Performance comparison of intraosseous devices and setups for infusion of whole blood in a cadaveric swine bone model

a b s t r a c t

Objectives: Intraosseous (IO) access can provide a critical bridge for blood product infusion when peripheral ve- nous access is not obtainable. Successful pressurized IO infusion requires flow rates sufficient to preserve life, but with infusion pressures low enough to avoid clinical complications (e.g., hemolysis, bone damage, fat emboli). However, the optimal method for pressured IO delivery of blood was unknown.

Methods: Three trained physicians infused 500 mL of whole blood through a 15-gauge, 45 mm IO catheter into fresh, high bone density cadaveric swine proximal humeri. Participants applied eight different pressure infusion strategies: (1) gravity, (2) pressure bag, (3) pressure bag actively maintained at or above 300 mmHg, (4) hand pump, (5) hand pump with pressure bag, (6) push-pull with 10 mL syringe, (7) push-pull with 60 mL syringe, and a (8) Manual Rapid Infuser in a randomized within-subjects design (30 trials per method, 240 trials total). The primary outcomes of flow rates, mean and Peak pressures, and user ratings were contrasted using ANOVA at p < 0.05.

Results: The Manual Rapid Infuser conferred the highest flow rates (199 +- 3 mL/min) and most favorable user ratings, but also the highest mean and peak pressures. Push-pull conferred the next highest flow rates (67 +- 5 mL/min for 60 mL, 56 +- 2 mL/min for 10 mL) and pressures, with intermediate-to-high user ratings. Hand pump flow rates were essentially identical with (45 +- 4 mL/min) or without (44 +- 3 mL/min) pressure bag, with high user ratings without a pressure bag. Pressure bag and gravity methods conferred low flow rates and user ratings.

Conclusions: Some pressured IO infusion methods can achieve flow rates adequate to serve as a resuscitative bridge in the massively hemorrhaged trauma victim, but flow rates and pressures vary greatly across IO pressur- ized infusion methods. Manual Rapid Infuser and push-pull methods conferred high flow rates but also relatively high pressures, highlighting the importance of using in vivo models in future research to assess the possible clin- ical complications of using these promising methods. Combined, present findings highlight the importance of studying pressurized IO methods towards preserving the life of the critically injured trauma victim.

Published by Elsevier Inc.

  1. Introduction

massive hemorrhage is one of the leading causes of potentially pre- ventable traumatic death after injury [1,2]. Early, prehospital blood product transfusion is associated with decreased mortality rates in

* Corresponding author at: Department of Emergency Medicine, Naval Medical Center San Diego, 34,800 Bob Wilson Drive, San Diego, CA 92134.

E-mail address: [email protected] (J.D. Auten).

critically injured patients at risk for hemorrhagic shock [3,4]. The effec- tive resuscitation of the critically injured requires rapid, reliable vascular access that can replace blood lost during hemorrhage. Intraosseous (IO) catheters can provide a critical bridge for blood prod- uct infusion when peripheral venous access is not obtainable [5,6]. However, despite the extensive role IO catheters have played in prehos- pital and trauma resuscitations, the optimal method to infuse blood products intraosseously remains poorly described [5,7]. IO blood prod- uct transfusion in the adult trauma patient requires an infusion pressure

https://doi.org/10.1016/j.ajem.2022.01.039 0735-6757/Published by Elsevier Inc.

above gravity to achieve adequate flow rates and overcome the higher bone density and blood viscosity in accordance to Darcy’s Law [8-11].

IO flow rates must be sufficiently rapid enough to preserve life, but in- fused at pressures low enough to avoid potential clinical complications, such as intravascular hemolysis, bone damage, or arterial fat emboli [8,10,12]. Prior research demonstrates that blood product IO infusions at pressures greater than 4000 mmHg cause intravascular hemolysis in live tissue swine models with bone densities similar to adult humans [10]. However, the optimal method for pressurized IO infusion is unclear. Current pressurized IO methods include gravity, pressure bags, in-line hand pumps, push-pull syringe infusion using a 3-way stopcock, and Manual Rapid Infusers [7,10,11,13]. Pressure bag augmented infusion is presently recommended because gravity does not provide Clinically meaningful flow rates in the adult population, but the pressure bag method only provides modest flow rates [7,10,11]. Importantly, no pub- lished studies to date were specifically designed to compare currently available pressurized IO methods in flow rates and pressures.

To fill this important gap in the literature, the present randomized prospective study was specifically designed to assess eight pressurized IO infusion methods in set-up time, flow rates, mean and peak pres- sures, and user ratings, using a translational swine Proximal humerus model with bone densities approximating adult humans. Accordingly, our null hypotheses were that the eight pressure methods would not significantly differ in set up times, flow rates, mean pressures, peak pressures, or end user ratings.

