Recently we received the following publication in this Journal: The effect of Helicopter EmergencyMedical Services on trauma patient mortality in the Netherlands, by de Jongh et al. [1]. We have read this article with a higher than normal level of interest as this directly appeals to our field of work and a few serious questions were raised at (all) the four HEMS centres in the Netherlands.First of all, the title of the paper suggests that the study represents the national situation, while in fact it is a single center study and therefore these data cannot by definition be extrapolated to the whole country. This paper was submitted without consulting the two trauma centers with an HEMS which prehospitally treated the patients involved in this study. Furthermore, previous papers [2,3,4] from other trauma centers showed conclusions that are very different to this study. Since this article is based on research on the effectiveness of the HEMS in the Netherlands, and the research was performed neither with involvement, nor information of one of the four HEMS stations, we like to make some remarks on this article from the perspective of the four HEMS stations.In their observational study, the authors conclude that 'HEMS treatment is associated with a nonsignificantly higher risk of in-hospital mortality for patients with TBI and a non-significantly lower risk for patients without TBI'. The adjusted OR's are 1.3 (95% CI 0.6-2.7) and 0.9 (95% CI 0.3-2.5) and since both the 95% CI's contain 1 (as do all 95% CIs in their study), indeed there is no statistical significance. The authors use the words non-significantly higher/lower risk to indicate this in their statement. But when using statistics to test for significance of findings and no statistical significance is found, only speculative conclusions can be drawn.Regarding methodology, we would also like to suggest that, to assess a possible effect of HEMS+EMS vs EMS-only on trauma patient mortality, investigators should stratify or adjust for EMS 'scoop and run' to regional hospitals (of a lower level of care) for primary stabilization and subsequent secondary transport to the definitive care facility (level I trauma centre with neurosurgical service).Also, in case of HEMS+EMS, delivery to regional hospitals can be omitted since primary stabilization takes place at the scene or 'en route' and patients are directly delivered to the level 1 trauma centre.Excluded from analysis, as described in methods, are patients who were directly transferred from the ED to another hospital. However, the authors do not provide information on the number of patients that have been excluded for this reason, and, if any, for what reason. As in the introduction it is stated that the regional hospitals in the county (Noord-Brabant) are capable of resolving acute problems. But it is unclear from the data provided what the level of care of these hospitals is, and whether or not they are adequately staffed for trauma care 24/7 within the time interval given.Selective transfer of patients that o...