Authors alone are responsible for opinions expressed in the contribution and for its clearance through their federal health agency, if required.

ORIGINAL ARTICLES Authors alone are responsible for opinions expressed in the contribution and for its clearance through their federal health agency, if required. MILITARY MEDICINE, 180, 7:780, 2015 Evaluating MEDEVAC Force Structure Requirements Using an Updated Army Scenario, Total Army Analysis Admission Data, Monte Carlo Simulation, and Theater Structure LTC Lawrence Fulton, MS USA (Ret.)*; Col Bernie Kerr, USAF MSC (Ret.) ; James M. Inglis, GS-13 ; LTC Matthew Brooks, PhD, MS USA (Ret.) ; CPT Nathaniel D. Bastian, MS USA ABSTRACT In this study, we re-evaluate air ambulance requirements (rules of allocation) and planning considerations based on an Army-approved, Theater Army Analysis scenario. A previous study using workload only estimated a requirement of 0.4 to 0.6 aircraft per admission, a significant bolus over existence-based rules. In this updated study, we estimate requirements for Phase III (major combat operations) using a simulation grounded in previously published work and Phase IV (stability operations) based on four rules of allocation: unit existence rules, workload factors, theater structure (geography), and manual input. This study improves upon previous work by including the new air ambulance mission requirements of Department of Defense 51001.1, Roles and Functions of the Services, by expanding the analysis over two phases, and by considering unit rotation requirements known as Army Force Generation based on Department of Defense policy. The recommendations of this study are intended to inform future planning factors and already provided decision support to the Army Aviation Branch in determining force structure requirements. INTRODUCTION The Army continues to analyze and adjust its force structure given increasingly limited resources. To support this analysis, the Medical Evacuation Proponency Directorate was asked to re-evaluate air ambulance requirements for supporting both Phase III (major combat operations or MCOs) and Phase IV (stability operations). Similar analysis was conducted previously in 2009 with an Army-approved, MCO scenario. In *Rawls College of Business, Texas Tech University, 703 Flint Avenue, Lubbock, TX 79410. Department of Health Administration, 1200 South Franklin Street, Mt Pleasant, MI 48859. Medical Evacuation Proponency Directorate, Army Medical Department Center and School, 1608 Stanley Road, Suite 47, Fort Sam Houston, TX 78234. Department of Health Administration, Room 250A 601 University Drive, Texas State University, San Marcos, TX 78666. Center for Integrated Healthcare Delivery Systems, Department of Industrial and Manufacturing Engineering, Pennsylvania State University, 355 Leonhard Building, University Park, PA 16802. The views expressed in this article are those of the authors and do not reflect the official policy or position of the U.S. Government, the Department of Defense, or any other official agency. doi: 10.7205/MILMED-D-14-00580 this previous study, Monte Carlo simulation using empirical patient and aircraft probability distributions resulted in requirement estimates for medical evacuation aircraft. These estimates suggested that between 0.4 and 0.6 airframes per admission would be appropriate to handle the workload of patients for the specific scenario. 1 The study presented here extends the 2009 analysis longitudinally, updates the analysis with a current scenario, and applies better methods. The results informed decision-makers regarding planning factors and potential rules of allocation that drive unit requirements in the Total Army Analysis (TAA), which is the Army s method of assessing force structure requirements. A discussion of the TAA process is provided later, as we first define some terms that are important to this study. The definitions of rules of allocation and planning factors need to be explicated to understand this study. In the Army, rules of allocation are associated with requirements estimation based on the TAA process, a process that will be described shortly. Planning factors, on the other hand, are used to estimate requirements necessary to support a specific operation. They are not used for force structure development per se. Understanding this distinction is important, because this study deals solely with rules of allocation (ROAs) rather 780

than planning factors, even though the ROAs may have implications for planners. ROAs are used for planning the future force, whereas planning factors estimate requirements for a specific operation. Since ROAs are used to estimate unit requirements as part of the TAA process, it is essential to understand the definition of these rules as well. The four ROAs used in the TAA process are (1) existence-based, (2) workload-based, (3) theater structure, and (4) manual input. Existence-based units link one unit s requirements to another (e.g., one unit X per each division). Workload-based rules link units to what work must be performed (e.g., one unit X per 1,500 Short Tons to be moved). Theater-structure rules provide requirements based on the theater geography and organizational structure (e.g., one unit X per corps in Theater Y). Manual input is used when standard allocation rules do not apply. 