[Skip to Navigation]
Sign In
Figure 1.  Decision Tree and Markov Model
Decision Tree and Markov Model

A, Decision tree comparing prehospital thawed plasma with standard care prehospital fluid resuscitation stratified by injury mechanism. B, Markov model for patients surviving 30 days posttrauma.

Figure 2.  Cost-effectiveness Acceptability Curves for Thawed Plasma and Standard Care Resuscitation
Cost-effectiveness Acceptability Curves for Thawed Plasma and Standard Care Resuscitation

Results of 10 000 iterations of a Monte Carlo probabilistic sensitivity analysis represented as cost-effectiveness acceptability curves across willingness-to-pay thresholds. At a willingness-to-pay threshold of $100 000, thawed plasma was cost-effective in 8140 of the iterations (81.4%).

Figure 3.  Two-Way Sensitivity Analysis of the Number of Air Medical Bases and Annual Trauma Volume
Two-Way Sensitivity Analysis of the Number of Air Medical Bases and Annual Trauma Volume

Two-way sensitivity analysis of the number of air medical bases per trauma center and the estimated average annual volume of trauma patients meeting PAMPer inclusion criteria. The blue area represents combinations of input variables where thawed plasma is the cost-effective strategy. The orange area represents combinations of input variables where standard care is the cost-effective strategy. With an annual eligible trauma volume of 5 patients, the average number of air medical bases per center needs to double for standard care to become the preferred strategy.

Table 1.  Decision Tree Input Assumptions
Decision Tree Input Assumptions
Table 2.  Incremental Cost, Effectiveness, and Cost-effectiveness Ratios Reported in Quality-Adjusted and Unadjusted Life-Years
Incremental Cost, Effectiveness, and Cost-effectiveness Ratios Reported in Quality-Adjusted and Unadjusted Life-Years
Supplement.

eMethods 1. Comparison of 24-hour transfusion volumes, ventilator days, intensive care unit, and hospital lengths of stay for each treatment by injury mechanism

eMethods 1. Comparison of 24-hour transfusion volumes, ventilator days, intensive care unit, and hospital lengths of stay for each treatment by injury mechanism

eTable 1. Average resource use by treatment received and injury mechanism

eMethods 2. Estimation of the annual plasma program cost per base

eTable 2. Annual thawed plasma costs per air medical base by travel distance and efficiency in 2017 USD

eMethods 3. Estimation of the annual plasma program cost per trauma center

eTable 3. Annual thawed plasma costs per trauma center by number of air medical bases

eMethods 4. Estimation of the annual plasma program cost per patient

eTable 4. Percentage of patients meeting PAMPer trial inclusion criteria (NTDB 2006-2017)

eTable 5. Air medical transport data from the University of Pittsburgh Medical Center (2019 estimates)

eTable 6. Annual PAMPer-eligible patient volume estimates for participating centers

eMethods 5. Input parameter distributions used in Monte Carlo probabilistic sensitivity analysis

eTable 7. Input parameter distributions used in Monte Carlo probabilistic sensitivity analysis

eMethods 6. Breakdown of the incremental cost by cost category

eTable 8. Incremental cost analysis by cost category

eFigure 1. One-way sensitivity analysis (tornado diagram) for model inputs

eFigure 2. Two-way sensitivity analyses of the number of air medical bases supplied by a given trauma center and the annual volume of patients meeting PAMPer inclusion criteria at various willingness-to-pay thresholds

eFigure 3. Two-way sensitivity analysis of 30-day mortality following blunt trauma for patients receiving thawed plasma and the proportion of blunt trauma

