Joshua B Brown1, E Brooke Lerner, Jason L Sperry, Timothy R Billiar, Andrew B Peitzman, Francis X Guyette. 1. Division of General Surgery and Trauma, Department of Surgery (J.B.B., J.L.S., T.R.B., A.B.P.), University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania; Department of Emergency Medicine (E.B.L.), Medical College of Wisconsin, Milwaukee, Wisconsin; and Department of Emergency Medicine (F.X.G.), University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania.
Abstract
BACKGROUND: Trauma activation level is determined by prehospital criteria. The American College of Surgeons (ACS) recommends trauma activation criteria; however, their accuracy may be limited. Prehospital lactate has shown promise in predicting trauma center resource requirements. Our objective was to investigate the added value of incorporating prehospital lactate in an algorithm to designate trauma activation level. METHODS: Air medical trauma patients undergoing prehospital lactate measurement were included. Algorithms using ACS activation criteria (ACS) and ACS activation criteria plus prehospital lactate (ACS+LAC) to designate trauma activation level were compared. Test characteristics and net reclassification improvement (NRI), which evaluates reclassification of patients among risk categories with additional predictive variables, were calculated. Algorithms were compared to predict trauma center need defined as more than 1 unit of blood in the emergency department; spinal cord injury; advanced airway; thoracotomy or pericardiocentesis; ICP monitoring; emergent operative or interventional radiology procedure; or death. RESULTS: There were 6,347 patients included. Twenty-eight percent had trauma center need. The ACS+LAC algorithm upgraded 256 patients and downgraded 548 patients compared to the ACS algorithm. The ACS+LAC algorithm versus ACS algorithm had an NRI of 0.058 (95% confidence interval [CI], 0.044-0.071; p < 0.01), with an event NRI of -0.5% and nonevent NRI of 6.2%. When weighted to favor changes in undertriage, the ACS+LAC still had a favorable overall reclassification (weighted NRI, 0.041; 95% CI, 0.028-0.054; p = 0.01). The ACS+LAC algorithm increased positive predictive value, negative predictive value, and accuracy. Over-triage was reduced 7.2%, while undertriage only increased 0.7%. The area under the curve was significantly higher for the ACS+LAC algorithm (0.79 vs. 0.76; p < 0.01). CONCLUSIONS: The ACS+LAC algorithm reclassified patients to more appropriate levels of trauma activation compared to the ACS algorithm. This overall benefit is achieved by significant reduction in overtriage relative to very small increase in undertriage. In the context of trauma team activation, this trade-off may be acceptable, especially in the current health care environment. LEVEL OF EVIDENCE: Therapeutic/care management study, level III; prognostic/epidemiologic study, level III.
BACKGROUND:Trauma activation level is determined by prehospital criteria. The American College of Surgeons (ACS) recommends trauma activation criteria; however, their accuracy may be limited. Prehospital lactate has shown promise in predicting trauma center resource requirements. Our objective was to investigate the added value of incorporating prehospital lactate in an algorithm to designate trauma activation level. METHODS: Air medical traumapatients undergoing prehospital lactate measurement were included. Algorithms using ACS activation criteria (ACS) and ACS activation criteria plus prehospital lactate (ACS+LAC) to designate trauma activation level were compared. Test characteristics and net reclassification improvement (NRI), which evaluates reclassification of patients among risk categories with additional predictive variables, were calculated. Algorithms were compared to predict trauma center need defined as more than 1 unit of blood in the emergency department; spinal cord injury; advanced airway; thoracotomy or pericardiocentesis; ICP monitoring; emergent operative or interventional radiology procedure; or death. RESULTS: There were 6,347 patients included. Twenty-eight percent had trauma center need. The ACS+LAC algorithm upgraded 256 patients and downgraded 548 patients compared to the ACS algorithm. The ACS+LAC algorithm versus ACS algorithm had an NRI of 0.058 (95% confidence interval [CI], 0.044-0.071; p < 0.01), with an event NRI of -0.5% and nonevent NRI of 6.2%. When weighted to favor changes in undertriage, the ACS+LAC still had a favorable overall reclassification (weighted NRI, 0.041; 95% CI, 0.028-0.054; p = 0.01). The ACS+LAC algorithm increased positive predictive value, negative predictive value, and accuracy. Over-triage was reduced 7.2%, while undertriage only increased 0.7%. The area under the curve was significantly higher for the ACS+LAC algorithm (0.79 vs. 0.76; p < 0.01). CONCLUSIONS: The ACS+LAC algorithm reclassified patients to more appropriate levels of trauma activation compared to the ACS algorithm. This overall benefit is achieved by significant reduction in overtriage relative to very small increase in undertriage. In the context of trauma team activation, this trade-off may be acceptable, especially in the current health care environment. LEVEL OF EVIDENCE: Therapeutic/care management study, level III; prognostic/epidemiologic study, level III.
Authors: Maarten J G Leening; Moniek M Vedder; Jacqueline C M Witteman; Michael J Pencina; Ewout W Steyerberg Journal: Ann Intern Med Date: 2014-01-21 Impact factor: 25.391
Authors: Richard A Falcone; Lynn Haas; Eileen King; Suzanne Moody; John Crow; Ann Moss; Barbara Gaines; Christine McKenna; David M Gourlay; Cinda Werner; David P Meagher; Lisa Schwing; Nilda Garcia; Deb Brown; Jonathan I Groner; Kathy Haley; Anthony Deross; Laura Cizmar; Rochelle Armola Journal: J Trauma Acute Care Surg Date: 2012-08 Impact factor: 3.313
Authors: Brian J Eastridge; Jose Salinas; John G McManus; Lorne Blackburn; Eileen M Bugler; William H Cooke; Victor A Convertino; Victor A Concertino; Charles E Wade; John B Holcomb Journal: J Trauma Date: 2007-08
Authors: Christian Waydhas; Markus Baake; Lars Becker; Boris Buck; Helena Düsing; Björn Heindl; Kai Oliver Jensen; Rolf Lefering; Carsten Mand; T Paffrath; Uwe Schweigkofler; Kai Sprengel; Heiko Trentzsch; Bernd Wohlrath; Dan Bieler Journal: World J Surg Date: 2018-09 Impact factor: 3.352
Authors: Andrew-Paul Deeb; Heather M Phelos; Andrew B Peitzman; Timothy R Billiar; Jason L Sperry; Joshua B Brown Journal: J Trauma Acute Care Surg Date: 2021-06-01 Impact factor: 3.697