Development of a risk score to predict peripherally inserted central catheter thrombosis in active cancer

Abstract Background Peripherally inserted central catheter (PICC) thrombosis is common. Aims To explore the prevalence of symptomatic PICC thrombosis and pulmonary embolism (PE)/deep vein thrombosis (DVT) in cancer and non‐cancer cohorts. In active cancer, we assessed the Khorana risk score (KRS) and Michigan risk score (MRS) for predicting PICC thrombosis and modifications to improve discriminative accuracy. Methods We reviewed consecutive cancer patients receiving chemotherapy through a PICC inserted April 2017 to July 2018. For each case, we identified a contemporaneous non‐cancer control. Results Among 147 cancer patients, median age 64 years, PICC duration 70 days (range, 2–452), 7% developed PICC thrombosis (95% confidence interval (CI) 3.6–12.2) and 4% (95% CI 2–9) PE/DVT. Among 147 controls, median age 68 years, PICC duration 18.3 days (range, 0.5–210), 0.7% (95% CI 0–4) developed PICC thrombosis and 2% (95% CI 0.4–6) PE/DVT. In our cancer cohort, no KRS < 1 patients developed PICC thrombosis (95% CI 0–11) compared with 9% (95% CI 5–16) in KRS ≥ 1 ( P = 0.12). PICC thrombosis occurred in 4.7% (95% CI 1.5–11.7) MRS ≤ 3 compared with 10.9% (95% CI 4.1–22.2) MRS > 3 ( P = 0.32). The addition of thrombocytosis, a variable from KRS, to MRS (modified MRS (mMRS)) improved discriminative value for PICC thrombosis (c‐statistic MRS 0.63 (95% CI 0.44–0.82), mMRS 0.72 (95% CI 0.58–0.85)). PICC thrombosis occurred in 1.4% (95% CI 0–8.3) mMRS ≤ 3 and 11.8% (95% CI 6.1–21.2) mMRS > 3 ( P = 0.02). More patients were categorised as low risk using mMRS ≤ 3 (47%) than KRS < 1 (22%). Conclusion Cancer patients had longer PICC durations and higher PICC thrombosis rates than those without (7% vs 0.7%). mMRS more accurately classified low PICC thrombosis risk than KRS <1(47% vs 22%). Prospective validation of mMRS is warranted.