eReferences

1.
Xu  J, Murphy  SL, Kockanek  KD, Arias  E.  Mortality in the United States, 2018.   NCHS Data Brief. 2020;(355):1-8.PubMedGoogle Scholar
2.
Davis  JS, Satahoo  SS, Butler  FK,  et al.  An analysis of prehospital deaths: who can we save?   J Trauma Acute Care Surg. 2014;77(2):213-218. doi:10.1097/TA.0000000000000292PubMedGoogle ScholarCrossref
3.
Holcomb  JB, Jenkins  D, Rhee  P,  et al.  Damage control resuscitation: directly addressing the early coagulopathy of trauma.   J Trauma. 2007;62(2):307-310. doi:10.1097/TA.0b013e3180324124PubMedGoogle Scholar
4.
Holcomb  JB, Tilley  BC, Baraniuk  S,  et al; PROPPR Study Group.  Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial.   JAMA. 2015;313(5):471-482. doi:10.1001/jama.2015.12PubMedGoogle ScholarCrossref
5.
Sperry  JL, Guyette  FX, Brown  JB,  et al; PAMPer Study Group.  Prehospital plasma during air medical transport in trauma patients at risk for hemorrhagic shock.   N Engl J Med. 2018;379(4):315-326. doi:10.1056/NEJMoa1802345PubMedGoogle ScholarCrossref
6.
Adams  PW, Warren  KA, Guyette  FX,  et al; PAMPer Study Group.  Implementation of a prehospital air medical thawed plasma program: is it even feasible?   J Trauma Acute Care Surg. 2019;87(5):1077-1081. doi:10.1097/TA.0000000000002406PubMedGoogle ScholarCrossref
7.
Husereau  D, Drummond  M, Petrou  S,  et al; ISPOR Health Economic Evaluation Publication Guidelines-CHEERS Good Reporting Practices Task Force.  Consolidated Health Economic Evaluation Reporting Standards (CHEERS)—explanation and elaboration: a report of the ISPOR Health Economic Evaluation Publication Guidelines Good Reporting Practices Task Force.   Value Health. 2013;16(2):231-250. doi:10.1016/j.jval.2013.02.002PubMedGoogle ScholarCrossref
8.
Sapiano  MRP, Jones  JM, Savinkina  AA, Haass  KA, Berger  JJ, Basavaraju  SV.  Supplemental findings of the 2017 National Blood Collection and Utilization Survey.   Transfusion. 2020;60(S2)(suppl 2):S17-S37. doi:10.1111/trf.15715PubMedGoogle ScholarCrossref
9.
Owens  PL, Russo  CA, Stocks  C. Frequency and costs of hospital admissions for injury, 2004: Statistical Brief #18. Published November 2006. Accessed October 10, 2020. https://www.ncbi.nlm.nih.gov/books/NBK63490/
10.
Dasta  JF, McLaughlin  TP, Mody  SH, Piech  CT.  Daily cost of an intensive care unit day: the contribution of mechanical ventilation.   Crit Care Med. 2005;33(6):1266-1271. doi:10.1097/01.CCM.0000164543.14619.00PubMedGoogle ScholarCrossref
11.
Brown  JB, Smith  KJ, Gestring  ML,  et al.  Comparing the air medical prehospital triage score with current practice for triage of injured patients to helicopter emergency medical services: a cost-effectiveness analysis.   JAMA Surg. 2018;153(3):261-268. doi:10.1001/jamasurg.2017.4485PubMedGoogle ScholarCrossref
12.
MacKenzie  EJ, Rivara  FP, Jurkovich  GJ,  et al.  A national evaluation of the effect of trauma-center care on mortality.   N Engl J Med. 2006;354(4):366-378. doi:10.1056/NEJMsa052049PubMedGoogle ScholarCrossref
13.
Delgado  MK, Staudenmayer  KL, Wang  NE,  et al.  Cost-effectiveness of helicopter versus ground emergency medical services for trauma scene transport in the United States.   Ann Emerg Med. 2013;62(4):351-364.e19. doi:10.1016/j.annemergmed.2013.02.025PubMedGoogle ScholarCrossref
14.
Cameron  CM, Purdie  DM, Kliewer  EV, McClure  RJ.  Long-term mortality following trauma: 10 year follow-up in a population-based sample of injured adults.   J Trauma. 2005;59(3):639-646.PubMedGoogle Scholar
15.
Arias  E, Bastian  B, Xu  J, Tejada-Vera  B.  US state life tables, 2018.   Natl Vital Stat Rep. 2021;70(1):1-18.PubMedGoogle Scholar
16.
Gold  MR, Franks  P, McCoy  KI, Fryback  DG.  Toward consistency in cost-utility analyses: using national measures to create condition-specific values.   Med Care. 1998;36(6):778-792. doi:10.1097/00005650-199806000-00002PubMedGoogle ScholarCrossref
17.
Sanders  GD, Neumann  PJ, Basu  A,  et al.  Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: second panel on cost-effectiveness in health and medicine.   JAMA. 2016;316(10):1093-1103. doi:10.1001/jama.2016.12195PubMedGoogle ScholarCrossref
18.
Cameron  CM, Purdie  DM, Kliewer  EV, McClure  RJ.  Ten-year health service use outcomes in a population-based cohort of 21 000 injured adults: the Manitoba Injury Outcome Study.   Bull World Health Organ. 2006;84(10):802-810. doi:10.2471/BLT.06.030833PubMedGoogle ScholarCrossref
19.
US Bureau of Labor Statistics. CPI inflation calculator. Accessed October 10, 2020. https://www.bls.gov/data/inflation_calculator.htm
20.
Eichler  HG, Kong  SX, Gerth  WC, Mavros  P, Jönsson  B.  Use of cost-effectiveness analysis in health-care resource allocation decision-making: how are cost-effectiveness thresholds expected to emerge?   Value Health. 2004;7(5):518-528. doi:10.1111/j.1524-4733.2004.75003.xPubMedGoogle ScholarCrossref
21.
The World Bank. GDP per capita. Accessed November 2, 2020. https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?name_desc=true&locations=US
22.
Gruen  DS, Guyette  FX, Brown  JB,  et al.  Characterization of unexpected survivors following a prehospital plasma randomized trial.   J Trauma Acute Care Surg. 2020;89(5):908-914. doi:10.1097/TA.0000000000002816PubMedGoogle ScholarCrossref
23.
Moore  HB, Moore  EE, Chapman  MP,  et al.  Plasma-first resuscitation to treat haemorrhagic shock during emergency ground transportation in an urban area: a randomised trial.   Lancet. 2018;392(10144):283-291. doi:10.1016/S0140-6736(18)31553-8PubMedGoogle ScholarCrossref
24.
Pusateri  AE, Moore  EE, Moore  HB,  et al.  Association of prehospital plasma transfusion with survival in trauma patients with hemorrhagic shock when transport times are longer than 20 minutes: a post hoc analysis of the PAMPer and COMBAT clinical trials.   JAMA Surg. 2020;155(2):e195085. doi:10.1001/jamasurg.2019.5085PubMedGoogle Scholar
25.
Gruen  DS, Brown  JB, Guyette  FX,  et al; PAMPer Study Group.  Prehospital plasma is associated with distinct biomarker expression following injury.   JCI Insight. 2020;5(8):e135350. doi:10.1172/jci.insight.135350PubMedGoogle Scholar
26.
Gruen  DS, Guyette  FX, Brown  JB,  et al.  Association of prehospital plasma with survival in patients with traumatic brain injury: a secondary analysis of the PAMPer cluster randomized clinical trial.   JAMA Netw Open. 2020;3(10):e2016869. doi:10.1001/jamanetworkopen.2020.16869PubMedGoogle Scholar
27.
Reitz  KM, Moore  HB, Guyette  FX,  et al.  Prehospital plasma in injured patients is associated with survival principally in blunt injury: results from two randomized prehospital plasma trials.   J Trauma Acute Care Surg. 2020;88(1):33-41. doi:10.1097/TA.0000000000002485PubMedGoogle ScholarCrossref
28.
Paisley  S.  Identification of evidence for key parameters in decision-analytic models of cost effectiveness: a description of sources and a recommended minimum search requirement.   Pharmacoeconomics. 2016;34(6):597-608. doi:10.1007/s40273-015-0372-xPubMedGoogle ScholarCrossref
Original Investigation
September 22, 2021

