Page 164 - Ultrasonography in Prehospital and Emergency Medicine - Rein Ketelaars

162
Chapter 7
Therefore, we speculate that PHUS rarely changed but more likely supported the physician’s decision that transferring care to the ambulance paramedics was adequate. When a lower level of care was planned, PHUS was probably more supportive of this plan than that it was the sole cause for the plan to change. When these supported decisions are left out of the total, decisions are impacted in 132 (8.6%) of the PHUS examinations and in 126 (8.4%) of the patients. Although changing to lower level of care results in earlier re-availability of the HEMS team for subsequent scrambles or other patients in a multi-victim situation, and reduced operational costs, the reassurance might be justified only temporarily. Therefore, it may be unwise to rely on PHUS findings alone. We share Stengel’s concern and therefore argue it might be prudent to change emergency (prehospital) protocols into a standard two- time abdominal PHUS regime. Such a regime, however, may not be feasible in the time-crit- ical prehospital situation where other procedures, including transportation, have priority. In many cases, additional emergency department imaging is appropriate.
Strength and Limitations
The large sample of PHUS examinations (n = 1631) and the large subpopulation transported to our trauma center (n = 594) is a major strength of the study. Another strength is the way we handled the missing data on the PHUS-specific notes (970 out of 1631) resulting in only 92 cases being excluded.
There are also significant limitations. The conclusions from this single center study might not apply to other HEMS operations. The retrospective design and the database not designed for the study may have led to another standard of record keeping and recall bias; abdominal US examinations might have been performed but not recorded. Not every physician was trained in PHUS from day one leading to less ultrasound examinations in the first years of the study period. Experience of the physicians with PHUS is initially limited and is expected to increase over time. This might influence the diagnostic performance although we couldn’t demonstrate an improvement some time after the introduction of PHUS in our service. Such an effect is probably blurred by the ever-changing pool of active HEMS physicians.
The impact of US on treatment decisions was self-reported by the physicians. This introduc- es a high risk of bias. For follow-up studies, efforts to reduce the extent of bias should be undertaken. Furthermore, the retrospective design allows for the occurrence of selection bias even though this was minimized through our methods.