  1. Methods
    1. Study setting

We conducted this prospective, randomized study in a translational research laboratory at the Naval Medical Center San Diego (NMCSD). The NMCSD Institutional Review Board (NMCSD.2019.0010) approved this study according to Federal Policy for the Protection of Human Sub- jects (the Common Rule) and the revised Common Rule (HHS 45 CFR 46.102 and DoD 32 CFR 219.101, effective January 21, 2019). The protocol was also reviewed by the NMCSD Institutional Animal Care and Use Com- mittee and was not categorized as live animal research. All activities were conducted in compliance with the Department of Defense regulations.

    1. Model selection

We utilized recently euthanized swine proximal humeri, similar to prior swine cadaveric studies on IO catheter placement, in alignment with the Refine, Reduce, Reuse principle of animal research [14,15]. Based on prior research, cadaveric swine (Sus scrofa) weighing 70-90 kg were selected for this model because their bone density (>1 g/cm2) approximates the average 20- to 40-year-old male trauma pa- tient [10,11,17-19]. Proximal humerus samples were chosen because they demonstrate more rapid transfusion rates than the tibia or femur [20,21]. The specimens were provided by a third-party vendor (Sierra Medical, Whittier, CA) that specializes in biologic tissue procurement for the research and development industry. The proximal humeri were chilled after euthanasia and any fascia or muscle remnants were carefully removed prior to the study to prevent unnecessary obstruction of the emissary vessels on the surface of the bone. Bones with damaged cortex from harvesting after euthanasia were excluded. An EZ-IO(R) powered driver (Arrow(R) EZ-IO(R), Teleflex Medical, Co., Westmeath, Ireland) was used to insert a 15-gauge, 45 mm, IO needle into the greater tubercle of the proximal humerus at a 45-degree angle. Success- ful placement was confirmed by flushing 10 mL of crystalloid fluid and observing adequate flow out of at least three emissary vessels on the surface of the bone. Samples were excluded in they failed to demon- strate fluid flowing out of at least three emissary vessels. Cold-stored heparinized bovine (Bos taurus) whole blood was used as the infusion fluid as it has been previously shown to be closer in viscosity to

human blood than porcine blood [9,22]. Mean and peak pressure data were collected using a calibrated in-line digital pressure gauge (Ash- croft Inc., Stratford, CT) attached 3 in. proximal to the IO insertion site, similar to prior studies published by our study group [10,11,14,19,20].

    1. Study participants

Infusions were performed by U.S. Navy emergency medicine physi- cians (N = 3; WH, KL, MM). All participants had prior experience with placing IO needles, including the male staff physician (25 IO needles on Live patients), the male resident physician (15 IO needles on a com- bination of live patients and cadavers), and the female resident physi- cian (10 IO needle placements on Training models and cadavers). Each participant had prior experience with all infusion methods utilized in the study, except for the Manual Rapid Infuser (Manual RI). An educa- tional session on optimal device use for the Manual RI was performed by an expert user, after participants had completed half of the trials with the device. This methodological step was taken to determine the importance of device education on optimal performance with novel de- vices. The educational intervention included formal device in-service and individual feedback on observed use of the device. No other method included formal instruction.

    1. Infusion methods

Eight IO infusion methods and setups were chosen for this study be- cause they were previously described in the clinical medical literature [10,11,14,23]:

      1. Method 1 – gravity

A blood-filled donor bag was attached to an 80-in. Y-type blood tub- ing set with 200-um blood filter, pre-pierced Y-site, and secure lock (ICU Medical, Inc., Lake Forest, IL). The bag was hung approximately three feet above the surgical table [11,23].

      1. Method 2 and 3 – pressure bag

A blood-filled donor bag was attached to the 80-in. Y-type blood tubing set and placed inside a 1000 mL Unifusor(R) Pressure Infuser bag with piston gauge and stopcock valve (Statcorp Medical, Jackson- ville, FL). Pressure bags were initially inflated to 300 mmHg, and were either maintained at or above 300 mmHg via semi-continuous pumping by participants (PB300) or allowed to deflate naturally over time (PB), simulating a scenario wherein the user may be unable to monitor the pressure bag [10,11,23].

      1. Method 4 and 5 – in-line bulb hand pump

A blood-filled donor bag was attached to an 80-in. bulb pump blood tubing set with 200-um blood filter, pre-pierced Y-site, and a secure lock (ICU Medical, Inc., Lake Forest, IL). Investigators infused the blood by squeezing the 2-in. diameter in-line bulb pump with single-handed or double-handed palm compressions (Fig. 1). The hand pump was tested with a 1000 mL Unifusor(R) Pressure Infuser bag, which was set at or above 300 mmHg (HP300). The hand pump was also tested without the pressure bag assistance (HP) [23].