Evaluating the Cost-effectiveness of Prehospital Plasma Transfusion in Unstable Trauma Patients: A Secondary Analysis of the PAMPer Trial

Author Affiliations
  • 1Department of Surgery, University of Pittsburgh, Pittsburgh, Pennsylvania
  • 2Division of Trauma & Acute Care Surgery, Department of Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
  • 3Department of Emergency Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania
  • 4Division of Trauma & Critical Care Surgery, Department of Surgery, University of Tennessee Graduate School of Medicine, Knoxville
  • 5Department of Surgery, JPS Health Network, Ft Worth, Texas
  • 6Division of Trauma Surgery, Department of Surgery, University of Louisville, Louisville, Kentucky
  • 7Division of Trauma & Critical Care, Metrohealth Medical Center, Case Western Reserve University, Cleveland, Ohio
  • 8Division of Burn Surgery, Department of Surgery, Louisiana State University Health Sciences Center, New Orleans
  • 9Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania
JAMA Surg. 2021;156(12):1131-1139. doi:10.1001/jamasurg.2021.4529
Key Points

Question  Is thawed plasma transfusion for trauma patients in hemorrhagic shock during air medical transport cost-effective compared with standard care resuscitation?

Findings  In a cost-effectiveness model using data from a randomized multicenter clinical trial, prehospital thawed plasma was more cost-effective compared with standard care resuscitation with an incremental cost-effectiveness ratio of $50 467.44 per quality-adjusted life-year (QALY). The incremental cost of thawed plasma per QALY remained below the standard $100 000 per QALY threshold across sensitivity analyses.

Meaning  Prehospital thawed plasma may be lifesaving and cost-effective and should become a standard component of resuscitation for trauma patients in hemorrhagic shock during air transport.

Abstract

Importance  Prehospital plasma transfusion is lifesaving for trauma patients in hemorrhagic shock but is not commonly used owing to cost and feasibility concerns.

Objective  To evaluate the cost-effectiveness of prehospital thawed plasma transfusion in trauma patients with hemorrhagic shock during air medical transport.

Design, Setting, and Participants  A decision tree and Markov model were created to compare standard care and prehospital thawed plasma transfusion using published and unpublished patient-level data from the Prehospital Plasma in Air Medical Transport in Trauma Patients at Risk for Hemorrhagic Shock (PAMPer) trial conducted from May 2014 to October 2017, health care and trauma-specific databases, and the published literature. Prehospital transfusion, short-term inpatient care, and lifetime health care costs and quality of life outcomes were included. One-way, 2-way, and Monte Carlo probabilistic sensitivity analyses were performed across clinically plausible ranges. Data were analyzed in December 2019.

Main Outcomes and Measures  Relative costs and health-related quality of life were evaluated by an incremental cost-effectiveness ratio at a standard willingness-to-pay threshold of $100 000 per quality-adjusted life-year (QALY).

Results  The trial included 501 patients in the modified intention-to-treat cohort. Median (interquartile range) age for patients in the thawed plasma and standard care cohorts were 44 (31-59) and 46 (28-60) years, respectively. Overall, 364 patients (72.7%) were male. Thawed plasma transfusion was cost-effective with an incremental cost-effectiveness ratio of $50 467.44 per QALY compared with standard care. The preference for thawed plasma was robust across all 1- and 2-way sensitivity analyses. When considering only patients injured by a blunt mechanism, the incremental cost-effectiveness ratio decreased to $37 735.19 per QALY. Thawed plasma was preferred in 8140 of 10 000 iterations (81.4%) on probabilistic sensitivity analysis. A detailed analysis of incremental costs between strategies revealed most were attributable to the in-hospital and postdischarge lifetime care of critically ill patients surviving severe trauma.

Conclusions and Relevance  In this study, prehospital thawed plasma transfusion during air medical transport for trauma patients in hemorrhagic shock was lifesaving and cost-effective compared with standard care and should become commonplace.

Introduction

Trauma is the third leading cause of mortality in the US, accounting for 48 deaths per 100 000 persons in 2018.1 Hemorrhage is the most common preventable cause of trauma-related death.2 Identifying optimal early resuscitation strategies for patients in hemorrhagic shock is essential for preventing coagulopathy, exsanguination, and cardiac arrest. There is robust evidence to support a balanced transfusion strategy approximating whole blood, with a 1:1:1 ratio of plasma to platelets to packed red blood cells both in hospital and in the prehospital environment3,4; however, prehospital volume resuscitation strategies for patients with active hemorrhage are limited by the feasibility and availability of blood products. Recently, the multicenter, pragmatic Prehospital Plasma in Air Medical Transport in Trauma Patients at Risk for Hemorrhagic Shock (PAMPer) trial demonstrated reduced 30-day mortality for patients receiving prehospital thawed plasma (TP).5 Despite this benefit, concerns about cost and waste associated with TP have restricted its use in the prehospital setting. A recent study estimated that the cost of maintaining TP stores for prehospital resuscitation during air medical transport may exceed $24 000 per helicopter base annually.6 The primary objective of this study was to assess the cost-effectiveness of implementing a prehospital TP program using patient-level data from the PAMPer trial. We hypothesized that supporting prehospital TP is cost-effective for many trauma centers but is dependent on annual volumes of trauma patients in hemorrhagic shock transported by air and on plasma use and recycling capabilities.