2 In this study, we look at the current, existence-based ROAs (those that link one unit s requirements to the existence of another) that affect air ambulance requirements and proffer more realistic methods. Specifically, we suggest that all four types of rules associated with the TAA process be used to estimate medical evacuation requirements. Although authorizations might not meet the requirements because of funding constraints, documentation of the appropriate structure given strategic scenarios is important. Phase III and Phase V operations are discussed in this study as well as Enhanced Protective Posture (EPP), Army Support to other Services (ASOS), and Army Force Generation (ARFORGEN). Phase III operations are those that are deemed to be MCOs in a particular scenario. Phase IV operations are stability operations, post MCOs. EPP reflects the homeland defense mission. ASOS reflects a mission where the Army provides support to other service components (e.g., Navy and Air Force.) Finally, ARFOGEN reflects the rotational policy for Active Component and Reserve Component forces. In other words, the Army has goals for how many months a unit should remain in garrison versus deployed. ARFORGEN cycles reflect these goals. With these basic definitions provided, we present the results of our study in this order. First, we review the TAA process for the Army as well as the rules associated with that process. Second, we evaluate the existence-based rules from a modified TAA scenario recently evaluated by the Aviation Branch. We modify this scenario by reordering monthly demand signals and removing time components so that the estimates of requirements will be identical even though the scenario is fictitious. The original scenario components as well as the simulated scenario include several basic missions that often need to be performed simultaneously. SIGNIFICANCE This study is the first to evaluate longitudinally the air ambulance requirements necessary for supporting both Phase III (MCOs) and Phase IV (stability operations). It is the also the first of its type that proposes separate planning factors which might be used to adjust ROAs. We further suggest that aeromedical evacuation ROAs should be evaluated separately outside of U.S. Army Aviation Branch requirements, which is now the status quo. THE TAA PROCESS The TAA process provides a method for the Army to estimate unit requirements and then links these requirements to authorizations based on affordability and accepted risk. The outcome of the TAA is the Army s required structure and its budgeted future force (Program Objective Memorandum or POM), so careful input into the process is vital. 3 The TAA process uses strategic planning guidance and joint force capability requirements based on documents such as the National Military Strategy and Defense Planning Guidance to evaluate full-spectrum operation requirements and then resource these requirements given budgetary constraints. It is an annual process conducted in two phases, the requirements determination phase and the resourcing/ determination phase. The requirements determination phase generates so-called above the line combat forces necessary based on the scenarios, timelines, simulations, and analysis of strategic planning documents. In this phase, quantitative analysis is conducted by the Center for Army Analysis to provide estimates of requirements for below the line forces, those forces necessary to support the above the line combat forces. In the qualitative analysis phase, the initial budgeted force structure is created. This force structure eventually becomes the recommendation to the Secretary of Defense of how to allocate resources. The end state of TAA is the projected future Army force structure. 4 The requirements determination phase, that phase which requires analytical rigor, is the focus of this study. The pivotal components of the quantitative analysis are the ROAs because these rules drive the requirements. In this study, we propose a compilation of these rules based on the phase of operation and nature of the scenario. PREVIOUS WORK AND PROBLEM STATEMENT Some previous literature exists regarding estimations associated with the Army s TAA process. Fulton et al 1 provided the first study estimating the medical evacuation workload requirements for Phase III operations given a TAA scenario; however, this study did not evaluate Phase IV operations, rotational requirements, or other missions such as ASOS. The Army is designated the aeromedical evacuation provider for all services in theater, so this mission (an ASOS mission) is a known Department of Defense (DoD) requirement. 5 Unfortunately, there is currently no ASOS ROAs for Army aeromedical evacuation despite the known and stated requirement. Another study used the TAA patient admission stream associated with a set of stacked scenarios (scenarios that occurred in overlapping time intervals) to estimate the effects of deployment on the medical sustaining base. This study is relevant in that it implies analysis should include effects 781

