholds can be adjusted based on desired sensitivity and specificity.

CLINICAL IMPLICATIONS: The IMPROVE Bleeding Risk Score can be used to assess bleeding risk at admission for patients with medical diagnoses.

DISCLOSURE: The following authors have nothing to disclose: Brian Foster, Elizabeth Marx, Sarah Petteys, Paul Clark, Jordanna Hostler, Joshua Mitchell, Jacob Collen, Aaron Holley

No Product/Research Disclosure Information




Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543