Background An optimally structured prehospital trauma care system can reduce the serious consequences of many injuries. Whether all major trauma patients should be directed to a trauma centre (TC), bypassing closer lower-level hospitals, is a contentious issue. In particular, it is unclear if direct transport reduces risk of death.

Objective To explore the relationship between direct versus indirect transfer to a definitive care hospital on short-term survival following major trauma in Aotearoa-New Zealand (NZ).

Methods This prospective cohort study using administrative data analysed patients aged <85 years with major trauma (Injury Severity Score>12) attended by Emergency Medical Services (EMS) who were admitted to a TC, either directly or indirectly. Patients for whom there was no intermediary hospital closer than the TC (ie direct was the obvious option) were excluded. Propensity scores were obtained from a logistic regression model with directness of transport to definitive care as the outcome variable and all other available variables accessible at the time of EMS retrieval and considered to be related to mortality. Crude and adjusted mortality rate were estimated using a generalised linear Poisson regression model with a log-link function and robust standard errors.

Results Of 1,008 major trauma cases meeting the eligibility criteria, 370 (36.7%) had pathways to definitive care that involved one intermediary hospital. Similar percentages of direct and indirectly transported patients died within 30 days following the EMS call (8.9% and 10.0% respectively).

The propensity-weighted adjusted model estimated an 8% lower 2-week mortality (95% CI -41%, 44%) for those that were transported indirectly compared to directly. For 30-day mortality, the adjusted relative risk estimate was 21% lower mortality (95% CI -48%, 19%) for those transported indirectly.

Conclusion Study findings suggest that in NZ, major trauma patients secondarily transferred to TCs may have decreased mortality when compared to directly transported patients, although there was considerable uncertainty with these estimates. Residual confounding from the observational study design is a limitation as is the relatively small sample size.