not typically associated with deployable units. 2 In the case of medical evacuation units, the nondeployable structure supports some fixed based aeromedical evacuation coverage. Several other studies have used current operations for modeling the future force structure, 6 14 but none of these studies used DoD-approved scenarios. No study to date has evaluated separate ROAs for aeromedical evacuation units that realistically portray the demand based on the workload, the theater, and an existence-based requirement. With these studies in mind, we determined that a significant gap in the literature existed, and began our analysis of requirements. HISTORICAL AND CURRENT ROAs Before 2005, the air ambulance companies were subordinate to the medical evacuation battalion. The ROAs were also the planning factors associated with air ambulance companies. These were specified in FM 8-55, Planning for Health Service Support. 15 The rules were one air ambulance company in direct support of each division, one for every three separate brigades or armored cavalry regiment, one for every two divisions for general support, and one per theater for ship-to-shore and shore-to-ship missions. Theater requirements could either increase or eliminate the requirements for the ship-to-shore mission set, however, depending on the TAA scenario set. The first three rules were based on existence, whereas the final rule was based on theater structure given the inherently joint role of air ambulance units. When the air ambulance companies were assigned to the General Support Aviation Battalion (GSAB) of the Combat Aviation Brigade (CAB) in 2005, air ambulance companies were reduced from 15 aircraft to 12 aircraft, and the maintenance support was removed from the unit. The cost savings reaped by the Army resulted in a 48 aircraft reduction for one corps, four divisions, and one ACR mission given the ROAs listed previously. Prolonged operations in Iraq and Afghanistan resulted in additional analysis, and air ambulance units were restored to a 15 aircraft per unit requirement just to meet operational demands. 16 But the ROA exists only for the air ambulance company s parent battalion, the GSAB. Air ambulance approved unit requirements are one per GSAB in the CAB, and two per GSAB in the Theater Aviation Brigade, although some changes in these rules are under consideration. Reductions in Theater Aviation Brigade would therefore have a disproportionate impact on air ambulance companies despite true requirements, underscoring that linking aeromedical requirements to Aviation Branch structure has no basis in real mission requirements. RESEARCH QUESTION The research question for this study derives from a need to provide affordable sourcing solutions that represent reality. The study team did not focus on authorizations (funded requirements), but instead focused on the requirements phase. The research question was straightforward. Given various levels of risk and a DoD-approved scenario set, what are the appropriate air ambulance requirements for Phase III and Phase IV operations? The answers to this question would include a complete analysis of ASOS, ARFORGEN, which accounts for unit rotation and stability requirements, steady FIGURE 1. This stacked column chart represents the baseline evaluation of aeromedical evacuation companies for the set of scenarios. The x-axis reflects a monthly time component (12 years of scenario data). 782

state missions, ship-to-shore/shore-to-ship evacuation requirements, homeland security requirements, as well as Phase III and Phase IV requirements. CASE STUDY The specific case study in question is based on a representation of a 12-year set of scenarios. The set of scenarios itself is re-ordered for this article; however, for purposes of decision support, the actual DoD-approved scenario was used. To evaluate the requirements, we first looked at the steady state mission requirements based upon the current Aviation Branch existence rule of allocation for the GSABs, which drives aeromedical evacuation force structure. The number of companies required by month based solely on Aviation Branch rules for steady state operations around the world, deterrence missions to prevent war, EPP to protect the homeland, and Phase III combat operations is shown in Figure 1. You can see that the maximum number of companies required by using the Aviation Branch existence rules is 33. The 95th percentile requirement is 25 companies. But the existence-based analysis does not provide an appropriate analysis for air ambulance workload requirements. During Phase III operations, casualties drive workload as demonstrated by previous work. 1 To estimate workload-based demand, we used the patient admission stream provided by the Center for Army Analysis and Monte Carlo simulation methods. Specifically, we used the admission stream to estimate patient evacuation requirements based primarily on empirical distributions from air ambulance companies deployed to Iraq and Afghanistan. We then used distributions to estimate missions per aircraft, casualties evacuated per airframe, and ultimately calculated the airframes and air ambulance companies required for these distributions. Distributions derived from previously published work 1 with some distributional changes. All analyses were done in R Statistical Software. 