Investigation and Validation of the TEG6s During Rotary Wing Aeromedical Flight

Abstract

Introduction:

In order to improve rural and austere trauma care, hospital-based testing performed at the point of injury may shorten the time lapsed from injury to intervention. This study aimed to evaluate the use of the TEG6s^® device in a rotary wing aircraft. Prior attempts suffered from limitation related to lack of vibration mitigation.

Methods:

This was an investigator initiated, industry supported study. Haemonetics^® provided a TEG6s^® analyzer. The device underwent a standard validation. It was secured in place on the aircraft utilizing shipping foam for vibration mitigation. Donors provided 2 tubes of sample blood in one sitting. Paired studies were performed on the aircraft during level flight and in the hospital, using the Global Hemostasis with Lysis Cartridge. Both normal and presumed pathologic samples were tested in separate phases. Paired T-tests were performed.

Results:

For normal donors, mean R (minutes) for laboratory compared to the aircraft was 6.2 vs. 7.2 (p=0.025). Mean CRT MA (mm) was 59.3 and 55.9 ±7.3 (p<0.001) for lab and aircraft (p<0.001). Among normal donors, R was within normal range for 17/18 laboratory and 18/18 aircraft tests (p>0.99).

During the testing of pathologic samples mean R time was 14.8 for lab samples and 12.6 minutes for aircraft (p=0.02). Aircraft samples were classified as abnormal in 78% of samples, this was not significantly different than lab samples (p=0.5).

Conclusions:

The use of the TEG6s^® for inflight viscoelastic testing appears promising. While statistically significant differences are seen in some results, these values are not considered clinically significant. Classifying samples as normal or abnormal demonstrated a higher correlation. Future studies should focus on longer flight times to evaluate for LY30, takeoff and landing effects. Overall, this study suggests that TEG6s^® can be utilized in a prehospital environment, and further study is warranted.

Level of Evidence:

Level II, Diagnostic Tests or Criteria

INTRODUCTION

Hemorrhage after trauma remains a leading cause of mortality, and patients in austere or resource-limited environments tend to have worse outcomes. We hypothesize that to improve trauma care in these settings, traditional hospital-based testing and intervention need to be taken closer to the point of injury. By shortening the time lapsed from injury to intervention, we can save lives in these challenging environments.

Thromboelastography (TEG) is one hospital-based technology that has shown the ability to improve outcomes and decrease resource utilization in treating hemorrhagic shock.^1–5 By individualizing care and treating identified coagulopathies, civilian hospital systems have shown a decrease in blood product use by eliminating unnecessary transfusions. This same experience should occur in a resource-limited environment or during prolonged field care if TEG testing were available in that setting.

Early published lab and simulation-based data suggested the device would be stable and accurate in an aeromedical environment utilizing rotary-wing aircraft.^6–9 However, in practice, two prior studies trialed the device in a helicopter with disappointing results; those studies have significant limitations due to a lack of vibration mitigation.^10,11 Additional in-flight data is needed to assess the accuracy of those studies. This study aimed to investigate and validate the use of the TEG6s^® device during helicopter flight. If found to be accurate, this would create an opportunity for additional research on prehospital TEG use.

METHODS

This was an investigator initiated, industry supported study. Haemonetics provided a TEG6s^® analyzer and laptop with TEG manager through the Haemonetics grant program. The study design was developed by the investigator, and the investigator’s institution paid for disposable costs. The study was designed as a pilot study, focusing on device investigation and validation. Patient care was not affected at this stage. The TEG6s^® device is not currently approved for use inflight. See Figure 1 for a depiction of the study plan. This study was completed with institutional IRB approval (protocol #2104281553), and complies with the Equator guidelines for STARD 2015: An Updated List of Essential Items for Reporting Diagnostic Accuracy Studies (Supplemental Digital Content 1).¹²

Figure 1.

Study design for TEG6s validation during aeromedical flights

After delivery of the study device, it underwent a standard lab validation overseen by the manufacturer. Next, the device was placed onto an AirBus EC-135 for inflight testing. The device was secured between the rear crew seats with shipping foam utilized by Haemonetics, for padding and vibration reduction (Figure 2). Shipping foam was grade 29–400, cut and laminated to form a shell conforming to the analyzer exterior profile, with additional openings made to allow cartridge insertion, power, and ventilation. The foam fits within dimensions 314 × 236 × 385mm. The device underwent evaluation for electromagnetic compatibility of portable electronic devices as required by the Federal Aviation Authority (FAA) for electrical equipment carried in the aircraft not installed with FAA-approved data (Supplemental Digital Content 2).¹³ This testing was performed to ensure no interference occurred between the TEG6s^® device and the operations of the aircraft. Once this was completed, the device was cleared for use on this individual aircraft only.

