Secondary Logo

Journal Logo

Original Articles

Improved survival in UK combat casualties from Iraq and Afghanistan


Penn-Barwell, Jowan G. MB ChB; Roberts, Stuart A.G. MB ChB; Midwinter, Mark J. CBE MD; Bishop, Jon R.B. PhD

Author Information
Journal of Trauma and Acute Care Surgery: May 2015 - Volume 78 - Issue 5 - p 1014-1020
doi: 10.1097/TA.0000000000000580


The decade since the invasion of Iraq in 2003 (Operation Telic) and the subsequent widening of operations in Afghanistan (Operation Herrick) has resulted in a sustained level of casualties not seen by the UK Military since the Korean War, when the United Kingdom suffered 1,078 killed in action, 2,674 wounded, and 1,060 missing or taken prisoner.1

The UK Defence Medical Services (DMS) have taken a deliberate approach to continuous performance improvement by measurement, identification, and promulgation of effective techniques and serially developing medical training to reflect recent practice to improve trauma team preparation. This process is collectively known as the Major Trauma Audit for Clinical Effectiveness (MACE).2

During the decade of conflict, the UK DMS trauma system, through a process of critical analyses of injuries and outcomes, has evolved to include advanced “buddy-buddy” care, where fighting troops deliver first responder assistance to a wounded colleague (including use of hemostatic dressings and application of tourniquets); the prehospital deployment of physicians delivering advanced resuscitation skills and decision making to the point of wounding via retrieval helicopter (e.g., airway interventions, including rapid sequence induction and intubation, transfusion of blood and fresh frozen plasma, ability to perform thoracostomies); field hospital damage-control resuscitation and damage-control surgery protocols with measures to recognize and mitigate against acute traumatic coagulopathy; continuing intensive care in flight during repatriation to the United Kingdom and integrated National Health Service delivered continuing care in a single center concentrating experience and expertise in contemporary injuries of combat.

This study sought to examine and report the temporal changes in the injury patterns of UK Military casualties from the conflicts of Iraq and Afghanistan during the last decade. The secondary aim was to determine changes in survival following combat trauma during this period.


The study was registered and approved by the Joint Medical Command institutional review process. The UK Military Joint Theatre Trauma Registry (JTTR) is an electronic database of prospectively gathered information on all casualties collected by trained trauma nurse coordinators working both in deployed medical facilities in Iraq and Afghanistan and in the Royal Centre for Defence Medicine (RCDM) in Birmingham, United Kingdom. Information on those killed in action is provided following post mortem. All fatalities and traumatically injured casualties that trigger a “trauma alert” on presentation to deployed UK medical facilities or subsequently requiring return to the UK following injury are included. The military definition of casualty to encompass those both killed and injured is used throughout this study. The database is managed by the Clinical Information and Exploitation Team and administered by UK Defence Statistics.2

The JTTR was searched for all UK casualties injured or killed in Iraq and Afghanistan between 2003 and 2012. Injuries were coded between 2003 and 2007 using the 1998 revision3 of the Abbreviated Injury Scale (AIS), which was also used to define the distribution of injuries about specific body regions.3 From 2007 to 2012, a military specific version of the AIS was adopted, and the new weighted scoring was retrospectively applied, so all injury data in this study were based on AIS 2005-Military scores.5 New Injury Severity Scores were used as an anatomic measure of injury.6

Personnel years at risk (PYAR) were calculated between 2008 and 2012 from UK Defence Statistics data. This was based on computerized records of every day spent in either of the two operational theaters by each service person; these figures were summed for each calendar year and divided by 365 to give the PYAR, that is, the equivalent number of personnel deployed for 12 months. For 2003 to 2007, detailed pay records were not available; hence, the information was extrapolated from the Ministry of Defence figures on troop levels contained in memoranda to the UK Parliament and was regarded as less precise.7,8 All PYAR figures exclude Special Forces because possible Special Forces deployments are not released by the Ministry of Defence.

Statistical Analysis

Data were grouped into calendar year cohorts according to date of injury, and logistic regression was used to examine relationships between year of injury and specific variables. In all models, year of injury was coded as a continuous variable on the range 1 to 10 corresponding to years 2003 to 2012.