17 Figure 2 depicts the basic simulation flowchart. In the simulation, we evaluated daily rather than monthly medical evacuation airframe requirements during Phase III operations and stability operations. This approach provides a more realistic assessment of requirements. From our analysis, we estimated a requirement for 15 additional companies to handle the patient evacuation workload. Although this does not cover the entirety of all requirements, it provides coverage for about 95% of the mean workload for MCO patients during Phase III. One should note that this company requirement is based solely on Army admissions and does not account for any other mission set. We overlaid the known medical evacuation workload requirements generated from the simulation as well as other mission sets, and subtracted out the existence-based companies FIGURE 2. This simple flow chart depicts the analysis performed to estimate workload requirements. Poi indicates a Poisson distribution. U indicates a uniform distribution with minimum and maximum parameters. 783

FIGURE 3. This figure depicts the medical evacuation unit requirements to handle workload-based, existence-based, and theater-based requirements. The +MCO Wkld, 95% indicates the 95th percentile workload requirement on top of assets already available because of existence rules. This bolus derives from the simulation of workload requirements. already in theater based on the Aviation Branch ROAs. Mission sets not included in the current ROAs that we included in our analysis were the ship-to-shore/shore-to-ship mission during scenario-driven times, the ASOS mission, and a European Union Theater Commitment requirement. These results (which omit ARFORGEN requirements) are depicted in Figure 3. The maximum requirement is 48 aeromedical evacuation companies, and the 95th percentile requirement is 29. The minimum (steady state) requirement is 16.3 companies, not including ARFORGEN requirements. Next, we evaluated ARFORGEN requirements. ARFORGEN allows units to have a rotational replacement from within the force structure. For example, an active component unit serving 1 year in a theater might expect to be replaced and have 1 year back on station to recover, reset, retrain, and redeploy. This is called a 1:1 ARFORGEN cycle. Reserve Component units typically serve approximately a year in theater every 3 years. We model that as a 1:3 cycle (1 year in theater and 3 years in garrison). We recognize that the intensity of Phase III (major combat) operations precludes unit replacement considerations. (The military will be less concerned with replacement operations and time on station when major combat is ongoing.) So the focus of our analysis was on nonpeak (Phase I, II, IV, and V) operations. We evaluated a 1:1 ARFORGEN cycle for Active Component (AC) units (in other words, 1 year in theater followed by 1 year back in garrison) and a 1:3 ARFORGEN cycle for Reserve Component (RC) units (in other words, 1 year in theater followed by 3 years back in garrison). Figure 4 shows the results of applying ARFORGEN to a 50% AC to 50% RC mix (1 AC company for each 1 RC company), to a 60% AC to 40% RC mix (6 AC companies for 4 RC companies), and to a 70% AC to 30% RC mix (7AC companies for 3 RC companies). Because the AC units can cycle back into theater more quickly, the unit requirements are reduced when the percent AC forces increases in theater. (NOTE: ARFORGEN requirements do not apply to several of the mission sets, including homeland defense (EPP), as these missions are not rotational.) The maximum company requirements for the ARFORGEN analysis based on a 60/40 AC/RC split is 120, clearly infeasible given the number of units available in the inventory. The 95th percentile for this case is 82 units, still infeasible. Ignoring the peak operations (associated with Phase III), the steady state requirement is 58 companies to handle the rotational requirements. ARFORGEN for aeromedical evacuation units given this set of scenarios is infeasible. Aeromedical evacuation units should not expect ARFORGEN policies to apply to them if scenario sets similar to this occur. DISCUSSION AND CONCLUSIONS From this analysis, several critical findings emerge First, evaluating only existence-based rules for medical evacuation companies is grossly flawed and under-represents the 784

FIGURE 4. The requirements analysis for medical evacuation units becomes more complex when considering ARFORGEN. Clearly, ARFORGEN in peak operations is impossible regardless of the AC/RC mix. More interestingly, it is impossible in any phase of the CAB mission set provided. We simply have insufficient units to pay for the SS-3 unit requirements. magnitude of the requirements. Second, workload rules should be used on top of existence-based rules to evaluate medical evacuation requirements. The magnitude of the workload requirements portrayed in Figure 3 for Phase III operations exceeds all existence-based requirements. Third, manual rules are necessary to account for existing and known medical evacuation requirements. For example, the EU-committed structure must be included in any analysis, and the companies required for other mission sets (e.g., EPP) have to be evaluated. Fourth, theater structure cannot be ignored for medical