ADVERTISEMENT
AVPU Instead of GCS For Trauma
Reviewed This Month:
Is AVPU Comparable to GCS in Critical Prehospital Decisions? A Cross-Sectional Study
Authors: Janagama SR, Newberry JA, Kohn MA, et al.
Published in: Am J Emerg Med, 2022
Any of us who have worked in the field know how challenging it can be to calculate a patient’s Glasgow Coma Scale (GCS) score without using a cheat sheet. However, most field triage guidelines require a GCS calculation to evaluate the patient’s level of consciousness and determine appropriate destinations for trauma patients. The AVPU (alert, verbal, pain, unresponsive) scale is also likely familiar to most reading this column as a method for evaluating patient level of consciousness. AVPU is easier to remember for field providers in high-stress environments. If AVPU could be performed similarly to GCS, this simpler assessment could potentially be useful for field trauma triage.
With that in mind, the authors of this month’s study sought to assess whether prehospital AVPU categorization correlated with GCS severity. Further, the authors examined the relationship between AVPU and mortality rates among trauma patients.
Parameters
This was a cross-sectional study that utilized a convenience sample of trauma patients transported by a single prehospital agency that provides emergency care across 17 states in India. Seven states participated in this study. The study period was from November 16, 2015 to January 30, 2016. The agency responds to approximately 25,000 calls a day. EMTs that work for this agency must complete 52 days of training. The authors note this training is comparable to the US National Highway Traffic Safety Administration EMT-Basic curriculum.
In the study region patients are typically cared for by EMTs and transported to the nearest appropriate government hospital. Government hospitals are categorized as primary, secondary, and tertiary. Primary hospitals function only during the day, do not provide intensive care services, and do not have operating rooms. Secondary hospitals function around-the-clock, provide some intensive care services, and have operating rooms. Tertiary hospitals are teaching hospitals that operate 24/7, uniformly provide intensive care services, and all have operating rooms.
Data collection began at 9 AM and ended at 5 PM from each Monday to Saturday during the study period. Data was collected by trained research assistants who identified trauma-related calls in real time and notified the EMTs about potential enrollment into the study. GCS was categorized as severe injury (≤8), moderate injury (9–12), and mild injury (13–15). Only cases with both AVPU and GCS documented were included in the analysis. Cases were excluded with 1) hospital-to-hospital transfers where the length of stay at the initial treating hospital was greater than 12 hours; 2) ambulance dispatch issues such as a dispatch cancellation, duplicate dispatch, or incorrect dispatch data; 3) research assistant inability to contact the EMT; 4) injuries that primarily involved burns, electrocution, lightning, smoke inhalation, or isolated animal bites; 5) no evidence of trauma/injury on EMT arrival; 6) EMTs caring for more than one trauma patient during a single response; 7) EMT transfer of the patient to another ambulance provider; and 8) patient death before the ambulance arrived.
The primary outcome measure was death within 48 hours, following transfer of care to the hospital, for each AVPU category. The secondary outcomes were the differences in median GCS within each AVPU category as well as a comparison of AVPU versus GCS in the ability to predict death within 48 hours.
Results
During the study period there were 2900 cases that had both AVPU and GVS included for analysis. The difference between the median GCS for verbal (12) and the median GCS for pain (10) was not statistically significant (p=0.18) after multiple comparisons. However, the GCS severity categorization was significantly correlated with each AVPU category (p<0.001).
Mortality data was available for 75% (2184) of cases. Of those 3% (61) died. When compared to alert patients, the odds of death were 20 times higher for patients responding to pain (OR 19.4; 95% CI 8.4–42.5) and 100 times higher for unresponsive patients (OR 101.6; 95% CI 50.5–209.9). There was a 4- to 5-fold increase in the odds of death as severity increased from alert to verbal to pain to unresponsive. Similarly, there was nearly a 9-fold difference in the odds of death for each GCS severity category from mild to moderate to severe. The probability of death was comparable at each type of destination hospital: primary (0% [0/129], 95% CI 0.0–3.6), secondary (2.9% [38/1324], 95% CI 2.1–4.0), tertiary (3.5% [10/286], 95% CI 1.8–6.5), and private (2.8% [12/423], 95% CI 1.5–5.0). The probability of death did not vary even among the prehospital responses that did not have destination hospital data (4.5% [1/22], 95% CI 0.2–24.9). The ability of AVPU (area under the ROC curve 79.7%; 95% CI 73.4–86.1) to predict death was similar to GCS (area under the ROC curve 81.5%; 95% CI 74.8–88.2).
Limitations to this study included the use of a convenience sample and the inability to evaluate mortality data for all patients. Nevertheless, this was an interesting study that suggested AVPU documented by EMTs can be used as a surrogate for GCS in the prehospital environment. AVPU was comparable to GCS in predicting the probability of death within 48 hours and correlated closely with GCS severity categories. The authors concluded that “using AVPU to assess the level of consciousness in prehospital trauma protocols may simplify their global application without impacting the overall quality of care.”
As always, I hope you have an opportunity to review this manuscript in its entirety. There are some useful tables and figures that are helpful in interpreting the findings.
ABOUT THE AUTHOR
Antonio R. Fernandez, PhD, NRP, is the principal research scientist at ESO and serves on the board of advisors of the Prehospital Care Research Forum at UCLA.
Comments
This sham study does not actually seem to show anything. There is no significant data. AVPU itself does not tell anything clinically significant about the patient, and the statistics used are flawed. AVPU is assessed as part of GCS determination (eyes opening corresponds with AVPU levels). Sure gcs might not be the best for trauma patients but stupidly simplifying it shows no benefit, even with these supposed statistics. The statistics showed AVPU, which has much less data than GPS correlated well with it. I can say in a similar fashion that a scale of alive or probably dead correlated well with a GCS score. It still tells me nothing about the patient and does not help me whatsoever as a clinician or a researcher not to mention, that the statistics cannot be run correctly since of four e they correlate well when one is integrated well into the other. This should never have been approved by a peer reviewed journal, nor should it have been published
—Ke Ju