In Model 1, multinomial log linear regression was used to examine the distribution of total recorded injuries across body regions by year of injury. The categorical outcome variable, body region of injury, was coded with “abdomen” set as the reference level. Year of injury was modeled using a restricted cubic spline to allow for flexible nonlinear relationships between time and region of body injury. Model fitting of this complex data was performed using a quasi-Newton optimization method because of the potential for convergence problems using standard optimization algorithms. Injuries to body regions were assumed to be clustered by individual, and SEs were estimated using a cluster bootstrap approach based on 10,000 samples to account for this. In Model 2, New Injury Severity Score (NISS) was included as a continuous variable as both a main effect and as part of an interaction term with year of injury. The interaction between year of injury and NISS was statistically significant (p = 0.009). In each model, year of injury was modeled using restricted cubic splines to allow for flexible relationships. Model selection was based on Akaike information criterion.9 Logistic regression was used for this analysis, and fitting was performed using maximum likelihood estimation. In Model 2, the reference level of the outcome variable was coded as “fatality” (vs. “survival”).

Analyses were conducted using R and the libraries: stats,10rms,11effects,12 and nnet.13


The JTTR recorded 2,792 casualties injured or killed during service in Iraq and Afghanistan. These consisted of all those who killed or were injured following trauma. The mean (SD) age was 25.7 (5.9), and 2,746 (99%) were male. Of these casualties, 2,227 (80%) were a result of hostile action, with the remaining 565 (20%) resulting from incidents not involving enemy forces, for example, road traffic collisions. There were 608 fatalities (22% of all casualties) during this decade. The distribution of casualties and fatalities throughout the 10-year study period is shown in Table 1.

Number of Fatalities and Injured Survivors per Year in Respective Conflicts

The most common mechanism of injury was caused by explosive weapons, causing 1,592 casualties, representing 56% of the total casualties and 65% of those from hostile action. Gunshot wounds (GSWs) were the next most significant mechanism of injury, being the cause of 684 casualties, 28% of total casualties, and 31% of those from hostile action. Aside from the increased proportion of GSWs in 2003 during the invasion of Iraq, the relative proportion of casualties of GSW to injuries from explosive weapons remained approximately consistent as the conflicts in Iraq and Afghanistan evolved at approximately 1:3.

There were 14,071 injuries sustained in the 2,792 casualties over the study period, distributed across body regions as shown in Table 2. The extremities were the most commonly injured body regions, constituting 43% of all injuries. The relative distributions of injuries affecting the abdomen, thorax, spine, face, neck, and upper extremity remained relatively constant during the study period.

Injured Regions as Defined by AIS System5 per Year

Observed proportions of injuries by body region and year of study were analyzed for temporal trend using the Cochran-Armitage test. Although there is evidence that the proportion of injuries by body region is associated with year (p < 1e-11), the Cochrane-Armitage test is restricted to testing for linear monotonic trends. We wished to examine how the relative distribution of injuries across body regions has changed over time. Models that included nonlinear functions of time provided an improved fit to the data and were used for the analyses.

The relative risk ratio (RR) of sustaining an injury to the head, relative to sustaining an injury to the abdomen (which remained relatively constant), changed by a factor between 0.87 (95% confidence interval [CI], 0.78–0.97) and 0.79 (95% CI, 0.68–0.92) per year from 2006 to 2010. Relative risk ratios for head injuries for all unit changes in year are presented in Supplemental Digital Content 1 ( The RR of sustaining an injury to the lower extremity, relative to the abdomen, remained statistically indistinguishable from 0 between 2003 and 2010. The RRs changed by 1.44 (95% CI, 1.25–1.68) from 2010 to 2011 and by 1.80 (95% CI, 1.41–2.30) from 2011 to 2012. RRs for lower extremity injuries for all unit changes in year are presented in Supplemental Digital Content 2 ( The predicted probabilities from Model 1 display the negative trend in the proportion of head injuries (Supplemental Digital Content 3, and the positive trend in the proportion of lower extremity injuries (Fig. 1) during the study period. Predicted probabilities and corresponding 95% CIs for receiving an injury to a specific body region for each year are presented in Supplemental Digital Content 4 (

Figure 1:
Distribution of upper and lower extremity injuries over time as proportion of total injuries. Shaded regions denote 95% CIs about the predicted values obtained from the multinomial logistic regression (Model 1). Dots denote observed proportions.