      1. Method 6 and 7 – push-pull

The push-pull (PP) method was tested with either a 10 mL (PP10) or 60 mL (PP60) BD Luer-lok(R) Tip syringe (Becton, Dickerson, and Co., Franklin Lakes, NJ). The PP10 and PP60 methods utilize a three-way high flow stopcock with rotating luer (ICU Medical, San Clemente, CA) that is toggled to the 90-degree setting by the user to fill the syringe when the plunger is pulled back, then toggled by the user to the in- line setting so that blood is driven into the bone when the plunger is de- pressed. A blood-filled donor bag was attached to the 80-in. Y-type blood tubing, with the three-way stopcock placed at the end of the blood tubing, 3 in. proximal to the digital gauge [10,23,24].

Image of Fig. 1

Fig. 1. In Line 80 in. Hand Bulb pump blood tubing set with 200-um blood filter, pre- pierced Y-site, and a secure lock (ICU Medical, Inc., Lake Forest, IL).

      1. Method 8 – manual rapid infuser

The LifeFlow(R) Rapid Infuser (410 Medical, Durham, NC) is a Manual RI that infuses crystalloids and blood through device-specific tubing which utilizes an in-line 10 mL syringe (Fig. 2). This Manual RI pulls fluid from the donor bag into the in-line 10 mL syringe when the handle is opened and ejects fluid forward into the bone when the handle is squeezed like a trigger. Manual RI training included a onetime tutorial to correctly prime and set-up the unique tubing within the device after half of the study trials were performed [14].

    1. Study protocol

A study coordinator informed the participants of their randomly assigned infusion method just prior to each trial (Fig. 3). Set-up time was recorded by a clinical research coordinator (ER) using a stopwatch, from when the participant first touched the infusion materials until all

components were completely connected and tubing was primed with blood. For pressure bag methods, set-up times also included inflation of the pressure bag. Infusion time began after set-up was complete and continued until the blood-filled donor bag was empty or until 15 min elapsed. A new swine proximal humerus was used for each trial to ac- count for possible bone marrow changes following higher pressure infu- sions. No swine proximal humerus samples were reused during this study. Outcomes included set-up time (seconds), flow rate (mL/min), mean pressure (mmHg), peak pressure (mmHg), and user ratings. Flow rate was calculated as the total volume infused divided by the infu- sion time. Mean pressures were calculated by recording and averaging the pressure gauge values every second for the first 60 s of the infusion. Peak pressure was defined as the highest pressure observed in the first 60 s of the infusion. Participants completed the user questionnaires for each trial at the end of the study. The following questions were scored from 1 to 5, corresponding to very unsatisfied to very satisfied:

  1. The device set up and hand fatigue were minimal. (Set-up and Fatigue)
  2. I would feel comfortable using this technique for blood transfu- sion in a massively hemorrhaged warfighter as a bridge to defin- itive central venous access. (Bridge to CVC)
  3. The blood product flow rates I observed with this technique can meet the demands of Damage control resuscitation. (Damage Control)
  4. I would feel comfortable using this method to resuscitate a criti- cally injured warfighter in a space confined setting like a rotary wing aviation asset or ground transport vehicle. (confined space)
  5. I feel comfortable that this technique could be used effectively in the hands of a corpsman/medic providing care close to the point of injury. (Medic)
  6. This method failed or malfunctioned during the intraosseous transfusion effort. (Malfunction)
  7. I trust this method to effectively initiate damage control resuscitation in the first 10-15 min after arrival or injury. (15 min)
    1. Data analysis

This study employed a prospective, randomized design to contrast eight infusion methods in set-up times, flow rates, mean infusion

Image of Fig. 2

Fig. 2. (A) LifeFlow(R) Manual Rapid Infuser, (B) AirCheck(R) including ball valve in drip chamber, (C) in-line 10 mL syringe, and (D) Force Reducer(R).

Image of Fig. 3

Fig. 3. Schematic diagram of the pressure infusion component arrangements (not to drawn scale).

pressures, peak infusion pressures, and user ratings. Tests of power using G*Power software (version 3.1, Faul et al., 2009) found that, assuming a 95% confidence interval and an effect size of f = 0.50 for omnibus testing, statistically significant differences would be real- ized on 80% of opportunities (power = 0.80) with as few as 72 trials total, with 14 trials per condition for pairwise testing. However, a similar study evaluating similar intravenous catheter infusion methods incorporated 30 independent trials for six different infusion methods [23]. Therefore, to ensure adequate power, this study was designed to have each of three participants complete ten trials for each of the eight infusion methods (30 trials per infusion method, N = 240 total infusion trials).

While 30 trials were planned for analysis of all eight methods, we noted significant differences between the use of Manual RI with and without formal educational training. The early Manual RI trials were beset by excessive infusion pressures, failure to disengage the ball valve in the drip chamber and handle breakage because participants were squeezing the handle too rapidly and placing excessive pressure on the force reducer component of the IV tubing (Fig. 2). Data prior to the educational training were therefore removed, so the Manual RI sam- ple size for formal analysis was n = 15. The pre- and post- educational training pressure data for Manual RI are provided for comparison (Appendix 3).

