Comparison of three risk assessment models for PICC-related thrombosis in patients with cancer: A prospective cohort study

Abstract

Background:

Peripherally Inserted Central Catheters-Related Thrombosis (PICC-RT) is a challenging complication associated with PICC in cancer patients. Risk assessment models (RAMs) help identify patients at high risk of PICC-RT. This study compares the predictive performance of the Michigan RAM, the Caprini RAM, and the Maneval RAM in cancer patients with PICCs. The results could help recommend an optimal RAM for PICC-RT risk assessment in practice.

Methods:

A prospective observational study was conducted in a large cancer center in Guangzhou, China. 281 patients with cancer who underwent PICC insertions were enrolled from April 2023 to January 2024. The extracted data included basic information on patients and catheters, the scores of the Michigan, Caprini, and Maneval RAMs, and the occurrence of PICC-RT. The sensitivity, specificity, and area under the receiver operating characteristics (AUC) values were used to compare the predictive performance of these three RAMs.

Results:

275 participants were finally included for data analysis. Eighteen patients (6.5%) developed symptomatic PICC-RT. The average time between catheter insertion and PICC-RT diagnosis was 83.40 ± 34.63 days (range 28–143 days). The sensitivity/specificity of the Michigan, Caprini, and Maneval RAM were 0.06/0.97, 0.44/0.78, and 0.89/0.68, respectively. The AUC of the Maneval RAM (0.85, 95% CI: 0.78–0.91) was higher than that of Michigan (0.50, 95% CI: 0.37–0.64) and Caprini RAM (0.62, 95% CI: 0.47–0.76).

Conclusions:

Maneval RAM has a better predictive value for PICC-RT than Michigan and Caprini RAM. Maneval RAM is recommended for predicting the risk of PICC-RT in patients with cancer.