The odds of surviving a given injury severity was examined by analyzing NISS as a continuous variable (Model 2). Analyses demonstrate a consistent improvement in survival year-on-year during the decade of the study as shown in Figure 2. Model 2 can be used to estimate the NISS value associated with a 50% probability of fatality for each year in the study period. The estimated probabilities of survival were obtained from Model 2 for every possible NISS value in each year. The smallest NISS value with a corresponding 95% CI lower limit that exceeds 50% was identified in each year. This 50% survival NISS value rose from 32.5 in 2003 to 59.6 in 2012 (Fig. 3).

Figure 2:
Plot of predicted probability of survival by NISS value for each year. Shaded regions indicate the 95% CIs for the predicted values obtained from the logistic regression model summarized in Supplemental Digital Content 5, (Model 2).
Figure 3:
NISS values associated with a predicted 50% or greater probability of survival predicted by the logistic regression model summarized in Supplemental Digital Content 5, (Model 2).


This study is the first that accurately quantifies a marked improvement in survival following major trauma on UK combat operations during the last decade. These findings provide unprecedented detail on the military casualties sustained by the UK Armed Services on combat operations. The results demonstrate that a majority of injuries from hostile action were primarily from blast and fragmentation trauma resulting from explosive devices, with the extremities being more likely than any other region to be injured.

This study echoes the finding of our US colleagues that support the belief that survival has improved over the conflict.14 However, previous attempts to quantify this improvement have relied on the case-fatality rate, that is, the ratio of fatalities to total casualties (killed and wounded). While we cite this figure in Table 1, the authors regard this methodology as vulnerable to multiple confounders, although it has the advantage of comparison with historical records, which is impossible with more sophisticated techniques such as NISS.

It is important to acknowledge that although obviously very important, mortality is a crude outcome measure. The UK JTTR is investigating the possibility of measuring subsequent disability, functional recovery, or patient-reported quality of life, but this has inherent challenges, and no major trauma registry has successfully incorporated this. There is limited evidence that casualties with severe injuries have returned to military service, and this can be regarded as a surrogate marker of functional recovery.15

The UK Military trauma system is not analogous to a civilian trauma network. Aside from the initial invasion of Iraq and special operations, the UK DMS did not deploy forward surgical teams and instead relied on rapid transit from point of wounding to a single field hospital in each operational theater. The distinction between prehospital and hospital care has become blurred with “hospital techniques,” that is, resuscitation with blood (from 2008), intubation, and thoracostomies being taken into the prehospital environment, with the overwhelming majority of UK casualty retrieval helicopters carrying a doctor capable of delivering these techniques with demonstrable improvements is survival.16

There is some distinction between the UK and US strategies in this regard. By virtue of the larger area of operations of US forces and resultant longer transit times, the US Military trauma system deploys forward surgical teams that perform initial damage-control procedures before transfer on to a field hospital.17,18 A further distinction is that US helicopters transferring casualties from point of wounding to a forward surgical team and onward to the field hospital are staffed by emergency medical technicians and paramedics rather than physicians.16 The UK also returns patients direct to a single treatment facility (the Royal Centre for Defence Medicine, Birmingham, United Kingdom), whereas the US returns patients to three main US-based hospitals after transit via the Landstuhl Regional Medical Center (Rhineland-Palatinate, Germany).

The structured MACE approach to improving combat casualty care is similar between the UK and US military medical services. Both operate a JTTR that are sufficiently aligned to allow coordinated joint research.16,19–21

The finding of this study that 70% of injuries resulting from hostile action are from explosive weapons is entirely consistent with the experience from the previous half century of conflict. In modern warfare, the majority of injuries have similarly resulted from explosive weapons, that is, landmines, rockets, grenades, improvised explosive devices, and mortars as demonstrated in Table 3. Our findings that the extremities are the most likely body region to be injured are consistent with the published literature from recent conflicts. Interestingly however, extremity injuries form a relatively smaller proportion of all injuries in this study (Table 3) compared with previous studies.