Methods
PAMPer Trial Description

This cost-utility analysis was performed using published outcomes and unpublished patient-level data from the PAMPer trial, which enrolled 501 patients from 9 trauma centers using 27 air medical bases between May 2014 and October 2017.5 All patients transported by air from the scene of injury or a referral emergency department to a participating trauma center were considered for inclusion. Inclusion criteria included at least 1 episode of hypotension (systolic blood pressure less than 90 mm Hg) and tachycardia (heart rate greater than 108 beats per minute) or any severe hypotension (systolic blood pressure less than 70 mm Hg) during transport. The intervention was transfusion of 2 units of TP during transport. The control group received crystalloid and up to 2 units of universal donor packed red blood cells, as dictated by local protocols. Each month, air medical bases were randomized to deliver the intervention or standard care (SC) resuscitation to eligible patients, and intraclass correlation was used to account for this cluster design. The primary outcome, 30-day mortality, was assessed by a modified intention-to-treat analysis. Multiple imputation was used for patients with unknown 30-day vital status. To our knowledge, this trial is the sole source of randomized, prospective, multicentric data regarding the efficacy and safety of prehospital TP during air transport and therefore served as the primary data source for input values for cost-effectiveness analysis. Data were analyzed in December 2019. University of Pittsburgh institutional review board approval was obtained for the original PAMPer trial at each participating site. Informed consent was waived after consultation with community members and public notification regarding the trial took place, as this was an intervention delivered in the emergency setting, often in patients unable to provide consent.

Model Design

A decision model was created using TreeAge Pro version R2.0 (TreeAge Software) to perform a cost-utility analysis from the payer perspective for 2 prehospital volume resuscitation strategies: SC and TP (Figure 1A). Demographic characteristics for the base case scenario were determined from PAMPer trial data (median [interquartile range] age, 45 [29-60] years; 72.6% male), as effectiveness measures used in the model were obtained from this population.5,7 Outcomes were stratified by injury mechanism to account for differences in resource use and outcomes for blunt vs penetrating trauma. Patients surviving 30 days were included in a Markov process to model long-term costs and outcomes (Figure 1B). In the first cycle, patients could die within 1 year of trauma or survive. Surviving patients entered a posttrauma state in which they could remain alive or die from background mortality. The Markov cycle length was 1 year, and the time horizon was 120 years, ensuring all patients entered the death state and incurred complete lifetime costs and benefits. This analysis was performed in accordance with the CHEERs reporting standards for cost-effectiveness analyses.7

Model Assumptions

Events, represented by branches at each chance node, were mutually exclusive and collectively exhaustive. Branches were populated with probabilities based on the relative incidence of blunt to penetrating trauma and 30-day mortality rates for each mechanism (Table 1).5,6,8-16

There were no differences in the rates of prehospital intubation, cardiopulmonary resuscitation, or injury severity between groups. Prehospital resuscitation costs were therefore considered equivalent and excluded from analysis except for direct intervention-related costs. The probability and cost associated with prehospital packed red blood cells transfusion were included given the difference between treatment arms. Prehospital and 24-hour transfusion volumes for all products were obtained from patient-level trial data. The incidence of adverse trial events and resuscitation-related complications were equivalent between arms and also excluded from the model.5

A Wilcoxon rank sum test was used to compare hospital lengths of stay, intensive care unit lengths of stay, and ventilator days between groups and across mechanisms. In the TP group, intensive care unit lengths of stay, ventilator days, and hospital lengths of stay were statistically different by injury mechanism at a 2-tailed P value threshold of 0.05 (eMethods 1 and eTable 1 in the Supplement). These costs were therefore included in the analysis.

For patients entering the Markov model, the probability of mortality within 1 year of trauma was obtained from published sources.11,12 Annual mortality rates beyond 1 year following severe trauma were calculated by applying a trauma multiplier of 7.33 to baseline mortality rates from US life tables.13-15 The primary outcome of this study was the incremental cost-effectiveness ratio (ICER), which represents the cost of obtaining an additional unit of benefit, measured in quality-adjusted life-years (QALYs).16,17 This value is derived by dividing the difference in costs by the difference in QALYs gained under each strategy. The secondary outcome was the cost of obtaining an additional unadjusted life-year. All costs and benefits were discounted at a rate of 3%.17

Costs

The total annual cost of a prehospital TP program was estimated as previously described.6 Attributable costs for plasma distribution, recovery, waste, and recycling were obtained from site-level accounting data. Estimates of carrying costs for each base by use and recycling efficiency were estimated from individual site data (eMethods 2 and eTable 2 in the Supplement). Rates of plasma use and recycling or waste ranged from 0% to 100% across sites allowing for a broad range of estimated carrying costs. The annual cost per base was multiplied by the average number of air medical bases per center to estimate the annual cost of supporting a TP program per trauma center (eMethods 3 and eTable 3 in the Supplement).

