Limitations of blood supply and walking blood bank implementation in Forward Resuscitative Surgical Detachments during large-scale combat operations: A Monte Carlo simulation model

Abstract

BACKGROUND:

Hemorrhage remains the leading cause of preventable battlefield mortality. Although early whole blood resuscitation improved outcomes during the Global War on Terror, this relied on permissive operational conditions unlikely to persist in future large-scale combat operations (LSCO). Projected LSCO casualty volumes may rapidly overwhelm current Role 2 Forward Resuscitative Surgical Detachments (FRSDs), and walking blood banks (WBB) may provide only minimal augmentation. This study evaluates blood supply resilience for a Role 2 FRSD across LSCO casualty scenarios and quantifies the incremental benefit of WBB augmentation.

METHODS:

A discrete-time, daily-step simulation model was developed in Python to model whole blood dynamics for a Role 2 FRSD over a 30-day conflict. Inventory was tracked using an age-segmented inventory. Baseline stock was 120 units with a resupply of 120 units every 3 days. A 100-donor WBB (56-day deferral) was incorporated. Daily casualties were stochastically generated across escalating intensity, with transfusion demand calculated assuming 20% of casualties required transfusion at 8 units per patient. Primary outcomes included days-to-failure (first unmet unit) and percent demand met over 30 days.

RESULTS:

Model-generated aggregate demand closely matched LSCO planning estimates (mean, 30-day demand 10,075.2 units vs. expected 10,080). Conventional supply alone failed rapidly once average casualties exceeded 30/day [median, 2 days; interquartile range (IQR), 1–5]. WBB augmentation extended survivability from 30 casualties/day to 11 days (IQR, 7–22) but declined to 5 days (IQR, 2–7) at 40 casualties/day and to 2 days (IQR, 2–4) at 50 casualties/day, with near-immediate failure at higher rates. Demand fulfillment declined sharply as casualty rates increased and was only modestly improved by WBB support.

CONCLUSIONS:

Current FRSD blood supply strategies are not sustainable under LSCO casualty conditions. WBBs provide limited surge buffering but cannot replace scalable resupply and forward sustainment in contested environments. (J Trauma Acute Care Surg. 2026;101: S17–S25. Copyright © 2026 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Association for the Surgery of Trauma.)

LEVEL OF EVIDENCE:

Therapeutic/Care Management; Level V.

Hemorrhage remains the leading cause of preventable mortality on the battlefield.^1,2 Advances in combat casualty care have underscored the importance of early blood resuscitation in trauma management, even exceeding the benefits of rapid evacuation.^3,4 Retrospective analyses from the Global War on Terror (GWOT) demonstrated that whole blood (WB) transfusion significantly improved survival among casualties in hemorrhagic shock.^5–7

However, the successes of casualty care during GWOT were enabled by a permissive operational environment characterized by air superiority, rapid medical evacuation, timely resupply, and a low casualty rate environment.^8,9 These conditions are unlikely to exist in a future peer-to-peer conflict. In such settings, Role 2 Forward Resuscitative Surgical Detachments (FRSDs) will likely need to provide prolonged casualty care beyond their current doctrinal capabilities. FRSDs are small, mobile medical units designed for damage control resuscitation and surgery close to the point of injury, but with the intention of rapidly evacuating their post-surgical patients to the next echelon of care.¹⁰ They are usually composed of ∼20 personnel, including surgeons, emergency medicine physicians, nurse anesthetists, critical care nurses, and surgical technicians. These medical units can receive up to 30 patients within 72 hours, but nursing care and logistics remain the key limiting factors in prolonged care.¹¹ Current doctrine defines FRSD blood storage capabilities anywhere from 40 to 200 units and requires frequent resupply.^12,13 These small teams could potentially be rapidly overwhelmed by mass casualty events, and blood supplies could be quickly depleted, highlighting the mismatch between the current medical force structure and projected peer-to-peer conflict casualty volumes. Modern military medicine, therefore, faces a pivotal transition from counterinsurgency to large-scale combat operations (LSCO), where prior assumptions of unrestricted medical support will no longer hold.^8,14

