Validating clinical threshold values for a dashboard view of the compensatory reserve measurement for hemorrhage detection
| Autor: | Mallory Wampler, Kevin K. Chung, John G. Myers, Tuan D. Le, Abdul Q. Alarhayem, Susannah E. Nicholson, Victor A. Convertino, Katie R Struck, Camaren Cuenca, Brian J. Eastridge, Michael Austin Johnson |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Adult
Male medicine.medical_specialty Population Hemorrhage Critical Care and Intensive Care Medicine Sensitivity and Specificity Machine Learning User-Computer Interface 03 medical and health sciences Injury Severity Score 0302 clinical medicine Risk Factors medicine Humans Blood Transfusion education education.field_of_study Blood Volume Receiver operating characteristic Vital Signs business.industry Trauma center Hemodynamics Shock 030208 emergency & critical care medicine Odds ratio Middle Aged Triage Confidence interval Logistic Models ROC Curve Emergency medicine Data Display Female Surgery Packed red blood cells business |
| Zdroj: | Journal of Trauma and Acute Care Surgery. 89:S169-S174 |
| ISSN: | 2163-0763 2163-0755 |
| DOI: | 10.1097/ta.0000000000002586 |
| Popis: | Background Compensatory reserve measurement (CRM) is a novel noninvasive monitoring technology designed to assess physiologic reserve using feature interrogation of arterial pulse waveforms. This study was conducted to validate clinically relevant CRM values with a simplified color-coded dashboard view. Methods We performed a prospective observational study of 300 injured patients admitted to a level I trauma center. Compensatory reserve measurement was recorded upon emergency department admission. Data collected to complement the analysis included patient demographics, vital signs, lifesaving interventions, Injury Severity Score (ISS), and outcomes. Threshold values of CRM were analyzed for predictive capability of hemorrhage. Results A total of 285 patients met the inclusion criteria. Mean age of the population was 47 years, and 67% were male. Hemorrhage was present in 32 (11%), and lifesaving intervention was performed in 40 (14%) patients. Transfusion of packed red blood cells was administered in 33 (11.6%) patients, and 21 (7.4%) were taken to the operating room for surgical or endovascular control of hemorrhage. Statistical analyses were performed to identify optimal threshold values for three zones of CRM to predict hemorrhage. Optimal levels for red, yellow, and green areas of the dashboard view were stratified as follows: red if CRM was less than 30%, yellow if CRM was 30% to 59%, and green if CRM was 60% or greater. Odds of hemorrhage increased by 12-fold (odds ratio, 12.2; 95% confidence interval, 3.8-38.9) with CRM less than 30% (red) and 6.5-fold (odds ratio, 6.5; 95% confidence interval, 2.7-15.9) with CRM of equal to 30% to 59% (yellow) when compared with patients with CRM of 60% or greater. The area under the receiver operating characteristic curve for three-zone CRM was similar to that of continuous CRM (0.77 vs. 0.79) but further increased the ability to predict hemorrhage after adjusting for ISS (area under the receiver operating characteristic curve, 0.87). Conclusion A three-zone CRM could be a potentially useful predictor of hemorrhage in trauma patients with added capabilities of continuous monitoring and a real-time ISS assessment. These data substantiate easily interpretable threshold dashboard values for triage with potential to improve injury outcomes. Level of evidence Diagnostic, level II. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|