For the purposes of this analysis, it was necessary to distribute the total annual cost per trauma center on a per patient basis. This required estimating the total annual number of patients meeting PAMPer trial inclusion criteria per trauma center. This estimate was obtained using a combination of National Trauma Databank and individual site-level data (eTables 4, 5, and 6 and eMethods 4 in the Supplement). This volume estimate was varied widely in sensitivity analyses to account for inherent uncertainty.

Total 24-hour in-hospital blood, plasma, and platelet transfusion requirements were calculated from patient-level trial data for patients sustaining blunt vs penetrating trauma in each treatment arm, and the associated costs were included. The costs of blood components were obtained from the 2017 National Blood Collection and Utilization Survey (Table 1).8

Data from the Healthcare Cost and Utilization Project8 and published literature were used to estimate the average daily cost of hospitalization following traumatic injury, accounting for the contribution of intensive care unit admission and mechanical ventilation. The model reflects the major concentration of hospital-associated costs at the beginning of a hospitalization.9,10 Postdischarge costs for the first posttrauma year were estimated from the National Study on the Costs and Outcomes of Trauma.12 Lifetime health care costs for subsequent years were estimated by inflating costs obtained from the Centers for Medicare and Medicaid Services mean annual health expenditures report by a factor of 1.45 to account for injury severity.11,18All costs were converted to November 2020 dollars using the US Bureau of Labor Statistics’ Consumer Price Index.19

Outcomes and Utilities

The effectiveness of each strategy was estimated in unadjusted life-years and QALYs. Health state utilities were obtained from the literature for hospitalization, the first year following traumatic injury, and subsequent posttrauma years (Table 1).11-13 Postdischarge health state utilities were decreased by 30% to represent the effect of surviving severe injury on health-related quality of life.12 A disutility was applied to the QALYs gained for each cycle of the Markov model based on the patient’s age during that cycle to account for the effect of aging on quality of life.16

Statistical Analysis

To test the assumptions in this model, a 1-way sensitivity analysis was conducted on each input over a wide range of plausible values (Table 1). Two-way sensitivity analyses were performed to evaluate the effect of varying related inputs. To evaluate the impact of transfusing TP solely for patients, with blunt injuries, the probability of blunt injury was set to 1, and the number of eligible patients was reduced proportionally. A Monte Carlo probabilistic sensitivity analysis was performed to account for the uncertainty in all model inputs simultaneously, where values for every input parameter were randomly sampled 10 000 times from the distributions for each parameter (eMethods 5 and eTable 7 in the Supplement). From these results, a cost-effectiveness acceptability curve was generated to illustrate the proportion of simulations in which each strategy was preferred at a willingness-to-pay threshold (WTP) of $100 000 per QALY.

Results

The PAMPer trial included 501 patients in the modified intention-to-treat cohort. Median (interquartile range) age for patients in the thawed plasma and standard care cohorts were 44 (31-59) and 46 (28-60) years, respectively. Overall, 364 patients (72.7%) were male. Prehospital TP delivered 6.71 QALY at a cost of $241 256.47 compared with 5.85 QALY at a cost of $198 359.15 for SC, yielding an ICER of $50 467.44 per QALY. TP and SC delivered 14.76 and 12.88 unadjusted life-years, respectively, yielding an ICER of $31 318.99 per unadjusted life-year (Table 2). Detailed analysis of cost revealed that most of the incremental costs associated with TP was attributable to the in-hospital (TP, $75 857.49 vs SC, $64 698.62) and postdischarge lifetime care (TP, $160 697.06 vs SC, $131 197.72) of critically ill patients surviving severe trauma, accounting for 26.0% and 68.8% of the incremental cost, respectively (eMethods 6 and eTable 8 in the Supplement). Prehospital costs, including the per-patient cost of the TP program, were higher for the intervention (TP, $3317.20 vs SC, $186.32); however, only 7.3% of the total incremental cost was attributable to prehospital resuscitation. Patients receiving TP required fewer packed red blood cell (4.9 vs 6.9), platelet (0.6 vs 1.2), and plasma (2.3 vs 3.3) transfusions in the first 24 hours postinjury, generating an average transfusion-related cost savings of $891.77.

On 1-way sensitivity analysis, the model was most sensitive to variations in the percentage of blunt trauma, patient age, and mortality rate following blunt trauma (eFigure 1 in the Supplement). Clinically reasonable variations in model inputs did not change strategy preference at a WTP of $100 000 until the median patient age increased by 34.8% to 69 years, or the proportion of blunt to total trauma decreased to 49.0%. For all other variables, TP remained the preferred strategy across all values within the sensitivity analysis range. Plasma use and recycling capabilities of individual centers did not have a large impact on the results. The ICER for TP at centers with 0% or 100% plasma recycling were $49 729 per QALY and $52 435.80 per QALY, respectively. A 2-way sensitivity analysis varying the numbers of bases and eligible patients showed TP remained cost-effective at centers when the average number of air bases was doubled and the volume of trauma patients in hemorrhagic shock transported via air was as low as 5 patients per year at a WTP threshold of $100 000/QALY (Figure 2). This analysis was expanded across various WTP thresholds and revealed TP remained the preferred strategy for centers with 4 air medical bases and annual eligible patient volumes of 40 and 10 patients at WTP thresholds of $50 000 per QALY and $75 000 per QALY, respectively (eFigure 2 in the Supplement). A 2-way sensitivity analysis of the proportion of blunt trauma and mortality following blunt trauma and receipt of TP revealed a robust preference for TP (eFigure 3 in the Supplement). The ICER decreased to $37 735.19 per QALY when only considering patients with blunt injuries. Probabilistic sensitivity analysis showed TP was the preferred strategy in 8140 of 10 000 iterations (81.4%) at a WTP of $100 000 per QALY (Figure 3).

