Page 101 - Journal of Special Operations Medicine - Summer 2015

in the military context is to be achieved, then active in- senior authors from multiple, longer deployments sug-
volvement of the surgeon will be essential. The authors gests that the activity level was not unusual.
consider it vital that deployed TNCs feel able to seek
clarification from the operating surgeon, while surgeons A further limitation is that our selection criteria focused
could contribute to improvement by describing injuries on patients likely to have more severe injuries. This
in language consistent with the dictionary when writing could lead to greater apparent disparity between physi-
the operative record. Together, the TNC and the sur- cians and registry staff, as external injuries are predomi-
geon should generate a consensus injury description to nantly minor and would be easily recognized as such by
describe the injury in terms of the most accurate “fit” both groups. However, these more complex injuries are
in the scoring dictionary. Existing collective predeploy- precisely those in which detail may be lost in the written
ment training provides an ideal opportunity to estab- record and, thus, are most relevant to our hypothesis.
lish collaboration by the inclusion of suitable casualty
scenarios. 29 This study did not verify the accuracy of the surgeons’
assessment of injuries, and none of our investigators had
A further potential contributor to our findings is that been formally certified in AIS methodology. However,
UK TNCs are trained for specific operations but rarely the UK surgical training syllabus requires competence in
have significant exposure to trauma data capture before the use of trauma scoring systems, and previous work
30
deployment and are unlikely to return subsequently to has found no significant difference in performance be-
that role. Detailed training in anatomy, radiology report tween trained and untrained raters. 18
writing, and injury description is required to undertake
the TNC role effectively. Establishing a Defence Medi- This study does not identify whether the difference
cal Services TNC cadre, working within the civilian UK in assigned scores arises from variability between in-
Major Trauma Networks, would allow participation in dividual raters, from documentation that does not
the ongoing UK Trauma Audit & Research Network convey a sufficient level of detail to permit accurate
(TARN) activity, leading to increased expertise in the retrospective coding, or from a different interpretation
matching AIS(M) descriptions to computed tomography of injuries arising from the different professional back-
scan reports, surgical operation notes, and postmortem grounds of physicians and registry personnel. Because
studies. While civilian and military wounding patterns our study was intended to compare the perceptions of
differ, such exposure would enable TNCs to deploy with the treating clinicians with registry output, it was not
greater experience in trauma audit and governance. possible to have the same clinicians score each case.
Logistical constraints meant the sample size was in-
A secondary finding of this study is that the JTTR may sufficient to perform multilevel model-based analysis
not accurately record mechanism of injury data when an of cross-classified data (not all cases assessed by all
31
IED strike to a military vehicle precipitates a rollover or reviewers) that could account for this. This should be
other accident in which patients sustain injuries typical studied further to refine registry process and maximize
of a more conventional motor vehicle incident. Finally, data quality and analysis.
the failure of investigators to identify and record all eli-
gible patients almost certainly relates to their primary Conclusion
responsibility of providing surgical care to the wounded.
This reinforces the importance of the deployed hospital This study highlights discrepancies in injury severity
establishment including registry personnel whose role is scoring between surgeons and TNCs, with surgeons
dedicated to information capture. This principle should recording more torso, limb, and external injuries, and
also apply to data capture for research projects in the TNCs including more injuries to the head and face. Be-
deployed environment. cause injury scoring is vital to the analysis of the efficacy
of interventions at all stages of the trauma care pathway,
Study Limitations from point of wounding to discharge from rehabilita-
Data collection for this study was conducted by two tion, it is important to minimize such differences. Both
British military surgical residents in the second half of a TNCs and treating clinicians should be actively involved
7-week deployment. This brief operational tour limited in the injury scoring of patients. Collaboration will im-
the sample size that could be collected, but it was con- prove both accuracy and reliability.
sidered sufficient for a preliminary investigation. It is
possible that this period was unrepresentative of normal Acknowledgments
activity and that greater casualty flow might have af-
fected the accuracy of physician coding by reducing the The authors thank the various consultant surgeons and
time available to match injuries to detailed descriptions radiologists who participated in this study. The Clinical
in the AIS(M) dictionary. However, the experience of the Information Exploitation Team and Defence Statistics

Variability in Injury Severity Scoring After Combat Trauma 91