Demographics and injury characteristics

Demographic data are presented in Table 1. One-hundred and nine patients were identified from ambulance (68 patients) and helicopter (41 patients) records. Twenty-three (23/109, 21%) patients were excluded due to not meeting inclusion criteria. Reasons for exclusion are shown on Fig. 1. Eighty-six patients were included in full data analysis. The number of cases each year is shown on Fig. 2. Twenty-six patients (26/86, 30%) had an ISS > 15. The breakdown of injury mechanism is shown in Table 2. Seventy-two patients (72/86, 84%) had an isolated extremity injury to which the tourniquet was applied, whilst the remaining patients had the tourniquet applied to an extremity injury in the context of polytrauma.

Table 1 Patient clinical and demographic characteristicsFig. 1

Flowchart showing patient inclusion and reasons for exclusion

Fig. 2

Frequency of indicated and non-indicated tourniquet use over the study period. Percentage of non-indicated/indicated is shown with each year

Table 2 Mechanism of injury for 86 patients with PHTQ appliedTourniquet

Tourniquet application data are presented in Table 3. There were 88 tourniquet applications to 88 extremities in 86 patients. In two patients (2/86, 2.3%), two tourniquets were applied to two different extremity injuries. Fourteen cases (14/88, 16%) had prolonged tourniquet time (> 120 min). Tourniquet was not effective (as defined as having pulses present or ongoing bleeding) in 12 cases (12/88, 14%).

Table 3 Tourniquet application data for 88 tourniquet applications to 88 limbsPrehospital interventions

Prehospital vital parameters are presented in Table 1. Sixty-six patients (66/86, 77%) had a shock index < 1. Thirteen patients (13/86, 15%) were intubated at the scene. Estimated blood loss was only recorded for 41 patients (41/86, 48%). Median blood loss for these cases was 1000 mL (350, 1000). Thirteen patients (13/86, 15%) received blood transfusion prior to arrival in ED. Median volume of fluid transfused prior to arrival in ED was 150 mL (0, 500).

Emergency department interventions and disposition

ED vital parameters and laboratory parameters are shown in Table 1. Sixty-six patients (66/86, 77%) of patients had a shock index < 1. Massive transfusion protocols were activated in 21 patients (21/86, 24%). Most patients (51/86, 59%) did not receive blood transfusion. Computed tomography angiography (CTA) was performed in 29 patients (29/86, 34%). Three patients (3/86, 3.5%) received no interventions in ED and went straight to the operating room. Seven patients (7/86, 8.1%) had wound irrigation and closure in ED and were discharged without admission.

Primary outcome

Tourniquet use was deemed non-indicated in 68 cases (68/88, 77%, 95%CI 67-86%). There was a statistically significant increase in the proportion of non-indicated to indicated tourniquet use during this period (p = 0.03). Forty-three extremities (43/88, 49%) did not have a potentially haemorrhaging arterial Injury. In cases with a potentially haemorrhaging arterial injury (44/88, 50%), haemostasis was able to be maintained after tourniquet removal using local non-surgical methods in 25 cases (25/44, 57%). In 20 of these cases (20/25, 80%), haemostasis was achieved with direct pressure and/or limb positioning alone. In five cases (5/25, 20%) haemostatic dressings were also used in conjunction with direct pressure and limb positioning.

Among the forty-four extremities (44/88, 50%) with confirmed arterial injury, four (4/44, 9.1%) had two arterial injuries. Presence of arterial injury was confirmed in operating room by arterial repair in 38 cases (38/48, 79%), and arterial ligation in 10 cases (10/48, 21%). CTA confirmed the arterial injury in 6 cases (6/48, 13%) prior to operative management. One patient who had a tourniquet applied to two different extremity injuries had confirmed arterial injury in one limb but not the other. In one case with open fracture, arterial injury on the extremity to which the tourniquet was applied was unable to be determined as the patient died in ED from other injuries prior to investigation following prehospital arrest from bilateral pneumothoraces.

Upper limb arterial injuries accounted for 35 cases (35/48, 73%). The most commonly injured arteries were the radial artery (16/48, 33%) and the ulna artery (13/48, 27%). The most commonly injured lower limb arteries were the popliteal artery (4/48, 8.3%) and the posterior tibial artery (4/48, 8.3%). Thirty-three arterial injuries (33/48, 69%) were classified as complete injuries, 10 (10/48, 21%) were classified as partial injuries, and four (4/48, 8.3%) were not classified in the operation report.

The most common injury in extremities without arterial injury was an isolated venous injury, accounting for 28 cases (28/43, 65%). Open fractures were the second most common, accounting for 12 cases (12/43, 28%). The remaining cases included two degloving injuries (2/43, 4.7%) and one case of a mangled extremity (1/43, 2.3%).

Interventions

Seventy-seven patients (77/86, 90%) required intervention in the operating room. Eighteen patients (18/86, 21%) required only exploration, irrigation, and debridement. Thirty-nine patients (39/86, 45%) required some form of arterial vascular intervention (ligation, repair, or shunt placement). Thirty patients required nerve repair (30/86, 35%). Nine patients (9/86, 10%) underwent amputation. Three patients (3/86, 3.5%) had replantation of a limb. Median number of operative interventions was 1 (1, 2). Seven patients (7/86, 8.1%) had unplanned return to operating room.

Complications

Complication rates are shown in Table 4. Thirty-three patients (33/86, 38%) had at least one recorded complication which may be attributed to tourniquet use. Among the 16 patients (16/86, 19%) with nerve palsy, 11 (11/16, 69%) had a nerve injury secondary to the primary injury documented on the operation report. Among the four cases (4/86, 4.7%) of deep vein thrombosis, two occurred in the extremity the tourniquet was applied to, and two were pulmonary embolisms without a confirmed primary location.

Table 4 Complication rates

Among the nine patients (9/86, 10%) that underwent amputation in operating room, two were formalisation of partial traumatic amputations (2/9, 22%), three were amputated secondary to unsalvageable limb with associated arterial injury (3/9, 33%), and four were amputated without any evidence of arterial injury due to the extensive tissue loss unlikely to lead to meaningful functional outcome (4/9, 44%). Among the patients who required amputation without arterial injury, median tourniquet time was 160 min (121, 191).

Subgroup analysis

Two post-hoc subgroup analyses were performed for patients without arterial injury and for patients who had extended tourniquet time (Table 4). For cases without arterial injury, 30 (30/43, 70%) had no documented attempt at haemorrhage control prior to tourniquet placement. Six cases (6/43, 14%) that did not have arterial injury had extended tourniquet time, with median total tourniquet time 145 min (126, 174). Fourteen patients (14/42, 33%) who had no arterial injury had a recorded complication. Eleven patients (11/13, 85%) with extended tourniquet time had a recorded complication.

Clinical outcomes

Clinical outcomes are presented in Table 4. Three patients died (3/86, 3.5%), two of whom died prior to intervention in operating room from their injuries (one in cardiac arrest on arrival and one who went into cardiac arrest in ED), and one who died after intervention on day two of admission due to multiple organ failure secondary to traumatic shock. Twenty-three patients (23/86, 27%) required ICU admission with a median ICU stay of two (1,7) days. Median length of hospital stay was 3 (1, 16) days. Length of hospital stay was 1 day or less in 27 patients (27/86, 31%). Thirty-nine patients (39/86, 45%) required blood transfusions during admission. Sixty-six patients (66/86, 77%) attended planned follow-up clinics. Two patients (2/86, 2.3%) discharged against medical advice.