Figure 2.

TEG6s device secured on aircraft

Flight crew were trained on device and cartridge use. For inflight testing, sample analysis started during level flight. Once in level flight, the crew utilized a calibrated auto-pipette to transfer 0.5 milliliters of blood from an appropriately labeled and filled blue top (citrate) test tube to the TEG6s^® Global Hemostasis with Lysis Cartridge. Of note, an uninterruptable power supply was added during phase 1 of samples due to brief interruptions in the aircraft power inverter, causing a sample to be interrupted. Flight crews kept a log of results and any serious inflight events.

Inflight validation

To perform the inflight validation, simultaneous paired samples from a single donor were analyzed: one sample on the inflight study device and a second sample in the hospital clinical lab. The first phase of this validation protocol involved samples of healthy donor blood. Volunteers were consented by the study coordinator or primary investigator. Subjects were excluded if they were prescribed any blood thinners, antiplatelet medications, and NSAIDs within 12 hours of blood draw. After consenting, a red top tube was drawn as a waste, followed by two citrate tubes.

The second phase of this study involved blood from donors who were expected to have pathologic findings on TEG. These patients were identified and consented within the intensive care unit at our facility and were on anticoagulant drips or extracorporeal membrane support (ECMO).

Statistical analysis

Descriptive statistical analyses were used to summarize the data including proportions, means and standard deviation (±SD), and scatter-plots. A power analysis was performed to demonstrate that 18 paired samples would provide adequate power to ensure the device is accurate while inflight compared to a tabletop control. This sample size will achieve an 80% power to detect 0.7 standard deviations using a two-sided paired t-test at a significance level of 0.05. A paired t-test was utilized to assess the continuous outcomes such as Mean R (minutes) time and Mean CRT MA for paired data from aircraft and labs. Fisher exact test was utilized to compare categorical values. Bland-Altman plots were generated to demonstrate the robustness of the analysis. During phase two of the study, subjects that had normal coagulation on hospital TEG testing crossed over into the normal donor group for analysis.

RESULTS

Phase 1 – healthy donors

In phase one of our analysis, results were available for 18 samples. Subjects were evenly divided between male and female, and the mean age was 33. Mean R (minutes) time for lab vs aircraft was 6.2 ±1.6 compared to 7.2 ±1.6 (p=0.025) (Figure 3). Mean CRT MA (mm) was 59.3 ±5.6 and 55.9 ±7.3 and (p<0.001) for lab and aircraft, respectively. The average level inflight time for the helicopter utilized was 20 minutes, this limited the ability to obtain Ly30 results. CFF MA (mm) results were routinely <4 on the aircraft, while normal range results were obtained from lab samples.

Figure 3.

Comparison of results in healthy donors

When R-time results are classified as normal vs abnormal, 18/18 lab tests were labeled as normal, and 17/18 aircraft tests were labeled normal (p=1). CRT MA results demonstrated 17/18 normal response in lab results, and 13/18 results from aircraft samples (p=0.17). CFF and Ly30 were not further analyzed.

Phase 2 – pathologic samples

Presumed pathologic samples were collected from 18 patients: three patients on coumadin, nine patients on a heparin drip, five patients on a bivalirudin drip, and one on therapeutic enoxaparin. Six patients were female, and the mean age was 54 years. Bivalirudin patients were receiving ECMO therapy.

Nine samples demonstrated pathology in lab-based analysis. Amongst these samples, the mean R time was 14.8 ±2.2 for lab samples and 12.6 ±3.6 minutes for aircraft (p=0.02) (Figure 4). Aircraft samples were classified as abnormal in 7 of 9 (78%) samples, this was not significantly different than lab samples (p=0.5).

Figure 4.

Comparison of results for pathologic samples

In this phase, nine patients on a heparin drip crossed over into the healthy donor analysis due to normal R-times on lab-based testing. When these were analyzed with the subjects from phase one, the mean R time for lab and aircraft was 6.6 ±1.6 compared to 7.8 ±1.9 (p=0.001). When classified as normal vs abnormal, 27/27 lab tests were normal, and 23/27 aircraft tests were abnormal (p=.11).

Agreement analysis

Bland-Altman plots were generated from all samples for evaluation of R time (Figure 5), and for CRF MA (Figure 6). For R time, the 95% confidence interval for limits of agreement were −5.065 to 4.365. For CRF MA the 95% confidence intervals for limits of agreement were −3.89 to 16.34.

Figure 5.

Bland-Altman plot for R time measurement agreement analysis. Dashed lines represent 95% limits of agreement.

Figure 6.

Bland-Altman plot for CRT MA measurement agreement analysis. Dashed lines represent 95% limits of agreement.

