Author, date and country | Patient group | Study type (level of evidence) | Outcomes | Key results | Study weaknesses |
---|---|---|---|---|---|
O’Reilly et al, 2014,1 UK | 1592 casualties over a 6-year period. Introduction of prehospital blood transfusion (PHBTX) by MERT-E in 2008 created two cohorts ie, a pre-PHBTX and post-PHBTX era. 97 patients were matched. 26.9% received a PHBTX of PRBC and FFP. | Retrospective matched cohort study | Mortality (no PHBTX vs PHBTX) | 19.6% vs 8.2% (p<0.001) | Losses to follow-up. Multiple potential confounders were identified: recipients of a PHBTX received more prehospital interventions (eg, chest decompression, advanced airway intervention, tranexamic acid, larger total blood infusions, improved ratios of PRBC: FFP and shorter prehospital times). No statistical analysis was used to control for these confounders. |
Holcomb et al, 2015,2 USA | 885 prehospital trauma patients transported by two different HEMS (LF and OA) operations. Comparison was made between cohorts of 716 patients with LF with available blood products and 169 patients with OA resuscitated with crystalloid only. 19% of the LF received a PHBTX of PRBC and FFP. | Retrospective cohort study | Mortality at 6 h among those with critical ED disposition (admitted directly to the ICU, IR, OR or morgue) | OR 0.23 (95% CI 0.0062 to 0.890; p=0.033) | Differences in critical care capabilities between the HEMS were not discussed in detail. Ground platforms are excluded from analysis due to ‘gross inequities’ and represent selection bias. No matching of patients was attempted in this study. LF shared its governance with the major trauma centre, possibly representing a conflict of interests. No breakdown in injury type. Marked differences in transport times. |
Mortality at 24 h | OR 0.57 (p=0.117) | ||||
Mortality at 30 days | OR 0.71 (p=0.441) | ||||
Brown et al, 2015,3 USA | 8616 prehospital trauma patients transport by air to a level 1 trauma centre. Matched cohort of 213 was created. 2.9% received a PHBTX of PRBC. | Retrospective matched cohort study | Survival at 24 h | AOR 6.32 (95% CI 1.88 to 21.14; p <0.01) | Single-centre study using a single HEMS operation. Potential for selection bias. Missing data. Initial large crystalloid infusions. Survival bias. PRBC transfusion only. |
Survival in hospital | AOR 4.32 (95% CI 0.76 to 24.72; p=0.10) | ||||
Brown et al, 2015,4 USA | 1415 civilian patients with blunt trauma transferred to a trauma centre. A matched cohort of 113 was created. 3.5% received a PHBTX of PRBC±plasma. | Retrospective cohort study | Mortality at 24 h | AOR 0.02 (95% CI 0.01 to 0.69; p=0.04) | Small numbers of transfusions. 2 h cut-off creating selection bias. Missing data. No description of the capabilities of the prehospital provider. No data regarding type of transfusions or ratios of blood products. Survival bias. Blunt trauma only. |
Mortality at 30 days | AOR 0.12 (95% CI 0.03 to 0.61; p=0.01) |
MERT-E, medical emergency response team; FFP, fresh frozen plasma; HEMS, helicopter emergency medical service; LF, life flight; OA, other agencies; ICU, intesive care unit; AOR, adjusted odds ratio; PRBC, packed red blood cells.