Relative Proportions of GSW to Blast Injuries, and Proportion of Wounds Involving the Extremities

The management of polytrauma is complex, and when casualties are injured on an overseas battlefield, this complexity is increased significantly. In an already complex system of care, pinpointing specific factors responsible for improved odds of survival is challenging. From the presented results, it is not possible to conclude which specific intervention and system changes improved outcomes. We propose that the improvement in survival demonstrated in this study is likely caused by the aggregation of multiple summative improvements in techniques across a system that adopts an “end-to-end” approach, blurring the boundaries between point of wounding treatment, prehospital en route care, receiving field hospital management, and in-flight care during repatriation to continuing care in the National Health Service.

If such results are to be achieved in the already fast improving civilian trauma sector, no single or limited number of advances are necessarily required or indeed expected in the near future; rather, smaller cumulative improvements across the care pathway could yield significant benefit in trauma outcomes.

Management of trauma in deployed UK Military medical facilities is both consultant led and consultant delivered. With the high tempo and unpredictability of military operations during the last decade, consultants have gained experience across multiple previous deployments. This knowledge is further consolidated by the cyclical predeployment training system through which clinicians returning from deployment instruct their colleagues about to deploy via the bespoke Military Operational Surgical Training (MOST) course run with the assistance of the Royal College of Surgeons of England since 2007. This team-based training involves rehearsing damage-control resuscitation and surgical techniques on cadaveric material and third-generation simulation mannequins with the complete team of surgeons, anesthetists, emergency physicians, and theater staff using current equipment and protocols. It is possible that improvements in UK Military trauma system performance achieved during the last 10 years might be lost at the cessation of hostilities. The associated decreased exposure to severe combat trauma may result in a loss of some of the gains in survival demonstrated by this study in the initial phases of subsequent conflicts.

There has been a significant improvement in the understanding of resuscitation with blood products during the last 10 years. The traditional concept of restoring a casualty’s hemoglobin concentration by administering packed red blood cells has been augmented by the administration of packed red blood cells and fresh frozen plasma at approximately a 1:1 ratio with early platelet administration.27 This strategy was improved in the second half of the study period by massive transfusions being guided by real-time, near-patient thrombelastography (introduced from 2009), which has supplanted traditional measures of “clotting,” allowing a tailored correction of coagulopathy as part of the resuscitation phase.28 Large volumes of blood products that are often required in catastrophic exsanguinating hemorrhage are infused through high-volume blood warmers at rates of up to 1 L/min. In addition, the use of tranexamic acid has been shown to improve survival following major trauma in both civilian and military studies and is now routinely administered at the prehospital setting within the military trauma system.19,29 These improvements in care were directly resulting from the systematic MACE system of analyzing practice and outcomes and rapidly altering doctrine and training accordingly as well as the ability to increase the numbers available for analyses by pooling UK and US JTTR data.16,19

Two or more anesthetists manage the casualty’s airway, ventilation, anesthetic, and resuscitation. An operating department practitioner and a transfusion technician aid them. Adapting these structures to the civilian setting will require recognition that a team of doctors and technicians, rather than a single doctor, is required for the anesthetic/resuscitation of the severely injured patient in the way described.30

Helicopters have been routinely used by the military to transport casualties from near point of wounding to surgical facilities since the US-Vietnam conflict.31 During the conflicts of the last decade, the UK Military developed a consultant-led retrieval service that permits prehospital advanced interventions. This service has been shown to improve outcomes when compared with non–physician-led conventional helicopter casualty retrieval in a selected group of patients with Injury Severity Scores (ISSs) between 16 and 50.16

Personal protective equipment improved during the study period and now offers protection to the head, eyes, and torso as standard with personnel able to increase protection of the neck, shoulders, thorax, groin, and thighs, depending on perceived threat.

We would like to acknowledge the inherent weaknesses of this study. First, although our results quantify improved survival, an observational study of this type is clearly unable to identify specific treatment or intervention responsible for this effect. Second, data modeling is inherently imperfect. The high shrinkage estimate of Model 2 implies that it does not demonstrate significant overfitting; therefore, we are confident that the model estimates are reliable.