Daily battle casualties in LSCO are projected to rise exponentially. For example, recent US war game simulations predict as many as 50,000 casualties, up to ∼3,000 per day, during a major engagement with a peer adversary.¹⁵ Similarly, the ongoing Russo-Ukrainian war has already produced over 300,000 casualties, with ∼500 per day.¹⁶ By contrast, during the peak of GWOT, such as Operation Enduring Freedom and Operation Iraqi Freedom, average casualties were much lower at ∼5 and 13 per day, respectively, and only rarely exceeded sustained rates of 50 per day during major offensives.¹⁷

Another critical vulnerability lies in current resuscitation contingency plans, including the walking blood bank (WBB), which many FRSD rely upon for emergency transfusion.¹⁸ WBB programs mobilize healthy on-site service members as immediate fresh WB donors. However, donor availability and regeneration are finite, as each donor can give only one unit of blood (2 units in extremis) and should not donate more frequently than every 8 weeks.^19,20 Even when a large donor pool is available, LSCO conditions exacerbate the inherent fragility of WBB logistics.²¹ Frontline surgeons in the Russo-Ukrainian war have reported significant challenges related to unreliable donor availability, attrition, and inability to reliably store blood due to power loss and security threats.¹⁶ As a result, the WBB model, while invaluable during brief casualty surges, buys hours, not days of capacity in high-intensity conflict. Traditional military blood utilization planning, outlined by the Office of the Joint Staff Surgeon, assumes that ∼20% of casualties will require transfusion and estimates that 8 units of blood will be required per patient.²² However, this estimate fails to account for the highly stochastic, distributed, and attrition-heavy dynamics of LSCO. Deterministic planning factors and linear supply models are inadequate to predict overall utilization, depletion rates, and systemic vulnerabilities. A new, adaptable paradigm is required to ensure blood product support under conditions of variable casualty surges and contested logistics.

To address this gap, this study uses a discrete-event Monte Carlo simulation to evaluate blood supply resilience across LSCO scenarios. By modeling dynamic casualty generation and WBB integration, this analysis examines system failure points and identifies critical thresholds for collapse. The primary objective of this study was to quantify days to failure and percent transfusion demand met for a Role 2 FRSD blood system across LSCO casualty rate scenarios, and to evaluate the incremental effect of WBB augmentation under variable resupply conditions.

METHODS

Historical casualty data from primary US military conflicts were used to define casualty rate ranges. These scenarios reflect escalating intensity levels, from sustained combat operations to major high-intensity offensive engagements, consistent with military doctrine for peer-adversary conflicts. This work did not meet the criteria for human subject research and was therefore exempt from institutional review board oversight. The study was conducted and reported in accordance with the TRIPOD+AI reporting guidelines for prediction model development and evaluation, and the completed checklist is provided as Supplemental Digital Content (SDC 1, http://links.lww.com/TA/F429).

Model Development

A discrete-time, daily-step simulation model was developed in Python (v3.9) using the NumPy, Pandas, and Matplotlib libraries to evaluate blood product availability at a Role 2 FRSD. The model’s architecture was designed to characterize the day-to-day evolution of on-hand inventory, expiry-related losses, donor availability, and the net balance of units supplied versus units demanded under varying operational conditions.

The inventory system was modeled using age-segmented batches. This enabled precise tracking of shelf life and a strict First-In, First-Out (FIFO) withdrawal policy, consistent with clinical inventory management. The model managed two distinct inventory types: refrigerated WB with a 35-day shelf life, and stored WBB units with a 21-day shelf life, which is comparable to current JTS CPG guidelines for stored whole blood (SWB).²³ The simulation models a 30-day conflict by repeating a logical sequence for each day, divided into three phases: upkeep, demand, and response (Fig. 1).

Figure 1.

Core model schematic.

Daily Upkeep and Resupply

The initial WB stock was set at 120 units in line with current recommendations.²⁴ The initial stock started at day zero of shelf life. In addition, there are 100 available donors to start the simulation, each capable of providing 1 unit of blood. Once a donor is used, they enter a 56-day deferral before being able to donate again in line with current guidelines.^25,26 Replenishment consists of a 120-unit resupply every 3 days, as well as the renewal of 20 donors every 14 days to simplify accounting for donor deferral periods. At the start of each simulated day, the model performs routine logistics. All extant WB and stored WBB inventory batches are aged 1 day, and any units at or beyond their expiration date are removed and recorded as spoilage. Concurrently, the system calculates its total blood supply by accounting for the initial supply, donors, and resupply.