Data were analyzed using analysis of variance (ANOVA), with pairwise comparisons to localize statistically significant differences between pressure infusion methods. Each result was confirmed using a non-parametric equivalent, Kruskal-Wallis for omnibus test- ing and Mann-Whitney U for pairwise testing. Differences were con- sidered statistically significant at the p < 0.05 threshold. Results are expressed as mean values +- standard error of the mean. All analyses were conducted in SPSS statistical software (version 23, IBM Corp., Chicago, Illinois).

  1. Results

The three participants (WH, KL, MM) did not significantly vary in set-up times, flow rates, mean pressures, or peak pressures, so their data were combined for statistical analysis.

    1. Set-up times

Set-up times were fastest for gravity and hand pump methods, aver- aging less than 1 min (Appendix 1). Push-pull and Manual RI set-up times averaged between 1 and 1.5 min. Pressure bag and hand pump plus pressure bag methods set-up times averaged roughly 2 min.

    1. Flow rates

Manual RI flow rates averaged 199 +- 14 mL/min, roughly three times the flow rate of both push-pull methods. This is equivalent to transfusing a 500 mL bag of blood in 2.5 min, compared to 7.5 min for PP60 and 9 min for PP10 (Fig. 4). The HP and HP300 demonstrated in- termediate flow rates, equivalent to transfusing a 500 mL bag of blood in 11 min. The pressure bag, whether left unattended (PB) or main- tained at 300 mmHg (PB300), could transfuse a 500 mL bag in 30 min. Infusing 500 mL blood with gravity alone would take over 2 h.

    1. Mean infusion pressures

Mean infusion pressures largely mirrored the order of the flow rate results (Fig. 5). Manual RI mean infusion pressures were roughly double the mean infusion pressures of the push-pull syringe methods. The ad- dition of a pressure bag to the hand pump tubing system increased mean infusion pressures by 50% compared to hand pump alone. The gravity method had the lowest mean infusion pressures, less than one-third the mean infusion pressures of the pressure bag methods.

    1. Peak infusion pressures

Manual RI peak pressures were 20% higher than PP10 and twice the pressure of PP60 (Fig. 6). PP10 was over 1000 mmHg higher than PP60. Peak pressures for HP were intermediate in comparison to Manual RI, PP10 and PP60. HP300 averaged 300 mmHg higher than HP alone. The gravity and pressure bag methods had the lowest peak infusion pres- sures, less than one-third those observed with other methods. Two of the 30 PP10 trials (7%) and two of the 15 Manual RI trials (13%) exceeded the theoretical pressure threshold of 4000 mmHg.

Manual RI Push-Pull 60

Push-Pull 10 HandPump+PB HandPump

PB PB300














0 50 100 150 200 250


Manual RI Push-Pull 10

Push-Pull 60 HandPump+PB HandPump

PB PB300













68 +


0 500 1000 1500 2000 2500 3000 3500 4000


Fig. 4. Flow rates (mL/min) of each infusion strategy.

*p < 0.05 versus next higher flow rate.

+p < 0.001 versus next higher flow rate.

    1. User ratings

Appendix 3 displays the user rating values and rankings, with poorer rankings indicated by darker shades. The Manual RI received the most favorable ratings overall, and the highest ratings in 6 of 7 categories. HP had the second highest ratings composite, followed by PP60 and PP10, in spite of poor ratings for set-up & hand fatigue and lack of fail- ures or malfunction. HP300 scored lowest for set-up & hand fatigue and PB scored lowest for utility in confined space. Gravity scored favor- ably for easy set-up & hand fatigue and lack of failures or malfunctions due to the simplicity of its system, but scored low or lowest in all other categories. PB300 scored lowest overall.

  1. Discussion

Our study is unique in contrasting eight previously described IO pressure infusion methods in set-up time, flow rate, mean and peak pressure, and end user ratings. This line of investigation provides a foun- dation for addressing clinical complications in future live tissue models, all towards determining the ideal pressurized IO method to preserve life in the prehospital phase of damage control resuscitation. The optimal IO infusion method will ideally achieve high flow rates at safe infusion pressures while being effectively performed near the point of injury [27]. Our threshold for safe infusion pressures was defined in prior research, which demonstrated that pressures above 4000 mmHg was associated with clinical complications like intravascular hemolysis [8,10]. However, our study’s 4000 mmHg threshold is theoretical since prior research did not quantify how high above 4000 mmHg hemolysis occurred [10].