Technical and device errors

Five additional samples were collected but did not result on the helicopter. In two cases, this was due to power interruptions, two due to screen malfunctions, and one due to user error in loading adequate sample volume.

DISCUSSION

The TEG6s^® device shows potential for inflight use on a rotary wing aircraft, and warrants continued investigation. While statistically significant differences are seen in R time and CRT, these values were deemed not clinically significant. Furthermore, in this experiment, the results would not have led to unnecessary intervention. Utilizing a classification system of normal and abnormal demonstrated a high correlation between both samples. Additional study is needed on pathologic blood samples to increase the power of this analysis and to demonstrate accuracy for other abnormalities seen on TEG testing. Further research into vibration mitigation strategies may also increase the accuracy of the device inflight.

Patients in rural and austere environments have worse outcomes after severe injury. Distance, and hence time, to definitive care are major drivers in this disparity.^14–16 In order to improve outcomes, damage control principles must start closer to the point of injury in resource-limited environments. This has been demonstrated by the military utilizing specialized units to bring additional therapies to the point of injury.¹⁷ These lessons have carried into civilian emergency medical services (EMS) care, with an increase in agencies utilizing blood products, an expansion of the products available for transfusion, and an increase in antifibrinolytics use to begin early resuscitation.^18–21 These strategies have improved patient outcomes and mortality.^22–25

The use of prehospital TEG is the next evolution in care for hemorrhagic shock. Military and civilian guidelines already recommended utilizing viscoelastic testing during damage control resuscitation.^26–28 Deploying this advanced testing into the prehospital space will benefit the patients and healthcare systems. In the future, EMS could begin guided resuscitation, utilizing limited products such as plasma or whole blood only when coagulopathy is present. The routine administration of tranexamic acid could also be tailored only to patients who demonstrated fibrinolysis and avoided in patients with fibrinolytic shutdown. While it is known that hypofibrinogenemia is commonly seen early in trauma, we were unable to demonstrate this in flight. Moreover, at least 30 minutes is required in order to obtain Ly30 using the TEG6s^® Global Hemostasis with Lysis Cartridge. Other devices may be able to evaluate fibrinolysis more quickly, but they require in flight validation and testing.²⁹

Another important factor to consider is the ability to provide useful data to a trauma center that receives patients from large rural catchment areas. Level I or II trauma centers that serve such regions often receive patients several hours after the injury has occurred. During this extended period between the injury and receiving medical attention, the patient’s blood’s ability to clot can be impaired, leading to detectable changes in TEG. Even if prehospital providers don’t have all the resources to transfuse patients based on TEG results, having that data 30–45 minutes earlier at the definitive trauma center would allow the immediate initiation of goal-directed therapy.

This study did have some limitations. The investigators attempted to synchronize the start of each sample, but some variation was inherent to the nature of the separated devices. The exact effect this may have had on results is unknown. The need for more data on Ly30 will have to be addressed in future studies with longer flight times. With regards to the agreement analysis, unfortunately, acceptable limits of agreement were not established a priori; as such, evaluation of the Bland-Altman plots must be made with clinical judgment. The pathological phase was also unexpectedly underpowered, which limits the interpretation of our results on pathological samples.

It should be reiterated that this device is not currently designed for use in the aeromedical environment, and the effect of long-term use on accuracy and device failure is still unknown. And based on these data this device could only be considered for the AirBus EC-135; in order to utilize the device on other airframes, future work will require investigating the limits of vibration that the TEG6s can operate within. Future studies will also be needed to investigate the effect of take-off and landing on sample analysis.

Limitations

An unexpected weakness of this study was the CFF MA result. This portion of the TEG tracing routinely did not demonstrate fibrinogen contribution to clotting on the aircraft, while normal in the laboratory. The authors hypothesize there could be a vibration interference but cannot confirm that as a cause. Future studies will be needed to determine the cause of this deficiency and find a mitigation strategy. However, cryoprecipitate is currently rarely available in prehospital settings, so this concern does not currently limit patient care possibilities.

CONCLUSION

Overall, this study suggests that the TEG6s^® can be utilized in a prehospital and rotary wing transport environment. Additional study is needed to confirm these results and further evaluate the degree of accuracy possible so as to avoid unindicated transfusion. This has the potential to improve prolonged or forward field care and make an impact on rural trauma care. Utilizing TEG-guided resuscitation could improve outcomes from hemorrhagic shock and improve resource utilization.

Supplementary Material

Supplemental Digital Content 1

SDC 1. STARD 2015 Checklist: An Updated List of Essential Items for Reporting Diagnostic Accuracy Studies.

Supplemental Figure 1

SDC 2. Federal Aviation Authority (FAA) form for electrical equipment carried in the aircraft not installed with FAA-approved data