The Ottawa score performs poorly in cancer patients with incidental pulmonary embolism

Background: The Ottawa score was previously developed to predict recurrent venous thromboembolism (VTE) in cancer patients with VTE. The performance of this score in patients with incidental VTE is currently unclear. Aim: To evaluate the performance of the Ottawa risk score in cancer patients with incidental pulmonary embolism included in an international, prospective, observational cohort study. Methods: The score was used to classify patients as high (≥1), intermediate (0), or low risk (≤−1). The discriminative performance of the score was estimated by calculating the cumulative incidence of recurrent VTE for all groups, the time-dependent c-statistic, and the sub-distribution hazard ratio (SHR), using a competing risk approach. Results: Of the 691 patients for which the Ottawa score could be calculated, 25 (3.6%) had recurrent VTE during 6-month follow-up and 38 (5.5%) during 12-month follow-up. The c-statistics of the continuous score at 6 and 12 months were 0.45 (95% CI, 0.36–0.54) and 0.51 (95% CI, 0.46–0.59), respectively. The 6-month cumulative incidences of recurrent VTE for those at low, intermediate, and high risk were 3.9% (95% CI, 1.5–8.4), 3.6% (95% CI, 1.9–6.2), and 3.6% (95% CI, 1.8–6.5), respectively. A sensitivity analysis restricted to the on-treatment period yielded similar results. None of the Ottawa risk score items were significantly associated with recurrent VTE. Conclusion: In cancer patients with incidental pulmonary embolism, the Ottawa risk score has a poor predictive value for recurrent VTE, which does not support the use of the score in this patient population.