Third, we acknowledge that NISS is an anatomic measure only and might underestimate survival in a young and physically robust military population. Both the UK and US JTTR collect Trauma and Injury Severity Scores (TRISS), which also incorporates physiologic variables; however, these calculations are both based on coefficients for either blunt or penetrating injury developed in the 1980s. There is currently no coefficient for explosive injury, and given that this is the most common injury mechanism in the registry, we regard this as a significant limitation with the use of TRISS in this population.

The output from Model 2 in which NISS is incorporated as a continuous variable is summarized in Supplemental Digital Content 5 ( For this model, the likelihood ratio test comparing the specified model against the null model returned p values of less than 0.0001. The le Cessie–Van Houwelingen–Copas–Hosmer goodness-of-fit test statistics for Model 2 gives a p value of 0.314. In addition, this model exhibits high discriminatory power with a C statistic of 0.982 and a shrinkage estimate of 0.9944.

This study shows a dramatic improvement in survival over the 10 years, caring for UK casualties across the conflicts in Iraq and Afghanistan. These results suggest that future military trauma system performance metrics will need reconsideration to be sensitive to change from the current performance level. More sophisticated outcome measures the performance of combat casualty care systems including morbidity and functional recovery are required to drive future improvement.


J.P-B. conceived of this study; J.P-B, S.A.G.R. and M.J.M. contributed to the study design. J.R.B.B. contributed to data analyses and graph preparation. J.P-B., S.A.G.R., and M.J.M. prepared the manuscript.


We acknowledge the hard work, dedication, and professionalism of all the members of the Defence Medical Services and the National Health Service that cared for the casualties described in this work. We also thank the Clinical Information and Exploitation Team, Joint Medical Command, and UK Defence Statistics (Health) for collecting, collating, and identifying appropriate data for this article. We also thank Brig. Tim Hodgetts L/RAMC for his role in leading the development of the Joint Theatre Trauma Register and Major Trauma Audit for Clinical Effectiveness (MACE) process.


J.P.B., S.A.R. and M.J.M. are serving officers in the Royal Navy. All authors have completed the ICMJE uniform disclosure form and declare no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; and no other relationships or activities that could seem to have influenced the submitted work.