Manual RI Push-Pull 10








63 +




Push-Pull 60 HandPump+PB HandPump

PB300 PB


0 500 1000 1500 2000


Fig. 5. Mean pressures (mmHg) for each infusion strategy.

*p < 0.05 versus next higher flow rate.

**p < 0.01 versus next higher flow rate.

+p < 0.001 versus next higher flow rate.

Fig. 6. Peak infusion pressures (mmHg) of each infusion strategy.

*p < 0.05 versus next higher flow rate.

**p < 0.01 versus next higher flow rate.

+p < 0.001 versus next higher flow rate.

In the present study, Manual RI infusion rates were significantly faster than all other infusion methods with potentially lifesaving flow rates (199 ml/min), similar to transfusing whole blood through a 16- gauge peripheral intravenous catheter [25,26]. The Manual RI received the most favorable rankings for easy set-up, minimal hand fatigue, ease of use in confined settings, and confidence in its ability to meet the demands of damage control resuscitation by Prehospital providers. The Manual RI requires only one hand allowing the user to multitask which is essential if there are limited personnel at the point of injury.

Formal training before using the novel Manual RI device is an impor- tant issue that had not been previously reported [14]. Our participants initially primed the tubing incorrectly, resulting in infusion pressures exceeding 5000 mmHg. Participants also did not allow for adequate re- coil within the Force Reducing chamber resulting in two incidences of participants breaking the trigger handle by applying excessive grip strength. However, following formal Manual RI user training, our partic- ipants produced high flow rates without device malfunction. They also rated the Manual RI highly during exit questionnaires, including ratings for avoiding device failure or malfunction.

Considering safety, flow rate, simplicity, and device availability, our study favors the use of PP60 to achieve lifesaving flow rates with a po- tentially safer pressure profile as previous described by both British and Israeli prehospital providers [22,32]. Consistent with prior research, both push-pull methods were significantly faster than gravity or pres- sure bag methods [10,14]. However, PP10 demonstrated lower flow rates and higher peak pressures than PP60, with 2 trials exceeding 4000 mmHg. PP60 had no instances of peak pressures exceeding the 4000 mmHg threshold. Although the push-pull methods require both hands for operation, the materials are readily available in most prehos- pital settings and required no formal user training. Users reported mar- ginally lower hand fatigue using push-pull with a 10 mL syringe compared to the 60 mL syringe, which was consistent with prior re- search demonstrating greater hand fatigue with increasing syringe size [31].

Our study is unique from prior research in that the hand pump (HP) infusion method had not been previously evaluated in comparison to push-pull, pressure bag, or Manual RI methods [10,11,13,14,19]. The hand pump had intermediate flow rates with potentially safer infusion pressures than other methods. The HP and HP300 flow rates were supe- rior to the pressure bag method which is considered the safest IO infu- sion method in current research [10,19]. We did not find a benefit in adding a pressure bag to the hand pump systems, as the added pressure bag increased infusion pressures without Noticeably increasing flow rates.

PB and PB300 showED flow rates and infusion pressures greater than gravity alone. PB and PB300 flow rates were essentially identical to each other. The flow rates seen with PB and PB300 were noted to

be significantly lower than those observed in prior in vivo models. These differences are likely related due to a combination of factors including IO catheter tip positioning within the medullary or cancellous bone, a reduced number of patent intraosseous emissary veins and the de- creased plasticity of our cadaveric bone model when compared to living tissue [33]. The gravity IO infusion method cannot be recommended for resuscitation of patients with hemorrhagic shock because the flow rates are futile. Present findings are consistent with the current body of evi- dence in the medical literature suggesting that some pressure above gravity is required for IO blood transfusions [28,29]. The gravity method also received the lowest participant ratings overall (Appendix 3).

Prehospital providers may have limited personnel and support, so ease of use and comfort while implementing a manual IO pressure method is important to maximize the capabilities of smaller medical teams. User ratings indicate that Manual RI and HP were easy to use in confined spaces and could be operated with one hand, which is espe- cially important for limited personnel required to multitask on-scene and during aeromedical transport [27,30]. While participants provided highest rating for Manual RI, HP, and PP60 methods for comfort in use, evaluating clinical complications was not possible in the present in vitro study. Future research utilizing in vivo models are needed to contrast Manual RI, HP, PP60, and PB to better understand which method best balances flow rates, pressures, and usability in many situ- ations that require emergency IO transfusion.

    1. Limitations

The primary limitation of our study is that swine anatomy may not directly translate to the adult trauma population. However, porcine bone was utilized because it has been shown to be an excellent transla- tional model for transfusion research [34,35]. Although this model does not generate values that precisely correspond to live humans, this model is able to directly contrast various IO pressurized infusion methods in relative differences. Identifying these differences provides a foundation for future studies utilizing a live animal model. We ad- hered to the “Three R’s” (Refine, Reduce, Reuse) of animal research by choosing to reduce live animal subjects by the reuse of cadaveric tissue in a model similar to previous intraosseous research [15,16]. The post- mortem collection of our study samples may have led to slower flow rates secondary to the possibility of some intraosseous emissary vessels becoming occluded after procurement. This likely explains why our push-pull and pressure bag flow rates were slower than previously de- scribed live swine models utilizing whole blood [10,11,13,14]. However, our Pressure measurements were consistent with infusion parameters from these prior in vivo IO studies [10,11,13,14,19].

