Page 68 - Journal of Special Operations Medicine - Spring 2016

surgical elements in western Afghanistan from 17 Feb- SPSS Statistics, version 22 (IBM Corp.; http://www.ibm
ruary 2012 through 15 August 2012. However, due to .com). Vital signs were compared using a Pearson corre-
mission requirements, one of the sites was only able lation to compare individuals over time. Data are given
to participate until 12 April 2012. All patients treated as mean ± standard deviation unless otherwise indi-
at the three facilities were included; however, patients cated. A p-value of less than .05 was considered statisti-
without a NATO 9-line or MIST report were excluded cally significant. Triage accuracy was determined using
from subsequent analysis. This study was approved as a a kappa statistic. Interpretation of the kappa statistic (k)
performance improvement project by the Human Pro- was made using the method of Landis and Koch. 5
tections Administrator for the Joint Combat Casualty
Research Team.
Results
The NATO 9-line and MIST report for all medical evac- During the 6 months of the study period, a total of 179
uation requests were compared with the findings on ar- casualty evacuations were brought to the three Role
rival at the forward surgical element and entered into a II facilities (268 individual patients). Casualty evacua-
database. Information collected from the NATO 9-line tions carried a mean of 1.5 ± 0.9 patients (range, 1–6
and MIST report included the number of patients, triage patients). Demographics of the patients are presented in
category for each patient, mechanism of injury, suspected Table 1.
injuries, and vital signs (pulse, blood pressure, and respi-
ratory rate). If updates were given while en route, only Table 1 Demographics of Patients
information from the initial report was recorded. Characteristic No. (%)*
Mean age, years, ± SD 24.5 ± 10.1
Information collected after arrival at the Role II facility
included the number of patients, mechanism of injury, Male sex, n (%) 258 (97)
injuries, vital signs, Glasgow Coma Scale (GCS) score, Mechanism of injury, n (%)
and triage category. The first set of vital signs and GCS Blunt 70 (26)
score recorded in the medical record at the Role II facil- Penetrating 65 (24)
ity were entered into the database. Injuries were entered
into the database after evaluation and treatment of the Explosion/blast 117 (44)
patient to minimize the potential for missed injuries in Burn 9 (3)
the database. Patient demographics were also collected Other 7 (3)
after arrival at the Role II facility and included the pa- Revised Trauma Score ± SD 11.2 ± 2.0
tient’s age and sex. Injury severity was determined us-
ing the Revised Trauma Score, which is based upon the *Unless otherwise indicated. SD, standard deviation.
patient’s GCS score, systolic blood pressure, and respi-
ratory rate. These variables were taken from those re- After evaluation and treatment, 93 patients were clas-
3
corded on arrival to the Role II facility. sified as having been a true Priority I, 33 patients as
Priority II, 37 as Priority III, and four as Priority IV.
After care had been rendered, the treating provider at A triage category at the point of injury had only been
the Role II facility (generally the surgeon) was asked to assigned to 167 (93%) of the patients; the remainder
make an assessment of the patient’s injuries and how had no category assigned in the NATO 9-line report.
long they believed care could have been delayed without For those individuals triaged at the point of injury, 102
deleterious effects to the patient. This information was individuals (61%) were labeled as Priority I, 38 (23%)
referenced against FM 4-02.2, and the provider was were Priority II, 25 (15%) were Priority III, and two
4
asked to determine which triage category accurately de- (1%) were Priority IV. The point of injury triage was
scribed the patient. This was treated as the true triage accurate for 80% of individuals. However, 15% of ca-
category against which triage accuracy at the point of sualties were overtriaged and 6% undertriaged. In gen-
injury was measured. eral, triage accuracy was greatest at the two extremes of
categories (Figure 1). The k statistic for triage was 0.619
Statistical analysis was performed using available data. (p < .001), suggesting a substantial correlation between
As a NATO 9-line report was the only requirement for the point of injury triage and medical needs on arrival at
requesting evacuation, a MIST report was frequently the Role II facility.
absent. In the absence of a MIST report, only the num-
ber of patients and triage category were available for Each NATO 9-line report requested evacuation for be-
analysis. Case deletion was not performed if data were tween one and six patients, with a mean of 1.5 ± 0.9
missing, nor was imputation used to replace missing patient per request. There was excellent correlation (k =
values. Statistical analysis was performed using IBM .850) between the number of casualties reported at the

52 Journal of Special Operations Medicine Volume 16, Edition 1/Spring 2016