Stochastic Demand Generation

Next, the model simulates the combat “event” of the day. A random number of casualties is generated from a randomized distribution (eg, an “average” of 20 casualties/day is drawn from a range of 0–40). This stochastic approach accounts for the unpredictable patterns of combat, allowing for days with minimal activity and days with mass casualty events, thereby randomizing daily blood demand. This casualty number is translated into a total blood requirement, using a historical rate in which 20% of casualties require transfusion, with an allocation of 8.0 units per transfused casualty per the Office of the Joint Staff Surgeon’s current planning guidelines.^24,26

Tiered Response and WBB Activation

The model then attempts to meet this demand by using available blood sources in a specific, escalating order. The system first uses its conventional on-shelf, prestocked, nonexpired WB inventory to meet as much of the demand as possible. If a shortfall remains after all stored inventory is exhausted, the emergency WBB capability is activated. This triggers a same-day collection from the available donor pool.

The model’s scope is operational; it does not simulate blood-group compatibility, cross-matching delays, clinical triage, or non–expiry-related losses. The specific parameters governing this architecture were defined to represent known operational constraints and test scenarios, as detailed in Supplementary Digital Content 2 (http://links.lww.com/TA/F430). These variables were selected to model the primary levers of logistics, demand, and policy that influence system resilience. This core architecture and its parameters were then adapted for analyses by modifying the objectives, casualty loads, and termination logic to answer different research questions.

30-Day Cumulative Shortfall Validation

The first analysis was designed to validate the expected range of total demand and total unmet demand (shortfall) over a fixed 30-day operational period. A 10,000-trajectory Monte Carlo simulation was executed, where each trajectory ran for the full 30 days, regardless of whether a shortfall occurred. This allowed for the accumulation of total unmet demand over the entire horizon.

For this analysis, daily casualties were drawn from a discrete, randomized distribution of 0 to 420 casualties. The primary outputs were the distributions of 30-day cumulative total demand and 30-day cumulative unmet demand.

System Stress Test and Failure Dynamics

The second Monte Carlo analysis was designed as a stress test to observe LSCO casualty rate scenarios. The purpose was to understand system failure points and the dynamics of WBB activation under different operational tempos.

Casualty rates were modified to account for an average of 10 to 200 casualties per day in intervals of 10. The simulation continued until the first day that all sources of blood (WB, SWB, and WBB) ran out. This design allowed for the capture of “days-to-failure” as the primary metric, with secondary outputs including the first day of WBB activation and the total duration of WBB use.

Quantifying Operational Limitations of Conventional Versus Augmentation With WBB

A discrete-event Monte Carlo simulation was definitively employed to evaluate 20 distinct casualty scenarios. These scenarios tested target average daily casualty rates ranging from 10 to 200 in 10-unit increments. Within each scenario, the actual daily casualty count was determined stochastically by drawing from a randomized distribution between zero and twice the target rate. The model was simulated 10,000 times for each of the 20 scenarios over a 30-day operational horizon, totaling 6,000,000 simulated combat days.

Two strategies were compared: a conventional WB supply only and a conventional supply augmented with a WBB. The primary outcome was the day of failure, defined as the first day a single required unit of blood could not be provided. This metric was designed to identify the operational limits of existing recommendations. At the conclusion of each daily step, the system was assessed; if demand was fully met, the simulation proceeded to the next day. Any shortfall resulted in immediate termination and the recording of that day as the failure point. Results were aggregated as median values and the interquartile range (IQR) to accurately characterize system survivability across non-normally distributed outcomes.

Quantifying Cumulative Demand Met

To evaluate the system’s performance after the initial supply collapse, the simulation model was executed for the entire 30-day analytical horizon. While initial “day of failure” metrics identified the first instance of supply exhaustion, the fixed-horizon approach permitted each 10,000-trajectory Monte Carlo simulation to run for the full 30-day operational period regardless of failure status. This methodology enabled the calculation of the “demand met %,” defined as the ratio of total units supplied to total units demanded over the 30-day conflict horizon.

