Risk Factors for Postdischarge Venothromboembolism After Colorectal Resection.

BACKGROUND

Current guidelines recommend extended-duration thromboprophylaxis for all abdominal oncologic resections. However, other high-risk patients may benefit from extended thromboprophylaxis.

OBJECTIVE

The purpose of this study was to identify risk factors for postdischarge venothromboembolism after colorectal procedures.

DESIGN

This was a retrospective cohort study.

DATA SOURCES

The New York Statewide Planning and Research Cooperative System database (2005-2013) was the data source for this study.

STUDY SELECTION

Colon and rectal resections were evaluated. Cases with in-hospital mortality or length of stay ≥30 days were excluded.

MAIN OUTCOME MEASURES

Postdischarge venothromboembolism was defined at 30-days after the procedure requiring representation to the emergency department or hospital admission with a new diagnosis of venothromboembolism using International Classification of Diseases, Ninth Revision, codes. Factors associated with postdischarge venothromboembolism were then evaluated using a hierarchical bivariate analysis. A hierarchical mixed-effects model was created using a manual stepwise approach assessing variables meeting p < 0.1 on bivariate analysis.

RESULTS

Among 128,163 patients, postdischarge venothromboembolism occurred in 0.7% (n = 789) of the population. Multiple factors were associated with postdischarge venothromboembolism on bivariate analysis. On multivariable analysis, benign conditions requiring operative intervention remained at high risk, with ulcerative colitis imparting an 93% increased odds when compared with other resections (OR, 1.93 (95% CI: 1.30-2.86); p = 0.001). Advanced malignancies (stages III and IV) were associated with increased postdischarge venothromboembolism risk, whereas stage I and II malignancies were not. The only protective factor was a laparoscopic procedure (OR, 0.80 (95% CI: 0.67-0.95); p = 0.010). There was no significant difference in procedure type after controlling for primary diagnosis.

LIMITATIONS

This was a retrospective analysis of administrative data with inherent limitations. Only patients who presented with postdischarge venothromboembolism to a hospital within New York State were captured.

CONCLUSIONS

This study identifies risk factors for postdischarge venothromboembolism and suggests that ulcerative colitis increases risk for postdischarge venothromboembolism whereas Crohn's disease does not. Ulcerative colitis postdischarge venothromboembolism rates exceeded even those of malignancy, suggesting that a future study is necessary to determine the efficacy of extended duration thromboprophylaxis in high-risk benign conditions, such as ulcerative colitis.