Discussion

This study described the similarities and differences in treating high-energy femur fractures in level I and level II trauma centers in Israel. The main findings were that epidemiology regarding age; mechanism and severity of injury; associated injuries and comorbidities were similar among the two hospital categories. We also found similarities regarding the fracture types: method of fixation and intraoperative complication rates. Differences between the two hospital types were found in transport methods to the hospital; time from admission to surgery; staged versus simultaneous surgery and surgeon skill level. No differences were found regarding complications or reoperation rates between the two centers.

The similarities and differences in trauma centers in this study seem counterintuitive, especially in regard to the severity of trauma dictating evacuation to trauma centers, which was the model used for the establishment of trauma systems in Israel. These results, however, do not reflect the data published by the Israel National Trauma Registry.10 According to this recent report, dealing with more than 36 000 trauma admissions, the number of patients with an ISS>16, patients who were hospitalized in ICUs and overall mortality, were all significantly higher in LITCs. The likely reason for this contradiction is that the study may be underpowered due to a relatively small sample size. Only about 34% of the patients from both study groups were with an ISS of 16 and above, up to a total of 80 of the 235 patients. Another finding regarding the Israel triage system is that most ICU ambulances and helicopters transport their patients to LITCs; however, this fact did not affect the patient population, as mentioned earlier. A previous study dealing with evacuation of patients from multicasualty events in Israel11 had shown some severe triage errors in the emergency evacuation to trauma centers. However, it is noteworthy to mention that some ‘level II’ trauma centers in Israel are well equipped and staffed, and are short of only one or two criteria to become an LITC, and can therefore offer similar care for many patients akin to that received at an LITC.

A significant difference between the two types of trauma centers is shorter time from injury to surgery and more simultaneous surgical procedures being done in LITCs. This can be explained by the availability of additional operating rooms and more on-call surgical disciplines (eg, neurosurgery, maxillofacial surgery, cardiothoracic surgery) in LITCs, as mandated by regulations.6 This fact, however, did not significantly alter the outcome in our group of patients as regards morbidity, mortality, or outcome of femur fracture fixation.

An interesting finding was that unsupervised residents performed about 47% of procedures in LITCs in contrast to only 21% in RTCs, again without significant differences in fracture healing or postoperative complications. A significant correlation with unsupervised resident surgery was the time of day (ie, after 15:00 and before 07:00 ). This topic is not thoroughly dealt with in the literature. In a 1999 study from Edinburgh,12 around 50% of all orthopedic trauma procedures were performed without the attendance of a senior surgeon and in an additional 20% the senior trauma surgeon was present but not scrubbed. There was no mention of different outcomes between supervised and unsupervised procedures. In an Australian study, where more than 6000 orthopedic trauma procedures were studied, 59% were unsupervised.13 Surprisingly, complications such as hardware failure and malunion were more frequent in the supervised group. In contrast to these studies, two more recent studies do not support these conclusions. In one study comparing a level II teaching trauma center with a non-teaching hospital with similar resources and activities14 where residents were present in the trauma center, severe errors in triage, resuscitation, and adverse outcomes were significantly more common. In a recent study dealing with proximal femoral fracture in the elderly, mortality and complications were significantly higher in patients treated without the supervision of a scrubbed senior surgeon.15

In our study, despite the high number of associated injuries, 66% of the patients were with an ISS<16, with an average age of 33, and therefore were more resilient than the average polytrauma patient. Another explanation is the fact that femur shaft fractures treated with antegrade, reamed, locked intramedullary nails have extremely high union rates.16 ,17 Since about half of the fractures were of the more simple type, the results may be more ‘forgiving’ for less trained surgeons.

There are some limitations in our study. First, despite the design, it seems to be underpowered given the low observed complication rate observed in actual fracture surgery. The fact that there were no observed differences in patient profiles aforementioned strengthens this point. A more meticulous analysis of the results, such as measuring malunions in terms of length and rotation, or evaluation of intraoperative blood loss, may have yielded different results. Also, a more detailed analysis of reduction quality and outcome regarding extension of the fracture into the joints (hip, knee), as has happened in many cases, might have influenced the study. The fact that since the time of this study several dedicated orthopedic trauma surgeons were introduced to some level I centers may change the results if this study were to be repeated today.

In conclusion, femoral shaft fracture can be successfully treated both in LITCs and in RTCs in the State of Israel. Some further additional work, both in research and policy implementation, is required in order to define the minimal training level of surgeons performing these procedures. A more refined outcome analysis and triage criteria for emergency services treating patients with trauma with femoral shaft fractures are needed.