RESULTS

Model Validation Against LSCO Demand Estimates

To validate the model’s aggregate demand generation, its output was compared against recent planning factors for LSCO, which anticipate a requirement of ∼10,080 units of WB per month.²⁵ The Monte Carlo analysis produced 10,000 30-day cumulative demand distributions with a mean of 10,075.2 units (Fig. 2). This near-exact alignment with established estimates validates the model’s core architecture.

Figure 2.

A Monte Carlo analysis of 10,000 trajectories produced a 30-day cumulative demand distribution with a mean of 10,075.2 units.

System Feasibility and the Limits of WBB Augmentation

The analysis of system feasibility is displayed in Figure 3. The results plot the median day until first failure, defined as the first day when a single unit of required blood could not be provided.

Figure 3.

Comparison of the average number of casualties per day versus the median days until supply failure.

The feasibility of a conventional-only supply collapses rapidly. At an average of just 30 casualties per day, the conventional supply is exhausted in a median of 2.0 days (IQR, 1.0–5.0). Augmentation with a WBB extended survivability at this casualty rate to a median of 11.0 days (IQR, 7.0–22.0) but still falls short of achieving sustainability for 30 days.

This benefit, however, is rapidly overcome by modest increases in the casualty rate. At 40 casualties per day, the WBB-augmented blood supply fails at a median of 5.0 days (IQR, 2.0–7.0). At 50 casualties per day, median system survivability with WBB-augmented supply is only 2.0 days (IQR, 2.0–4.0). An average of 60 to 90 casualties per day shows the potential for only one day of blood supply, while anything above an average of 90 casualties per day demonstrates immediate blood supply failure.

Estimating Additional Supply Required

The amount of demand met was quantified by dividing the cumulative supply met by the sum of the cumulative supply met and the unmet demand. These results demonstrate that as casualty rates increase, the ability to meet the demand for blood products falls precipitously. In addition, the utilization of a WBB is unable to significantly bolster the demand (Fig. 4).

Figure 4.

The demand met percentage over the increasing average casualties per day.

DISCUSSION

This analysis quantifies the rate at which Role 2 FRSDs exhaust blood resources across a broad range of LSCO casualty conditions using a stochastic, discrete-event Monte Carlo framework. Model validation demonstrated near-exact alignment with established LSCO planning estimates, generating a mean 30-day blood demand of 10,075.2 units, compared with the expected 10,080 units. This concordance supports the model’s ability to represent aggregate demand, providing a credible foundation for evaluating system performance under variable casualty intensities.

Under current doctrinal assumptions, the modeled system remained feasible without failure, defined as exhaustion of blood supply, or WBB augmentation at an average of 20 casualties per day. However, the transition to unsustainable performance occurred over a narrow operational range. When average casualty rates reached 30 per day, the system experienced rapid failure within a median of 2 days (IQR, 1–5), while WBB augmentation delayed failure to a median of 11 days (IQR, 7–22), demonstrating that WBB activation can meaningfully extend short-term capacity.

Even with WBB augmentation, however, the blood supply remained highly vulnerable to modest increases in casualty load. At 40 casualties per day, WBB-supported supply fell to a median of 5 days (IQR, 2–7), and at 50 casualties per day, supply decreased to a median of 2 days (IQR, 2–4). Above this threshold, system failure approached immediacy, reflecting a non-linear collapse phenomenon unable to keep pace with sustained high-volume demand.

Beyond the timing of first failure, the proportion of cumulative demand met of a 30-day conflict horizon declined precipitously with escalating casualty rates. WBB augmentation shifted the demand-met curve modestly upward across all scenarios but did not restore durable blood supply performance once casualty intensity exceeded moderate thresholds. For example, demand fulfillment approached a functional inflection point around 50 casualties per day, at which point both conventional supply and WBB augmentation met only about half of the total projected blood demand. This finding has operational implications. In LSCO settings where resupply is contested and casualty generation is stochastic, falling below the majority demand-fulfilled threshold implies that rationing decisions, triage constraints, and preventable mortality become inevitable rather than isolated events.