1. Hickey M. The Korean War: The West Confronts Communism, 1950–1953. 1st ed. Woodstock, NY: John Murray; 1999: 408.
2. Smith J, Hodgetts T, Mahoney P, Russell R, Davies S, McLeod J. Trauma governance in the UK Defence Medical Services. J R Army Med Corps. 2007; 153 (4): 239–242.
3. Association for the Advancement of Automotive Medicine: The Abbreviated Injury Scale 1990 Revision - Update 98 Barrington, IL: Association for the Advancement of Automotive Medicine; 1998.
4. Baker SP, O’Neill B, Haddon W Jr, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma. 1974; 14 (3): 187–196.
5. Association for the Advancement of Automotive Medicine, ed. The Abbreviated Injury Scale 2005—Military. Des Plaines, IL: Association for the Advancement of Automotive Medicine; 2005.
6. Osler T, Baker SP, Long W. A modification of the injury severity score that both improves accuracy and simplifies scoring. J Trauma. 1997; 43 (6): 922–925.
7. Ingram A. Parliamentary Written Answers London: Hansard; 2006. Available at: Accessed October 5, 2013.
8. Ministry of Defence. Operations in Afghanistan [Memorandum]. London, England: Ministry of Defence; 2010. Available at: Accessed October 5, 2013.
9. Akaike H. A new look at the statistical model identification. IEEE. Transactions on Automatic Control. 1974; 19 (6): 716–723.
10. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2012.
11. Harrell FEJ. rms: Regression Modeling Strategies. R package 3.6-3 ed. Nashville, TN; 2013.
12. Fox J. Effect Displays in R for Generalised Linear Models. J Stat Software. 2003.
13. Venables WN, Ripley BD, Venables WN. Modern Applied Statistics With S. 4th ed. New York, NY: Springer; 2002: xi, 495.
14. Rasmussen TE, Gross KR, Baer DG. Where do we go from here? Preface. US Military Health System Research Symposium, August 2013. J Trauma Acute Care Surg. 2013; 75 (2 Suppl 2): S105–S106.
15. Penn-Barwell JG, Fries CA, Bennett PM, Midwinter M, Baker A. Mortality, survival and residual injury burden of Royal Navy and Royal Marine combat casualties sustained in 11-years of operations in Iraq and Afghanistan. J R Nav Med Serv. 2014; 100 (2): 161–165.
16. Morrison JJ, Oh J, DuBose JJ, O’Reilly DJ, Russell RJ, Blackbourne LH, Midwinter MJ, Rasmussen TE. En-route care capability from point of injury impacts mortality after severe wartime injury. Ann Surg. 2013; 257 (2): 330–334.
17. Chambers LW, Rhee P, Baker BC, Perciballi J, Cubano M, Compeggie M, Nace M, Bohman HR. Initial experience of US Marine Corps forward resuscitative surgical system during Operation Iraqi Freedom. Arch Surg. 2005; 140 (1): 26–32.
18. Beekley AC, Watts DM. Combat trauma experience with the United States Army 102nd Forward Surgical Team in Afghanistan. Am J Surg. 2004; 187 (5): 652–654.
19. Morrison JJ, Dubose JJ, Rasmussen TE, Midwinter MJ. Military Application of Tranexamic Acid in Trauma Emergency Resuscitation (MATTERs) Study. Arch Surg. 2012; 147 (2): 113–119.
20. Brown KV, Dharm-Datta S, Potter BK, Etherington J, Mistlin A, Hsu JR, Clasper JC. Comparison of development of heterotopic ossification in injured US and UK Armed Services personnel with combat-related amputations: preliminary findings and hypotheses regarding causality. J Trauma. 2010; 69 (Suppl 1): S116–S122.
21. Morrison JJ, Ross JD, Dubose JJ, Jansen JO, Midwinter MJ, Rasmussen TE. Association of cryoprecipitate and tranexamic acid with improved survival following wartime injury: findings from the MATTERs II Study. JAMA Surg. 2013; 148 (3): 218–225.
22. Owens BD, Kragh JF Jr, Wenke JC, Macaitis J, Wade CE, Holcomb JB. Combat wounds in operation Iraqi Freedom and operation Enduring Freedom. J Trauma. 2008; 64 (2): 295–299.
23. Owens BD, Kragh JF Jr, Macaitis J, Svoboda SJ, Wenke JC. Characterization of extremity wounds in Operation Iraqi Freedom and Operation Enduring Freedom. J Orthop Trauma. 2007; 21 (4): 254–257.
24. Jackson DS. Sepsis in soft tissue limbs wounds in soldiers injured during the Falklands Campaign 1982. J R Army Med Corps. 1984; 130 (2): 97–99.
25. Hardaway RM 3rd. Viet Nam wound analysis. J Trauma. 1978; 18 (9): 635–643.
    26. Reister FA. Battle Casualties and Medical Statistics; US Army Experience in the Korean War. Washington, DC: Office of the Surgeon General, US Army; 1973: xii, 172.
      27. Jansen JO, Morrison JJ, Midwinter MJ, Doughty H. Changes in blood transfusion practices in the UK role 3 medical treatment facility in Afghanistan, 2008–2011. Transfus Med. 2014.
      28. Allcock EC, Woolley T, Doughty H, Midwinter M, Mahoney PF, Mackenzie I. The clinical outcome of UK military personnel who received a massive transfusion in Afghanistan during 2009. J R Army Med Corps. 2011; 157 (4): 365–369.
      29. CRASH-2 trial collaborators, Shakur H, Roberts I, Bautista R, Caballero J, Coats T, Dewan Y, El-Sayed H, Gogichaishvili T, Gupta S, Herrera J, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet. 2010; 376 (9734): 23–32.
      30. Morrison JJ, Ross JD, Poon H, Midwinter MJ, Jansen JO. Intra-operative correction of acidosis, coagulopathy and hypothermia in combat casualties with severe haemorrhagic shock. Anaesthesia. 2013; 68 (8): 846–850.
      31. Neel S. Army aeromedical evacuation procedures in Vietnam: implications for rural America. JAMA. 1968; 204 (4): 309–313.

      Combat injuries; survival; war; injury severity

      Supplemental Digital Content

      © 2015 Lippincott Williams & Wilkins, Inc.