Our findings are limited to the use of a single site infusion of cold stored whole blood in a cadaveric swine humerus. Different anatomic sites, like the tibia or sternum, dual anatomic sites, or different fluid vis- cosities would likely confer different results [19,21]. Additionally, the mean and peak pressures were measured using an in-line digital pres- sure gauge and were therefore not true intermedullary pressures. While measuring intermedullary pressures would have been ideal, ex- ternal pressure readings achieved the purpose of our study was to allow for relative comparisons between infusion methods. This study did not include assessment of intravascular hemolysis or pulmonary histology to measure possible damage created by shear stresses from the pressures generated from various IO infusion devices and setups [7,36,37]. Future research using in vivo models should be conducted to assess clinical complications created by pressurized IO infusion, such as intravascular hemolysis, trabecular bony injury, or occlusive pulmo- nary arterial Fat embolism. Finally, the correlation between flow rates, mean pressures and peak pressures averaged near zero within six of the eight individual infusion methods. This limitation suggests that the IO catheter tip location within the medullary or cortical bone, may play a confounding role in overall performance within individual infu- sion methods. Future research on this topic should assess the impact

IO catheter tip placement in relation to the central medullary venous sinus has on infusion flow rates and pressures.

  1. Conclusion

The Manual Rapid Infuser method had the highest IO flow rates and most favorable user ratings in this in vitro study, but also showed high mean and peak pressures. Further, proper Training and practice are needed to avoid Manual RI device malfunctions. The Push-pull method with a 60 mL syringe conferred higher flow rates at lower peak pres- sures than push-pull with a 10 mL syringe. Intermediate performance was found for hand pump techniques, which provided three times the flow rates of the currently recommended pressure bag method. The gravity method of IO infusion produced flow rates that are futile for trauma resuscitation. Future research using in vivo models should as- sess clinical complications, such as intravascular hemolysis, bone dam- age, and arterial fat embolism that may occur from the sheer stresses created by pressurized IO infusion methods. The present study adds to a growing body of literature towards defining the optimal IO method that can balance flow rates that are rapid enough to preserve life, while minimizing short- and long-term clinical complications.

Author contributions

J.D.A., V.S.B., G.J.Z. designed this study. All authors contributed to the literature search. J.D.A., K.J.L. and E.R.R. managed institutional board submission. M.M.M., W.C.H., K.J.L., W.D.B., J.J.L, J.D.A., and E.R.R. collected data. J.D.A., K.J.L., A.D.E, G.J.Z., and V.S.B. performed Data interpretation. All authors contributed to manuscript preparation and revision.


Joint Services Symposium on Emergency Medicine, September 2020.

Military Health System Research Symposium, August 2020.


The views expressed in this article reflect the results of research con- ducted by the author and do not necessarily reflect the official policy or position of the Department of the Air Force, Department of the Navy, Department of Defense, nor the United States Government. I am a mili- tary service member of the United States government. This work was prepared as part of my official duties. Title 17 U.S.C. 105 provides that “copyright protection under this title is not available for any work of the United States Government.” Title 17 U.S.C. 101 defines a U.S. Gov- ernment work as work prepared by a military service member or em- ployee of the U.S. Government as part of that person’s official duties.


This work was supported by The Office of the Assistant Secretary of Defense for Health Affairs, endorsed by the Department of Defense, through the Congressional Directed Medical Research Programs Joint Program Committee-6, Award Number – W81XWH-20-2-0033.

Credit authorship contribution statement Katherine J. Lee: Writing – review & editing, Writing – original draft,

Project administration, Methodology, Investigation, Formal analysis,

Data curation, Conceptualization. Morgan M. McGuire: Writing – review & editing, Writing – original draft, Investigation, Data curation. Warren C. Harvey: Writing – review & editing, Writing – original draft, Investigation, Data curation. William D. Bianchi: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptual- ization. Alec D. Emerling: Writing – review & editing, Writing – original

draft, Resources, Project administration, Formal analysis, Data curation. Erin R. Reilly: Writing – review & editing, Writing – original draft, Soft- ware, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation. Vikhyat S. Bebarta: Writing – review & editing, Writing – original draft, Supervision, Methodology, Funding ac- quisition, Formal analysis, Conceptualization. Jason J. Lopez: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation. Gregory J. Zarow: Writing – review & editing, Writing – original draft, Validation, Software, Resources, Methodology, Formal analysis, Conceptualization. Jonathan D. Auten: Writing – review & editing, Writing – original draft, Supervision, Resources, Project admin- istration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Declaration of Competing Interest

All of the authors of this study declare this study was funded by The Office of the Assistant Secretary of Defense for Health Affairs, endorsed by the Department of Defense, through the Congressional Directed Medical Research Programs Joint Program Committee-6, Award Number – W81XWH-20-2-0033.