Discussion

This cost-effectiveness analysis demonstrated that prehospital plasma transfusion in hemodynamically unstable trauma patients transported by air was cost-effective compared with standard prehospital resuscitation at a WTP threshold of $100 000. This is a conservative cutoff according to World Health Organization guidelines, which suggest that interventions costing less than 3 times the national gross domestic product per capita represent good value.20 The 2017 per capita gross domestic product in the US was $60 062.22, which exceeds the base case ICER of $50 467.44 per QALY.21

Despite level 1 evidence demonstrating a survival benefit associated with prehospital TP, incorporation of this strategy into resuscitation algorithms during air transport has been limited, likely because of concerns regarding the economic and logistical burden of supporting these programs. Although prior analyses have quoted an estimated annual cost of $24 343 to $30 077 to supply an air base with TP,6 a formal cost-effectiveness analysis is necessary to illustrate how this cost compares with other resuscitation strategies when distributed across the benefits derived over the lifetime of the patient. We demonstrated that more than 90% of the incremental cost associated with TP was attributed to the hospital and lifetime care costs incurred by increased survival of severely injured patients. Post hoc analysis of the PAMPer trial demonstrated increased unexpected survivorship in patients receiving TP.22 Sustaining the most critically injured trauma survivors is an unequivocally costly but worthwhile endeavor.

The cost-effectiveness of prehospital TP depends on center-level characteristics, including the number of air bases supplied, the annual volume of trauma patients meeting PAMPer inclusion criteria, and the proportion of blunt trauma injuries. However, 1- and 2-way sensitivity analyses failed to reveal clinically meaningful scenarios wherein prehospital TP was not cost-effective. A 2-way sensitivity analysis of annual patient volume and number of bases per trauma center showed that SC became the preferred strategy when the number of bases exceeded the annual volume of eligible patients, a highly improbable scenario. Detailed examination of the 18% of iterations on probabilistic sensitivity analysis in which SC was preferred revealed combinations of older patients, fewer instances of blunt trauma, and high hospital or lifetime health care costs.

It is important to note that this analysis showed that prehospital TP was cost-effective when delivered to hypotensive trauma patients during air transport, as dictated by PAMPer trial protocols and inclusion criteria. The single-center Control of Major Bleeding After Trauma (COMBAT) trial of prehospital TP vs SC resuscitation in trauma patients during ground transport showed equivalent mortality between treatment arms, and the authors concluded prehospital TP was unjustified in a metropolitan setting with short travel distances.23 A combined post hoc analysis demonstrated a survival benefit for patients in both trials when transport times exceeded 20 minutes.24 The interquartile range for transport times in the PAMPer trial study cohort was 33 to 52 minutes. Routine administration of prehospital plasma during air transport for this population with these travel times was clinically effective as well as cost-effective. Affirming the cost-effectiveness of TP for certain patients in the ground transport setting requires additional prospective data from trials designed to select for those patients likely to benefit from TP, such as those with estimated transport times longer than 20 minutes, and accurate cost estimates for supporting a TP program at ground bases.

Secondary analyses of PAMPer and COMBAT have shown even greater benefit associated with TP in certain subgroups including blunt trauma, multiple-trauma, and traumatic brain injury.25-27 When considering only patients with blunt injuries, the ICER decreased from $50 467.44 to $37 735.19 per QALY. Restricting prehospital plasma to patients with multiple traumas and traumatic brain injuries may further increase the cost-effectiveness of this strategy; however, the decision to deliver prehospital plasma is made prior to a thorough understanding of the patient’s injury complex and must therefore be guided by easily ascertainable data, such as vital signs or injury mechanism. Additionally, this analysis showed prehospital TP was cost-effective even without selecting for these subgroups.

There are additional scenarios not discussed in this analysis wherein prehospital plasma during air transport may be increasingly cost-effective. States with centralized air medical transport systems may institute and coordinate statewide plasma distribution, restocking, and recycling to support more efficient resource use and spread costs over a greater number of eligible patients. Plasma formulations with greater shelf stability, such as liquid and freeze dried plasma, have the potential to further improve the cost-effectiveness of this strategy by reducing courier requirements and waste.

Limitations

This study has limitations. Model inputs were derived from multicenter data with variations across trial sites. Annual eligibility was estimated from single-site internal data and multisite enrollment data, which may not reflect actual trauma volume. Prehospital plasma resuscitation costs were estimated using accounting data from a small number of sites. Differences in the anatomy of air transport systems between centers and states may alter the logistics and costs associated with this strategy. In addition to true costs, hospital charges were used and may not accurately reflect actual payments by the health care system. The impact of variation in regional regulations and emergency medical services protocols, including air ambulance staffing for transfusion, on total costs was not considered. We attempted to account for uncertainty and variability with broad sensitivity analyses; however, certain values may fall outside the designated parameters of this model. In particular, there are limited data to support long-term cost projections for patients with a history of severe trauma. The long-term cost estimate was based on best available data identified by adhering to minimum search requirements.28 Beyond the first posttrauma year, we assumed the effect of trauma on health-related quality of life remained constant regardless of the time elapsed since injury and hospitalization. This analysis ignores nonmonetary costs, including environmental and social impacts. Additionally, although this analysis shows prehospital TP to be cost-effective, it does not answer the question of who should pay for this service or provide a model of how this cost could be passed on to various stakeholders, such as states, trauma centers, patients, or society.