These findings are particularly consequential because an average casualty rate of 30 per day to a single FRSD, sufficient to produce rapid collapse in this model, represents a conservative estimate for LSCO. Historical analyses of major conventional conflicts and contemporary war-gaming forecasts anticipate total battlefield casualties in the thousands per day across multiple FRSDs.¹⁴ Our results suggest that even modest allocation of that total load to a single FRSD quickly exceeds the sustainment capacity assumed by current blood supply guidance, indicating that doctrinal planning at the unit level requires adaptation to future operational environments.

Importantly, observed failures were driven by acute unmet demand rather than inventory spoilage. Blood products were consumed long before expiration, demonstrating that under high-intensity conditions, the limiting factor is not wastage but the fundamental mismatch between consumption and replenishment capacity. This distinction is critical for medical logistics planning, where strategies focusing on delivery cadence, throughout, and expansion become the dominant determinant of performance.

The results also clarify the role and limits of WBB augmentation in LSCO. While WBB capability provides meaningful surge buffering and has proven instrumental in counterinsurgency contingency planning, it is inherently bound by donor pool size, regeneration, and operational practicality. In this model, the 56-day donor deferral period and finite donor availability prevent WBB systems from sustaining repeated high-volume demand, reinforcing that WBBs function as a bridge for transient spikes rather than a scalable solution for prolonged high-intensity operations. In addition, donor mobilization entails a non-trivial operational trade-off, as personnel availability and unit cohesion may already be degraded. From a doctrinal perspective, this supports the principle that WBBs buy time but do not replace the need for robust, reliable resupply under LSCO conditions.

Closing the gap between projected demand and feasible forward supply will require a layered strategy that spans policy adjustments and capability development. The forecasted blood supply shortages in LSCO additionally warrant renewed interest in utilizing alternative human-derived emergency donor pathways to include wounded-in-action, disease non-battle injured, or even killed-in-action deceased-donor strategies. The concept of postmortem blood procurement has historical precedent, having been explored before the development of modern blood banking systems, but was largely abandoned as reliable donor programs and component therapy became established.²⁷ Although recently deceased casualties may theoretically represent a transient source of viable blood if collected rapidly after death, significant ethical, legal, and logistical challenges remain, including consent, infection screening, and operational feasibility.²⁸ In parallel, continued research is needed to operationalize next-generation products to mitigate anticipated human-derived shortages. Hemoglobin-based oxygen carriers (HBOCs), manufactured red blood cells, and advanced preservation technologies, including freeze-dried and spray-dried formulations, represent strategies to extend transfusion capacity in austere environments.^29,30 Recognizing the operational constraints imposed by cold chain dependence, DARPA’s Fieldable Solutions for Hemorrhage with bio-Artificial Resuscitation Products (FSHARP) program is actively developing shelf-stable WB substitutes intended to provide resuscitation support without refrigeration.³¹ Although these technologies remain under development, they highlight ongoing efforts to create deployable blood substitutes capable of augmenting conventional blood supplies. Beyond blood substitution, pharmacologic adjuncts such as arginine vasopressin, poly (ADP-ribose) polymerase, PEG-20k, and other anti-shock therapeutics warrant continued investigation for their potential to augment perfusion and reduce transfusion requirements in hemorrhagic shock.^32–35

The model’s planning factor of 8 units per transfused casualty could also reflect conservative assumptions about blood use. Combat surgical cohort data from the GWOT demonstrate that critically injured patients frequently require substantially higher transfusion volumes, with a median of ∼15 units of blood products administered within the first 24 hours after injury (IQR, 6–30 units). Notably, 28% of patients required 10–20 units, while 36% required more than 20 units, highlighting the significant strain that a single severely injured casualty can place on a constrained blood supply system.³⁶ These transfusion requirements reflect cumulative blood utilization across the continuum of care, from point of injury through Role III evacuation and resuscitation, and underscore the magnitude of demand associated with severe hemorrhagic trauma. Importantly, during LSCO, Role II facilities may become increasingly isolated as the battlespace expands and evacuation is degraded, possibly increasing transfusion requirements due to delayed definitive hemorrhage control and prolonged resuscitation. Accordingly, the depletion timelines presented here should be interpreted as optimistic estimates under sustained high-intensity combat operations, reinforcing that vulnerability persists even under deliberately conservative measures.

