Background: Colorectal cancer has the second highest mortality of any malignancy, and venous thromboembolism is a major postoperative complication.
Objective: This study aimed to determine the variation in incidence of venous thromboembolism after colorectal cancer resection.
Data sources: Following PRISMA and MOOSE guidelines (PROSPERO, ID: CRD42019148828), Medline and Embase databases were searched from database inception to August 2019 including 3 other registered medical databases.
Study selection: Two blinded reviewers screened studies with a third reviewer adjudicating any discordance. Eligibility criteria: Patients post colorectal cancer resection aged ≥18 years. Exclusion criteria: Patients undergoing completely endoscopic surgery and those without cancer resection. Selected studies were randomized controlled trials and population-based database/registry cohorts.
Main outcome measures: Thirty- and 90-day incidence rates of venous thromboembolism per 1000 person-years following colorectal cancer surgery.
Results: Of 6441 studies retrieved, 28 met inclusion criteria. Eighteen were available for meta-analysis reporting on 539,390 patients. Pooled 30- and 90-day incidence rates of venous thromboembolism following resection were 195 (95% CI, 148-256, I2 99.1%) and 91 (95% CI, 56-146, I2 99.2%) per 1000 person-years. When separated by United Nations Geoscheme Areas, differences in the incidence of postoperative venous thromboembolism were observed with 30- and 90-day pooled rates per 1000 person-years of 284 (95% CI, 238-339) and 121 (95% CI, 82-179) in the Americas and 71 (95% CI, 60-84) and 57 (95% CI, 47-69) in Europe.
Limitations: A high degree of heterogeneity was observed within meta-analyses attributable to large cohorts minimizing within-study variance.
Conclusion: The incidence of venous thromboembolism following colorectal cancer resection is high and remains so more than 1 month after surgery. There is clear disparity between the incidence of venous thromboembolism after colorectal cancer surgery by global region. More robust population studies are required to further investigate these geographical differences to determine valid regional incidence rates of venous thromboembolism following colorectal cancer resection.