Conclusions

In this study, prehospital TP was cost-effective across a range of patient and center-level variables. The cost of a prehospital TP program was small when distributed across patients and QALYs. The bulk of the observed incremental cost was attributable to in-hospital and lifetime care associated with improved survivorship after prehospital TP. Plasma should be a standard component of prehospital resuscitation during the air medical transport of trauma patients in hemorrhagic shock.

Back to top
Article Information

Accepted for Publication: July 13, 2021.

Published Online: September 22, 2021. doi:10.1001/jamasurg.2021.4529

Corresponding Author: Jason L. Sperry, MD, MPH, Division of Trauma & Acute Care Surgery, Department of Surgery, University of Pittsburgh Medical Center, 200 Lothrop St, Pittsburgh, PA 15213 (sperryjl@upmc.edu).

Author Contributions: Drs Hrebinko and Nicholson had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Hrebinko, Sperry, Miller, Neal, Yazer, Nicholson.

Acquisition, analysis, or interpretation of data: Hrebinko, Sperry, Guyette, Brown, Daley, Miller, Harbrecht, Claridge, phelan, Zuckerbraun, Nicholson.

Drafting of the manuscript: Hrebinko, Nicholson.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Hrebinko, Sperry, Brown, Nicholson.

Obtained funding: Sperry, Guyette, Nicholson.

Administrative, technical, or material support: Hrebinko, Daley, Miller, Harbrecht, Claridge, Neal, Yazer.

Supervision: Sperry, Brown, Claridge, Zuckerbraun, Nicholson.

Conflict of Interest Disclosures: Drs Sperry, Guyette, Claridge, and Phelan report grants from the Department of Defense during the conduct of the study. Dr Harbrecht reports grants from the Department of Defense during the conduct of the study and grants from the Department of Defense and personal fees from the American College of Surgeons Committee on Trauma Verification, Review, and Consultation Program and the University of Pittsburgh Data and Safety Monitoring Board outside the submitted work. Dr Neal reports nonfinancial support from Haima Therapeutics, personal fees from Janssen Pharma and Haemonetics, and grants from Instrumentation Laboratory outside the submitted work. Dr Yazer reports personal fees from Terumo Corporation during the conduct of the study and from Verax Biomedical, Grifols, Octapharma, Cook Medical Incorporated, Macopharma, and Cerus Corporation outside the submitted work. Dr Nicholson reports grants from the National Institutes of Health during the conduct of the study. No other disclosures were reported.

Funding/Support: Dr Nicholson receives support from the National Center for Advancing Translational Sciences (TL1TR001858).

