Literature DB >> 31248583

Quick risk assessment profile (qRAP) is a prediction model for post-traumatic venous thromboembolism.

Jotaro Tachino1, Kouji Yamamoto2, Kentaro Shimizu3, Ayumi Shintani2, Akio Kimura4, Hiroshi Ogura3, Takeshi Shimazu3.   

Abstract

OBJECTIVE: The Risk Assessment Profile (RAP) score is used as a tool of risk prediction in post-traumatic venous thromboembolism (VTE), but this scoring system is complicated to use in clinical settings due to its many variables. The objective of this study was to validate the utility of the RAP model and to develop a simpler risk prediction model for post-traumatic VTE.
METHODS: We conducted an observational study at two emergency and critical care centres in Japan between 2013 and 2016. Consecutive adult trauma patients who survived for 24 h or more after admission to the hospital were enroled. One prediction model (quick RAP model) was created with 6 variables based on clinical utility, experience, and thrombogenic mechanism from 17 variables in the conventional RAP model. We calculated diagnostic performance with 95% confidence interval (95% CI) by exact method.
RESULTS: We identified and analysed 859 patients. Twenty-six patients (3.0%) had VTE (17 with deep venous thrombosis alone, 2 with pulmonary embolism alone, and 7 with both). In the external validation, the RAP model had a sensitivity of 100% (95% CI, 86.8-100%) and specificity of 37.9% (95% CI, 34.6-41.3%). In contrast, the qRAP model had a sensitivity of 96.2% (95% CI, 80.4-99.9%) and specificity of 56.2% (95% CI, 52.7-59.6%). In the internal validation, receiver-operating characteristic curve analysis showed that the two models had similar area under the curve values that were not significantly different (0.832 and 0.800, respectively; RAP vs qRAP, p = 0.477).
CONCLUSIONS: We developed a practical, modified predictive model for VTE, the qRAP model, which appeared only slightly less accurate than the conventional RAP model and had the advantage of being simpler to use to predict VTE. In our dataset, the conventional RAP model was also evaluated as useful for the prediction of post-traumatic VTE.
Copyright © 2019 Elsevier Ltd. All rights reserved.

Entities:  

Keywords:  Post-traumatic venous thromboembolism; Prediction model; Risk assessment profile

Mesh:

Year:  2019        PMID: 31248583     DOI: 10.1016/j.injury.2019.06.020

Source DB:  PubMed          Journal:  Injury        ISSN: 0020-1383            Impact factor:   2.586


  1 in total

1.  Predicting venous thromboembolism in hospitalized trauma patients: a combination of the Caprini score and data-driven machine learning model.

Authors:  Lingxiao He; Lei Luo; Xiaoling Hou; Dengbin Liao; Ran Liu; Chaowei Ouyang; Guanglin Wang
Journal:  BMC Emerg Med       Date:  2021-05-10
  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.