Validation of Risk Assessment Models of Venous Thromboembolism in Hospitalized Medical Patients

Greene, M. Todd; Spyropoulos, Alex C.; Chopra, Vineet; Grant, Paul; Kaatz, Scott; Bernstein, Steven J.; Flanders, Scott A.

doi:10.1016/j.amjmed.2016.03.031

Cited by 80 publications

(108 citation statements)

References 20 publications

(16 reference statements)

Supporting

Mentioning

101

Contrasting

Unclassified

Order By: Relevance

“…Until now, only four RAMs have been externally validated together using one dataset which showed moderate results . Yet, most models have only been validated once, using different datasets, which limits comparison of model performances.…”

Section: Discussionmentioning

confidence: 97%

“…To our best knowledge this is the only external validation study (and from an independent group, ie, not related to any of the original development studies) thus far, that validated all RAMs, specifically in medical patients. This is in contrast to the study of Greene et al in which only four models were validated. Furthermore, we adhered to the Transparent Reporting of multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) checklist …”

Section: Discussionmentioning

confidence: 99%

“…However, for high‐risk medical patients, thromboprophylaxis remains underused as shown in the ENDORSE study where medical patients at risk of venous thrombosis actually received thromboprophylaxis in only 48% of cases . Possibly, this underuse could be explained by (a) a wide range of RAMs, not all known to clinicans, (b) a lack of generalizibility for most RAMs, and (c) at most modest accuracy as has recently been shown in a validation study (c‐statistic range of 0.58‐0.64) …”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Validation of risk assessment models for venous thrombosis in hospitalized medical patients

Cobben

Nemeth

Lijfering

et al. 2019

Research and Practice in Thrombosis and Haemostasis

View full text Add to dashboard Cite

Background Risk assessment models (RAMs) have been developed to predict which medical patients are at high risk for venous thrombosis (VT) and hence, should receive pharmacological thromboprophylaxis. Possibly due to the limited performance of the current models, guidelines differ on thromboprophylaxis regimens. Objective In this study all known RAMS were externally validated in medical patients in order to guide physicians in their thromboprophylaxis strategies. Patients and Methods A literature search was performed to find all published RAMs for medical patients at risk for VT. Data from a large population based case‐control study (MEGA study) designed to identify risk factors for a first VT were used to externally validate each RAM. Results 516 patients and 40 controls met the inclusion criteria for being a medical patient out of 4956 cases and 6297 controls). The c‐statistic, sensitivity, specificity, negative, and positive predictive values were calculated for each RAM. The literature search yielded 12 RAMs, none of which showed good discrimination (area under the curve range 0.56‐0.66), high sensitivity or specificity. The negative predictive value was high in all the RAMs (range 98.6‐99.4) and positive predictive value was low (range 1.1‐3.4). Conclusion This study shows that the discriminative power of currently available RAMs to predict VT in medical patients is limited. Therefore, clinical judgment should supplement RAMs to identify high‐risk medical patients for thromboprophylaxis. These results underline the importance of improving the performance of these models.

show abstract

Section: Discussionmentioning

confidence: 97%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Validation of risk assessment models for venous thrombosis in hospitalized medical patients

Cobben

Nemeth

Lijfering

et al. 2019

Research and Practice in Thrombosis and Haemostasis

View full text Add to dashboard Cite

show abstract

“…Padua and IMPROVE (International Medical Prevention Registry on Venous Thromboembolism) are 2 validated scores for risk stratification of acute medically ill patients . These RAMs demonstrated modest performance in validation data sets …”

Section: Introductionmentioning

confidence: 99%

Machine learning to predict venous thrombosis in acutely ill medical patients

Nafee

Gibson

Travis

et al. 2020

Research and Practice in Thrombosis and Haemostasis

View full text Add to dashboard Cite

Background The identification of acutely ill patients at high risk for venous thromboembolism (VTE) may be determined clinically or by use of integer‐based scoring systems. These scores demonstrated modest performance in external data sets. Objectives To evaluate the performance of machine learning models compared to the IMPROVE score. Methods The APEX trial randomized 7513 acutely medically ill patients to extended duration betrixaban vs. enoxaparin. Including 68 variables, a super learner model (ML) was built to predict VTE by combining estimates from 5 families of candidate models. A “reduced” model (rML) was also developed using 16 variables that were thought, a priori, to be associated with VTE. The IMPROVE score was calculated for each patient. Model performance was assessed by discrimination and calibration to predict a composite VTE end point. The frequency of predicted risks of VTE were plotted and divided into tertiles. VTE risks were compared across tertiles. Results The ML and rML algorithms outperformed the IMPROVE score in predicting VTE (c‐statistic: 0.69, 0.68 and 0.59, respectively). The Hosmer‐Lemeshow goodness‐of‐fit P‐value was 0.06 for ML, 0.44 for rML, and <0.001 for the IMPROVE score. The observed event rate in the lowest tertile was 2.5%, 4.8% in tertile 2, and 11.4% in the highest tertile. Patients in the highest tertile of VTE risk had a 5‐fold increase in odds of VTE compared to the lowest tertile. Conclusion The super learner algorithms improved discrimination and calibration compared to the IMPROVE score for predicting VTE in acute medically ill patients.

show abstract

“…Most VTE‐RAM used worldwide are derived by expert consensus or regression analysis to identify risk factors and all suffer from one or more deficiencies (Greene et al ., ). There is no internationally accepted RAM in use.…”

mentioning

confidence: 97%

Are we assessing venous thromboembolism (VTE) risk appropriately for hospitalised medical patients? The National VTE Risk Assessment Tool versus Padua Prediction Score

2020

View full text Add to dashboard Cite

Validation of Risk Assessment Models of Venous Thromboembolism in Hospitalized Medical Patients

Cited by 80 publications

References 20 publications

Validation of risk assessment models for venous thrombosis in hospitalized medical patients

Validation of risk assessment models for venous thrombosis in hospitalized medical patients

Machine learning to predict venous thrombosis in acutely ill medical patients

Are we assessing venous thromboembolism (VTE) risk appropriately for hospitalised medical patients? The National VTE Risk Assessment Tool versus Padua Prediction Score

Contact Info

Product

Resources

About