A clinical predictive model for post-hospitalisation venous thromboembolism in patients with inflammatory bowel disease

Abstract

Background: Patients with inflammatory bowel disease (IBD) are at increased risk of venous thromboembolism (VTE) during hospitalisation and potentially post-discharge. Aims: To determine the incidence and risk factors for post-discharge VTE in IBD patients and create a point of care predictive model to assess VTE risk. Methods: Hospitalised IBD patients were identified from our institutional discharge database between 2009 and 2016, and were assessed for VTE by chart review. Risk factors for VTE within 3 months of discharge were determined by univariable and multivariable logistic regression. A point of care model was created using variables from the univariate analysis with P < 0.05, and internally validated by bootstrap methods. Results: Sixty-six of 2161 eligible discharges (3%) were associated with VTE within 6 months of hospitalisation. The median time to event was 37 days (range 3-182 days). On multivariable analysis age >45 years (OR 3.76; 95% CI 1.80-7.89) and multiple admissions (OR 2.62; 95% CI 1.34-5.11) were independently associated with VTE risk. Our final model incorporated age >45 years, multiple admissions, intensive care unit admission, length of admission >7 days and central catheter and was able to discriminate between discharges associated with and without VTE (optimism-corrected c-statistic, 0.70; 95% CI 0.58-0.77). By limiting treatment to a high-risk group, extended thromboprophylaxis could be avoided in 92% of discharges with a miss rate of 1.6% (32/1982 discharges). Conclusion: Patients with IBD remain at risk of VTE after hospital discharge. Our model may help clinicians stratify which patients will benefit most from extended thrombophrophylaxis.