All of the authors of this study have no conflicts of interest to declare.


We thank Eric N. Brown, MD, Julia Dahlstrom, RN, Ted Morrison, PhD, Howard Greene, PhD, Ms. Lucy Betterton, and Andrew Schrader, DVM, and Division of Animal Resources staff for their expertise towards preparation and conducting this study.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2022.01.039.


  1. Eastridge BJ, Holcomb JB, Shackelford S. Outcomes of traumatic hemorrhagic shock and the epidemiology of Preventable death from injury. Transfusion. 2019;59(S2): 1423-8.
  2. Drake SA, Holcomb JB, Yang Y, Thetford C, Myers L, et al. Establishing a regional trauma preventable/potentially preventable Death rate. Ann Surg. 2020;271(2): 375-82.
  3. O’Reilly DJ, Morrison JJ, Jansen JO, Apodaca AN, Rasmussen TE, et al. Prehospital blood transfusion in the en route management of severe combat trauma: a matched cohort study. J Trauma Acute Care Surg. 2014;77(3 Suppl 2):S114-20.
  4. Shackelford SA, Del Junco DJ, Powell-Dunford N, Mazuchowski EL, Howard JT, et al. Association of Prehospital Blood Product Transfusion during Medical Evacuation of combat casualties in Afghanistan with acute and 30-day survival. JAMA. 2017;318 (16):1581-91.
  5. Lewis P, Wright C. Saving the critically injured trauma patient: a retrospective anal- ysis of 1000 uses of intraosseous access. Emerg Med J. 2015;32(6):463-7.
  6. Johnson M, Inaba K, Byerly S, Falsgraf E, Lam L, et al. Intraosseous infusion as a bridge to definitive access. Am Surg. 2016;82(10):876-80.
  7. Tactical Combat Casualty Care (TCCC). Guidelines for Medical Personnel. Available at. https://www.naemt.org/docs/default-source/education-documents/tccc/tccc- updates_092017/tcccmp-curriculum-1708/00-tcccmp-guidelines/tccc-guidelines- for-medical-personnel-170131.pdf?sfvrsn.798acd92_2; 2021.
  8. Harris M, Balog R, Devries G. What is the evidence of utility for intraosseous blood transfusion in damage-control resuscitation? J Trauma Acute Care Surg. 2013;75 (5):904-6.
  9. Simmons CT. Henry Darcy (1803-1858): immortalized by his scientific legacy. Hydrgeol J. 2008;16(6):1023-38.
  10. Auten JD, McEvoy CS, Roszko PJ, Polk TM, Kachur RE, et al. Safety of pressurized intraosseous blood infusion strategies in a swine model of Hemorrhagic shock. J Surg Res. 2020;246:190-9.
  11. Auten JD, McLean JB, Kemp JD, Rozsko PJ, Fortner GA, et al. A pilot study of four intraosseous blood transfusion strategies. J Spec Oper Med. 2021;18(3):50-6.
  12. Rubal BJ, Meyers BL, Kramer SA, Hanson MA, Andrews JM, et al. Fat Intravasation from intraosseous flush and infusion procedures. Prehosp Emerg Care. 2015;19 (3):376-90.
  13. Burgert JM, Mozer J, Williams T, Gostnell J, Gegel BT, et al. Effects of intraosseous transfusion of whole blood on hemolysis and transfusion time in a swine model of hemorrhagic shock: a pilot study. AANA J. 2014;82(3):198-202.
  14. Bianchi W, George T, McEvoy C, Piehl M, Manzano A, et al. Intravenous and intraosseous blood transfusion with three different Pediatric pressure transfusion strategies in an immature swine (Sus scrofa) model of Hemorrhagic shock: a pilot study. Mil Med. 2020;185(1):121-9.
  15. Lee B, Jeung K, Lee H, Lee S, Bae S, et al. Confirmation of intraosseous cannula place- ment based on pressure measured at the cannula during squeezing the extremity in a piglet model. Resuscitation. 2014;85(1):143-7.
  16. National Research Council. Guide for the care and use of laboratory animals. 8th ed.. Washington, DC: National Academies Press; 2011..
  17. Mitchell AD, Scholz AM, Pursel VG. Total body and regional measurements of bone mineral content and bone mineral density in pigs by dual energy X-ray absorptiom- etry. J Anim Sci. 2001;79(10):2594-604.
  18. Looker AC, Borrud LG, Hughes JP, Fan B, Shepherd JA, et al. Total body bone area, bone mineral content, and bone mineral density for individuals aged 8 years and over: United States, 1999-2006. Vital Health Stat. 2013;11(253):1-78.
