Key result
A regression model incorporating mean arterial pressure, SpO2, pulse rate, and age predicted 30-day mortality with an AUC of 0.848, performing significantly better than the NEWS-2 score.
Why the study?
Complications including bleeding increase significantly with INR levels between 4.5 and 10, making it crucial to identify high-risk patients who require closer follow-up and monitoring.
Cohort (n=263)
No
Effect estimate: AUC 0.848 (95% CI 0.799-0.898)
p-value: p=<0.001
Mean arterial pressure, SpO2, heart rate, and age independently predict 30-day mortality in patients with warfarin-induced INR of 4.5-10, and a derived regression model outperforms the NEWS-2 score for risk stratification.
Purpose: Although it has been shown that vitamin K administration does not benefit patients with INR levels between 4.5 to 10, there are studies in the literature showing that some complications including the risk of bleeding in these patients increase significantly. For this reason, it is crucial to select high-risk patients who present with elevated INR to apply closer follow-up and monitoring. The primary objective of our study is to define the predictors for 30-day mortality of the patients with an INR between 4.5 to 10 due to warfarin-related overanticoagulation. The secondary objective of our study is to derive a regression model which can predict mortality in 30 days and to compare the performance of this model with the National Early Warning Score-2(NEWS-2). Materials and Methods: We included patients older than 18 years old, admitted between the dates 01.01.2016 - 01.01.2022 who are using warfarin as medication and with an INR between 4.5 – 10 in our study. We excluded patients with trauma, major bleeding on admission or patients with missing data. For the regression model, backward-wald stepwise method was utilized. We used the Hosmer-Lemeshow test for the goodness of fit. For the overall performance of the model, we evaluated the Nagelkerke R Square, and the Reciever Operating Characteristics test. DeLong test was used to compare the area under the curves (AUC). Results: A total of 263 patients were examined in the statistical analysis. Mean arterial pressure, SpO2, pulse rate, and age were the independent predictors of 30-day mortality. The model have classified 81.4% of the patients correctly. The AUC of the regression model was 0.848 (0.799 to 0.898). The sensitivity of the model as a tool for mortality prediction was 94.1%, specificity 66.5%, and accuracy 71.9%. The AUC of the NEWS-2 score for 30-day mortality was calculated as 0.782 (95%CI = 0.715 to 0.849). The difference between the AUCs of our model and the NEWS-2 score was statistically significant. Conclusion: Mean arterial pressure, SpO2, heart rate, and age were the independent predictors for the 30-day mortality of patients with an INR between 4.5 to 10 due to overanticoagulation because of warfarin medication. The regression model we derived showed good overall discrimination and performed significantly better than NEWS-2 score.
No takes yet. Share an insight, caveat, or question.
Mehmet Muzaffer İslam (2023) conducted a cohort in Warfarin-related overanticoagulation (INR 4.5-10) (n=263). Predictors of mortality (mean arterial pressure, SpO2, pulse rate, age) was evaluated on 30-day mortality prediction model performance (AUC) (AUC 0.848, 95% CI 0.799-0.898, p=<0.001). A regression model incorporating mean arterial pressure, SpO2, pulse rate, and age predicted 30-day mortality with an AUC of 0.848, performing significantly better than the NEWS-2 score.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: