We evaluated tourniquet use in two ways. We categorized
patients by whether their tourniquets were applied
prehospital or in the ED, and also when they were placed
physiologically in relation to shock. ED tourniquet patients
were those patients who had a tourniquet first
placed on a limb in the ED; the other patients had a
tourniquet placed on a limb in the prehospital setting.
Shock was defined by medics or hospital providers
as a weak or absent radial pulse in an uninjured limb
without a tourniquet. Patients with tourniquets first
placed after the onset of shock were analyzed as “shock
present,” and all other patients were “shock absent”
before application of first tourniquet. This validated approach
is consistent with the clinical definitions used by
the Tactical Combat Casualty Care course, taught to all
military medics, and correlated with systolic blood pressures
as described by McManus et al. (14,15).
Survival rate was the primary outcome and morbidity
rate was the secondary outcome. We defined indicated
and appropriate use by the following criteria: indicated
use was medical (vessel lesion hemorrhage unresponsive
to a pressure dressing) or tactical for care under fire;
appropriate use entailed no misplacement (e.g., wrong
limb or distal to a wound), purposeful venous tourniquet,
or misuse (e.g., upside down).
Data were collected prospectively by the two on-site
investigators. Data were collected from the patients, their
providers or attendants, records, or medical reports such
as the morbidity and mortality reports. We had access to
electronic records of casualties. Data collected included
patient age, gender, application time (time between injury
and use) in minutes, setting of tourniquet application
(prehospital or ED), mechanism of injury, treatment (including
operative procedures, number of transfused units
[all blood products were summed]), injury severity
scores, abbreviated injury score (AIS), base deficit, systolic
blood pressure, international normalized ratio, initial
heart rate, injury description (e.g., traumatic ampu-