  19. Sulava E, Bianchi W, McEvoy CS, Roszko PJ, Zarow GJ, et al. Single versus double an- atomic site intraosseous blood transfusion in a swine model of Hemorrhagic shock. J Surg Res. 2021;267:172-81.
  20. Lairet J, Bebarta V, Lairet K, Kacprowics R, Lawler C, et al. A comparison of proximal tibia, distal femur, and proximal humerus infusion rates using the EZ-IO intraosseous device on the adult swine (Sus scrofa) model. Prehosp Emerg Care. 2013;17(2):280-4.
  21. Pasley J, Miller C, DuBose JJ, Shackelford SA, Fang R, et al. Intraosseous infusion rates under high pressure: a cadaveric comparison of anatomic sites. J Trauma Acute Care Surg. 2015;78(2):295-9.
  22. Windberger U, Bartholovitsch A, Plasenzotti R, Korak KJ, Heinze G. Whole blood vis- cosity, plasma viscosity and erythrocyte aggregation in nine mammalian species: reference values and comparison of data. Exp Physiol. 2003;88(3):431-40.
  23. Balaban O, Walia H, Tumin D, Bhalla T, Tobias JD. Efficacy of rapid fluid administra- tion using various setups and devices. Pediatr Emerg Care. 2019;35(8):539-43.
  24. Allon R, Epstein D, Shavit I. Prehospital transfusion of low titer cold-stored whole blood through the intraosseous route in a trauma patient with hemorrhagic shock. Transfusion. 2020;60(4):875-8.
  25. Hodge D, Fleisher G. Pediatric catheter flow rates. Am J Emerg Med. 1985;3(5):403-7.
  26. Plewa MC, King RW, Fenn-Buderer N, Gretzinger K, Renuart D, et al. Hematologic safety of intraosseous blood transfusion in a swine model of pediatric hemorrhagic hypovolemia. Acad Emerg Med. 1995;2(9):799-809.
  27. Butler FK, Holcomb JB, Shackelford S, Barbabella S, Bailey S, et al. Advanced resusci- tative care in tactical combat casualty care: TCCC guidelines change 18-01:14. J Spec Oper Med Winter. 2018;18(4):37-55.
  28. Hammer N, Mobius R, Gries A, Hossfeld B, Bechmann I, et al. Comparison of the fluid resuscitation rate with and without external pressure using two intraosseous infusion Systems for Adult Emergencies, the CITRIN (comparison of InTRaosseousinfusion sys- tems in emergency medicINe)-study. PLoS One. 2015.;10(12):e0143726.
  29. Ong ME, Chan YH, Oh JJ, Ngo AS. An observational, prospective study comparing tibial and humeral intraosseous access using the EZ-IO. Am J Emerg Med. 2009;27(1):8-15.
  30. Sperry JL, Guyette FX, Brown JB, Yaser MH, Triulzi TJ, et al. Prehospital plasma during air medical transport in trauma patients at risk for Hemorrhagic shock. N Engl J Med. 2018;379(4):315-26.
  31. Toshniwal G, Ahmed Z, Sengstock D. Simulated fluid resuscitation for toddlers and young children: effect of syringe size and hand fatigue. Paediatr Anaesth. 2015;25 (3):288-93.
  32. Hooper TJ, Brown CA, Letchford KA, Wright EF, Banks L. The effectiveness of blood product warming circuits in a simulated MERT environment. J R Army Med Corps. 2013;159(4):314.
  33. Jansen G, Leimkuhler K, Mertzlufft F. Intramedullary placement of intraosseous can- nulas inserted in the preclinical treatment of polytrauma patients: a retrospective, computed tomography-assisted evaluation. Anaesthesist. 2017;66(3):168-76.
  34. Rubal BJ, Meyers BL, Kramer SA, Hanson MA, Andrews JM, et al. Fat Intravasation from intraosseous flush and infusion procedures. Prehosp Emerg Care. 2015;19 (3):376-90.
  35. Dubick MA, Pfeiffer JW, Clifford CB, Runyon DE, Kramer GC. Comparison of intraosseous and intravenous delivery of hypertonic saline/ dextran in anesthetized, euvolemic pigs. Ann Emerg Med. 1992;21:498-503.
  36. Aerssens J, Boonen S, Lowet G, Dequekker J. Interspecies differences in bone compo- sition, density, and quality: potential implications for in vivo bone research*. Endo- crinology. 1998;139(2):663-70.
  37. Tsukamoto T, Pape HC. Animal models for trauma research: what are the options? Shock. 2009;31:3-10.