Lastly, these results prompt the development of adaptive decision support systems capable of optimizing blood distribution under uncertainty. Predictive analytics and machine learning approaches leveraging operational and medical data streams could enable dynamic forecasting of casualty flow, transfusion demand, and supply chain stress, supporting proactive redistribution and pre-positioning rather than reactive depletion.³⁷ Early proof-of-concept work in automated battlefield trauma systems suggests that integrated networks can improve situational awareness and resource allocation for scarce commodities, such as blood.³⁸ In future degraded environments, semi- or fully autonomous logistics decision-support may be essential to sustain resuscitation capacity independent of human oversight when resupply corridors are contested.^39,40

Limitations

This study has several limitations. First, the simulation intentionally abstracts the complex, dynamic realities of LSCO into a tractable operational model. Although key parameters were derived from historical combat data and current doctrinal guidance, real-world casualty generation, injury severity, transfusion requirements, and resupply conditions are highly variable and influenced by tactical context, environmental factors, and operational tempo, which the model cannot fully capture.

Second, the model relied on simplified transfusion demand assumptions, including a fixed proportion of casualties requiring transfusion and a fixed blood requirement per transfused casualty. In reality, transfusion requirements are heterogeneous and usually skewed, with a minority of massively transfused casualties consuming a disproportionately large share of total blood products. Consequently, the modeled demand may underestimate the consumption during periods of severe injury patterns, prolonged casualty holding, and delayed definitive hemorrhage control.

Third, WBB performance was modeled under idealized assumptions due to limited empirical data describing donor availability and regeneration under sustained LSCO stress. The simulation assumed a stable donor pool with fixed deferral intervals and immediate donor availability, and did not explicitly model donor attrition, operational tasking, or the time and manpower required for collection, screening, and delivery. These assumptions likely overestimate the practical contribution of WBB augmentation under contested conditions.

Fourth, the model did not incorporate several operational constraints that would be expected to degrade performance in near-peer conflict, including refrigeration failure, supply chain disruption, resupply reliability, delays in casualty evacuation, and prolonged casualty holding. The model was also blood type agnostic and did not account for compatibility constraints, testing delays, handling-related waste, or the need for adjunctive products and component therapy. WB was therefore treated as a surrogate for total resuscitation capacity rather than a complete representation of all transfusion requirements. In addition, the model represents a single Role 2 FRSD as an isolated node and does not simulate network-level behavior such as redistribution between elements, theater-level pooling, prioritization, or adaptive allocation strategies, which could assist in shortage mitigation in the absence of synchronized casualty events.

Finally, as a computational analysis, the findings reflect theoretical performance under predefined assumptions rather than empirical outcomes from LSCO-scale field validation. While Monte Carlo simulation provides probabilistic insight into stochastic failure dynamics, future work should refine and validate parameters using data from operational experience and exercises.

CONCLUSION

This discrete-event Monte Carlo simulation demonstrates that current military blood supply strategies, even under idealized “best-case” assumptions, are not sustainable at casualty rates anticipated in LSCO. In modeled LSCO scenarios, blood supply rapidly deteriorated at 30 casualties per day and collapsed beyond this threshold, with only ∼50% of blood supply demand met over the 30-day period at 50 casualties per day. System failure was driven by acute unmet demand rather than by storage expiration, indicating that the primary constraint is supply throughput, not inventory management. WBB augmentation delayed failure but remained a short-duration bridge rather than a scalable sustainment solution. Future readiness, therefore, requires doctrinal reform and a layered sustainment approach that includes a resilient resupply network. Blood product support must be treated as a strategic commodity comparable to ammunition and fuel, as insufficient transfusion capacity may become a determinant of casualty survivability and operational endurance in LSCO.

Supplementary Material

ta-101-s017-s001.pdf ^{(263.9KB, pdf)}

ta-101-s017-s002.pdf ^{(525.6KB, pdf)}