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

References
1.
Xu  J, Murphy  SL, Kockanek  KD, Arias  E.  Mortality in the United States, 2018.   NCHS Data Brief. 2020;(355):1-8.PubMedGoogle Scholar
2.
Davis  JS, Satahoo  SS, Butler  FK,  et al.  An analysis of prehospital deaths: who can we save?   J Trauma Acute Care Surg. 2014;77(2):213-218. doi:10.1097/TA.0000000000000292PubMedGoogle ScholarCrossref
3.
Holcomb  JB, Jenkins  D, Rhee  P,  et al.  Damage control resuscitation: directly addressing the early coagulopathy of trauma.   J Trauma. 2007;62(2):307-310. doi:10.1097/TA.0b013e3180324124PubMedGoogle Scholar
4.
Holcomb  JB, Tilley  BC, Baraniuk  S,  et al; PROPPR Study Group.  Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial.   JAMA. 2015;313(5):471-482. doi:10.1001/jama.2015.12PubMedGoogle ScholarCrossref
5.
Sperry  JL, Guyette  FX, Brown  JB,  et al; PAMPer Study Group.  Prehospital plasma during air medical transport in trauma patients at risk for hemorrhagic shock.   N Engl J Med. 2018;379(4):315-326. doi:10.1056/NEJMoa1802345PubMedGoogle ScholarCrossref
6.
Adams  PW, Warren  KA, Guyette  FX,  et al; PAMPer Study Group.  Implementation of a prehospital air medical thawed plasma program: is it even feasible?   J Trauma Acute Care Surg. 2019;87(5):1077-1081. doi:10.1097/TA.0000000000002406PubMedGoogle ScholarCrossref
7.
Husereau  D, Drummond  M, Petrou  S,  et al; ISPOR Health Economic Evaluation Publication Guidelines-CHEERS Good Reporting Practices Task Force.  Consolidated Health Economic Evaluation Reporting Standards (CHEERS)—explanation and elaboration: a report of the ISPOR Health Economic Evaluation Publication Guidelines Good Reporting Practices Task Force.   Value Health. 2013;16(2):231-250. doi:10.1016/j.jval.2013.02.002PubMedGoogle ScholarCrossref
8.
Sapiano  MRP, Jones  JM, Savinkina  AA, Haass  KA, Berger  JJ, Basavaraju  SV.  Supplemental findings of the 2017 National Blood Collection and Utilization Survey.   Transfusion. 2020;60(S2)(suppl 2):S17-S37. doi:10.1111/trf.15715PubMedGoogle ScholarCrossref
9.
Owens  PL, Russo  CA, Stocks  C. Frequency and costs of hospital admissions for injury, 2004: Statistical Brief #18. Published November 2006. Accessed October 10, 2020. https://www.ncbi.nlm.nih.gov/books/NBK63490/
10.
Dasta  JF, McLaughlin  TP, Mody  SH, Piech  CT.  Daily cost of an intensive care unit day: the contribution of mechanical ventilation.   Crit Care Med. 2005;33(6):1266-1271. doi:10.1097/01.CCM.0000164543.14619.00PubMedGoogle ScholarCrossref
11.
Brown  JB, Smith  KJ, Gestring  ML,  et al.  Comparing the air medical prehospital triage score with current practice for triage of injured patients to helicopter emergency medical services: a cost-effectiveness analysis.   JAMA Surg. 2018;153(3):261-268. doi:10.1001/jamasurg.2017.4485PubMedGoogle ScholarCrossref
12.
MacKenzie  EJ, Rivara  FP, Jurkovich  GJ,  et al.  A national evaluation of the effect of trauma-center care on mortality.   N Engl J Med. 2006;354(4):366-378. doi:10.1056/NEJMsa052049PubMedGoogle ScholarCrossref
13.
Delgado  MK, Staudenmayer  KL, Wang  NE,  et al.  Cost-effectiveness of helicopter versus ground emergency medical services for trauma scene transport in the United States.   Ann Emerg Med. 2013;62(4):351-364.e19. doi:10.1016/j.annemergmed.2013.02.025PubMedGoogle ScholarCrossref
14.
Cameron  CM, Purdie  DM, Kliewer  EV, McClure  RJ.  Long-term mortality following trauma: 10 year follow-up in a population-based sample of injured adults.   J Trauma. 2005;59(3):639-646.PubMedGoogle Scholar
15.
Arias  E, Bastian  B, Xu  J, Tejada-Vera  B.  US state life tables, 2018.   Natl Vital Stat Rep. 2021;70(1):1-18.PubMedGoogle Scholar
16.
Gold  MR, Franks  P, McCoy  KI, Fryback  DG.  Toward consistency in cost-utility analyses: using national measures to create condition-specific values.   Med Care. 1998;36(6):778-792. doi:10.1097/00005650-199806000-00002PubMedGoogle ScholarCrossref
17.
Sanders  GD, Neumann  PJ, Basu  A,  et al.  Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: second panel on cost-effectiveness in health and medicine.   JAMA. 2016;316(10):1093-1103. doi:10.1001/jama.2016.12195PubMedGoogle ScholarCrossref
18.
Cameron  CM, Purdie  DM, Kliewer  EV, McClure  RJ.  Ten-year health service use outcomes in a population-based cohort of 21 000 injured adults: the Manitoba Injury Outcome Study.   Bull World Health Organ. 2006;84(10):802-810. doi:10.2471/BLT.06.030833PubMedGoogle ScholarCrossref
19.
US Bureau of Labor Statistics. CPI inflation calculator. Accessed October 10, 2020. https://www.bls.gov/data/inflation_calculator.htm
20.
Eichler  HG, Kong  SX, Gerth  WC, Mavros  P, Jönsson  B.  Use of cost-effectiveness analysis in health-care resource allocation decision-making: how are cost-effectiveness thresholds expected to emerge?   Value Health. 2004;7(5):518-528. doi:10.1111/j.1524-4733.2004.75003.xPubMedGoogle ScholarCrossref
21.
The World Bank. GDP per capita. Accessed November 2, 2020. https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?name_desc=true&locations=US
22.
Gruen  DS, Guyette  FX, Brown  JB,  et al.  Characterization of unexpected survivors following a prehospital plasma randomized trial.   J Trauma Acute Care Surg. 2020;89(5):908-914. doi:10.1097/TA.0000000000002816PubMedGoogle ScholarCrossref
23.
Moore  HB, Moore  EE, Chapman  MP,  et al.  Plasma-first resuscitation to treat haemorrhagic shock during emergency ground transportation in an urban area: a randomised trial.   Lancet. 2018;392(10144):283-291. doi:10.1016/S0140-6736(18)31553-8PubMedGoogle ScholarCrossref
24.
Pusateri  AE, Moore  EE, Moore  HB,  et al.  Association of prehospital plasma transfusion with survival in trauma patients with hemorrhagic shock when transport times are longer than 20 minutes: a post hoc analysis of the PAMPer and COMBAT clinical trials.   JAMA Surg. 2020;155(2):e195085. doi:10.1001/jamasurg.2019.5085PubMedGoogle Scholar
25.
Gruen  DS, Brown  JB, Guyette  FX,  et al; PAMPer Study Group.  Prehospital plasma is associated with distinct biomarker expression following injury.   JCI Insight. 2020;5(8):e135350. doi:10.1172/jci.insight.135350PubMedGoogle Scholar
26.
Gruen  DS, Guyette  FX, Brown  JB,  et al.  Association of prehospital plasma with survival in patients with traumatic brain injury: a secondary analysis of the PAMPer cluster randomized clinical trial.   JAMA Netw Open. 2020;3(10):e2016869. doi:10.1001/jamanetworkopen.2020.16869PubMedGoogle Scholar
27.
Reitz  KM, Moore  HB, Guyette  FX,  et al.  Prehospital plasma in injured patients is associated with survival principally in blunt injury: results from two randomized prehospital plasma trials.   J Trauma Acute Care Surg. 2020;88(1):33-41. doi:10.1097/TA.0000000000002485PubMedGoogle ScholarCrossref
28.
Paisley  S.  Identification of evidence for key parameters in decision-analytic models of cost effectiveness: a description of sources and a recommended minimum search requirement.   Pharmacoeconomics. 2016;34(6):597-608. doi:10.1007/s40273-015-0372-xPubMedGoogle ScholarCrossref
×