Please use this identifier to cite or link to this item: https://hdl.handle.net/1/2737
Title: Performance of risk assessment models for venous thromboembolism in critically ill patients receiving pharmacologic thromboprophylaxis: a post hoc analysis of the PREVENT trial
Authors: Al-Dorzi, Hasan M;Arishi, Hatim;Al-Hameed, Fahad M;Burns, Karen Ea;Mehta, Sangeeta;Jose, Jesna;Alsolamy, Sami J;Ann I Abdukahil, Sheryl;Afesh, Lara Y;Alshahrani, Mohammed S;Mandourah, Yasser;Almekhlafi, Ghaleb A;Almaani, Mohammed;Al Bshabshe, Ali;Finfer, Simon;Arshad, Zia;Khalid, Imran;Mehta, Yatin;Gaur, Atul ;Hawa, Hassan;Buscher, Hergen;Lababidi, Hani;Al Aithan, Abdulsalam;Al-Dawood, Abdulaziz;Arabi, Yaseen M
Affliation: Central Coast Local Health District
Gosford Hospital
Issue Date: 2-Sep-2024
Source: Online ahead of print
Journal title: Chest
Department: Intensive Care
Abstract: The diagnostic performance of the available risk assessment models for venous thromboembolism in critically ill patients receiving pharmacologic thromboprophylaxis is unclear. For critically ill patients receiving pharmacologic thromboprophylaxis, do risk assessment models predict who would develop venous thromboembolism or who could benefit from adjunctive pneumatic compression for thromboprophylaxis? In this post hoc analysis of the PREVENT trial, we evaluated different risk assessment models for venous thromboembolism (ICU-VTE, Kucher, Intermountain, Caprini, Padua, and IMPROVE models). We constructed receiving operator characteristic curves and calculated the sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios. Additionally, we conducted subgroup analyses evaluating the effect of adjunctive pneumatic compression versus none on the study primary outcome. Among 2003 patients receiving pharmacologic thromboprophylaxis, 198 (9.9%) developed venous thromboembolism. With multivariable logistic regression analysis, the independent predictors of venous thromboembolism were APACHE II score, prior immobilization, femoral central venous catheter, and invasive mechanical ventilation. All risk assessment models had areas under the curve <0.60 except for the Caprini model (0.64, 95% confidence interval 0.60, 0.68). The Caprini, Padua and Intermountain models had high sensitivity (>85%) but low specificity (<20%) for predicting venous thromboembolism, whereas ICU-VTE, Kucher, and IMPROVE models had low sensitivities (<15%), but high specificities (>85%). The positive predictive value was low (<20%) for all studied cutoff scores, whereas the negative predictive value was mostly >90%. Using the risk assessment models to stratify patients into high- versus low-risk subgroups, the effect of adjunctive pneumatic compression versus pharmacologic prophylaxis alone was not different across the subgroups (p for interaction >0.05). The risk assessment models for venous thromboembolism performed poorly in critically ill patients receiving pharmacologic thromboprophylaxis. None of the models identified a subgroup of patients who might benefit from adjunctive pneumatic compression.
URI: https://hdl.handle.net/1/2737
DOI: 10.1016/j.chest.2024.07.182
Pubmed: https://pubmed.ncbi.nlm.nih.gov/39232999
Publicaton type: Journal Article
Keywords: Intensive Care
Appears in Collections:Health Service Research

Show full item record

Page view(s)

12
checked on Oct 11, 2024

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.