Background: Venous thromboembolism (VTE) is the third major cause of mortality in the world after heart attack and stroke. Most of patients admitted to ICUs are at high risk of VTE, especially with sedation, immobilization and mechanical ventilation. A number of studies have pointed out that woman present higher risk of VTE compared to men. That is why the prediction factors of such a risk should be analyzed seperatly according genders.
Aims: The purpose of this study is to develop a gender-specific computer-aided system to predict the risk of VTE complications, one hour early, for critically ill patients admitted to ICU.
Methods: All patients admitted to a 20 beds ICU between July 2019 and December 2020 were involved in the study. Exclusion criteria include patients with thromboembolic events prior to admission and patients having had other than a VTE-related event. Collected data included demographic characteristics, vital signs and laboratory measurements evaluated during the patient ICU stay. For each gender, the developed computer-aided system is based on logistic regression model to detect VTE complications risk one hour early.
Results: The performance of the developed computer-aided systems has been evaluated on a data set of 443 patients, 140 females (mean 63.8 +/- 17.3 years) and 303 males (56.36 +/- 19.8 years). Results show that the retained discriminant features are different for male and female. The AUC is higher for female than for male (85% against 71%). Moreover, the ability of the developed model to correctly predict a thromboembolic events reach 88% for females against 61% for males. The false negative prediction is higher for male (39% against 12% for females).
The performances of the developed computer-aided systems
Conclusions: These results demonstrate
(1) the performances of the computer-aided system to predict one hour early the risk of VTE complications which could be useful to provide adequate therapeutic strategy and
(2) the predictions model is gender-specific.
To cite this abstract in AMA style:
Labbene I, Mezghani N, Sellami W, Ghorbel A, Kadri F, Nouri N, Ferjani M. Gender-specific Risk Prediction of Venous Thromboembolism for Patients Admitted to Intensive Care Units (ICU) [abstract]. Res Pract Thromb Haemost. 2021; 5 (Suppl 2). https://abstracts.isth.org/abstract/gender-specific-risk-prediction-of-venous-thromboembolism-for-patients-admitted-to-intensive-care-units-icu/. Accessed December 6, 2023.« Back to ISTH 2021 Congress
ISTH Congress Abstracts - https://abstracts.isth.org/abstract/gender-specific-risk-prediction-of-venous-thromboembolism-for-patients-admitted-to-intensive-care-units-icu/