evacuation assets. In this case, the ship-to-shore mission should be identified for all potential off-shore aeromedical evacuation operations in the theater scenario. The Aviation Branch has developed some new ROAs for the TAA process that are pending approval. These new rules may authorize 5 to 10 additional companies during Phase III operations, well short of the 95th percentile requirement we detailed in this analysis. These rules still couple aeromedical evacuation requirements to aviation requirements. Our analysis suggests that this is an inappropriate way to estimate requirements. Given our analysis, we believe that is necessary to decouple the medical evacuation ROAs from the current Aviation Branch existence-based rules or to augment these existence rules with workload, theater, and manual input rules. If we do not accomplish this, then requirements will be underreported. Even if authorizations are unavailable for the unit structure required, at a minimum, the requirements can be documented and contingency plans made to ameliorate the potential delta should this scenario or one like it occur. Finally, it is important to mention limitations as well as future work. One of the major limitations in this analysis is that no consideration was made for the Geneva Convention requirements for evacuating non-u.s. injured. Second, we used data associated with admissions and converted them to evacuation requirements with very general distributions derived from Iraq and Afghanistan operations in the workload simulation. These distributions are likely to be different in future conflicts, which is why the parameters are flexible. Third, it is possible that ARFORGEN is just too expensive to generate requirements. In this case, we should be straightforward about the prospects of extended deployments for the medical evacuation community. Our future work will apply the same type of methods discussed here (as well as others) to other TAA scenarios in support of the Aviation and Medical decision-making processes. ACKNOWLEDGMENTS The researchers wish to acknowledge the Center for Army Analysis and the Center for AMEDD Strategic Studies for making data available for this study. 785

REFERENCES 1. Fulton L, McMurry P, Kerr B: A Monte Carlo Simulation of air ambulance requirements during major combat operations. Mil Med 2009; 174(6): 610 4. 2. Department of the Army: Total Army Analysis, Army Regulation 71 11. Washington, DC, U.S. Department of the Army, December 29, 1995. Available at http://www.army.mil/usapa/epubs/pdf/r71_11.pdf; accessed October 1, 2014. 3. Fulton L, Devore R, McMurry P: Estimating sustaining base hospital personnel requirements during extended operations. Mil Med 2010; 175(4): 238 46. 4. United States Army War College (USAWC): How the Army Runs: A Senior Leader Reference Handbook, Ed 26, pp 60 2. Carlisle, PA, USAWC Press, 2011. 5. Department of Defense: Directive 5100.01: Functions of the Department of Defense and its Major Components, p 30, 2010. Available at http://dtic.mil/whs/directives/corres/pdf/510001p.pdf; accessed October 6, 2014. 6. Fulton L, Lasdon L, McDaniel R, Nicholas CM: Two-stage stochastic optimization for the allocation of medical assets in stability operations. JDMS 2010; 7(2): 89 102. 7. Bastian N, Fulton L, Mitchell R, Pollard W, Wilson R: Force design analysis of the army aeromedical evacuation company: a quantitative approach. JDMS 2012; 10(1): 23 30. 8. Bastian N, Brown D, Fulton L, et al: Analyzing the future of army aeromedical evacuation units and equipment: a mixed methods, requirementsbased approach. Mil Med 2013; 178(3): 321 9. 9. Bastian N: A robust, multi-criteria modeling approach for optimizing aeromedical evacuation asset emplacement. JDMS 2010; 7(1): 5 23. 10. Fulton L, Bastian N, Wilson R: Carried away. Quality Progress 2015; 48(2): 32 8. 11. Grannan B, Bastian N, McLay L: A maximum expected covering problem for locating and dispatching two classes of military medical evacuation air assets. Optimization Letters 2014; OnlineFirst: 1 21. 12. Bastian N, Fulton L: Aeromedical evacuation planning using geospatial decision-support. Mil Med 2014; 179(2): 174 82. 13. Bastian N, Fulton L, Mitchell R, Pollard W, Wierschem D, Wilson R: The future of vertical lift: initial insights for aircraft capability and medical planning. Mil Med 2012; 177(7): 863 9. 14. Bastian N, Grannan B, Fulton L: Optimizing forward surgical team operating room scheduling for mass casualty events. Proceedings of the 14th INFORMS Computing Society Conference, Operations Research and Computing: Algorithms and Software for Analytics (Ed. Borchers et al.), 2015; pp 161 71. 15. Department of the Army: Field Manual 8 55: Planning for Health Service Support. Washington, DC, U.S. Department of the Army, September, 1994. Available at https://rdl.train.army.mil/catalog-ws/view/100.atsc/76dc22c3-a723-45b3-b6dc-c5d1ac1b82ae-1274316664931/ 8-55/toc.htm; accessed June 10, 2014. 16. Leipold J: Army to Increase Medevac Support, add new CAB, more UAVs, 2010. Available at http://www.army.mil/article/32603/; accessed January 10, 2014. 17. R Core Team: R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, 2013. ISBN 3-900051-07-0, Available at http://www.r-project.org/; accessed June 10, 2014. 786