American Journal of Emergency Medicine xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Brief Report
Variability in intraosseous flush practices of emergency physicians☆,☆☆,★ Joseph S. Sontgerath, MD a,⁎, Bernard J. Rubal, PhD b, Robert A. DeLorenzo, MD c, Trent L. Morgan, MD a, John A. Ward, PhD b a b c
Department of Emergency Medicine, Fort Sam Houston, TX 78234 Cardiology Service, San Antonio Military Medical Center, Fort Sam Houston, TX 78234 US Army Institute of Surgical Research, Fort Sam Houston, TX 78234
a r t i c l e
i n f o
Article history: Received 26 November 2013 Received in revised form 4 March 2014 Accepted 4 March 2014 Available online xxxx
a b s t r a c t Objective: Intramedullary pressure changes during intraosseous (IO) procedures have been implicated in the intravasation of bone marrow fat and with pain in conscious patients. The objective of this study was to demonstrate inter-provider variability in pressures generated during initial flush procedures. Methods: IO cannulas were inserted into the proximal tibiae and humeri by study personnel. A second cannula was placed in the mid diaphysis of each bone to record intramedullary pressures. Fifteen emergency physicians performed 60 flushes in random order in two cadavers while flush duration and IO pressure were continuously recorded. Providers were blinded to the flush pressures they generated and the flush techniques of others. Results: The median IO pressure (IOP) generated by providers was 903 mm Hg (range, 83-2941 mm Hg) and the median flush duration was 5.2 seconds (range, 1.0-13.4 seconds). Significant differences were noted among providers in peak IOP generated (analysis of variance P b .001). Providers were consistent in the forces they generated relative to each other. An inverse, nonlinear relationship was observed between flush duration and the peak IOP generated. Significant differences were noted in intramedullary flush pressures at flush sites within cadavers (analysis of variance P: cadaver #1 P b .001; cadaver #2 P = .012). Conclusions: The IO compartment pressures generated by physicians demonstrated significant interoperator variability with greater than 35-fold difference in flush forces, and an inverse relationship between intraosseous pressure and flush duration. It may be prudent practice for providers to extend the flush over several seconds, thus limiting maximal pressures. Published by Elsevier Inc.
1. Introduction Intraosseous (IO) cannulation is an accepted means to achieve vascular access when peripheral venous access is not available. With the advent of semi-automated cannulation systems more than a million IO cannulas have been utilized in the past decade [1] with only rare complications reported [2-4]. Although numerous studies have demonstrated the success of IO placement, the speed and efficacy of IO cannulation [5-7], or compared infusion
☆ Presentation: Variability in Intraosseous Flush Practices by Trained Emergency Physicians, Oral Presentation at the SAEM 2013 Annual Meeting in Atlanta, GA, May 14 to 18, 2013. ☆☆ Financial Disclosures: None ★ Disclaimer. The opinions or assertions contained herein are the private views of the authors and are not to be construed as reflecting the views of the Department of the Army or the Department of Defense. ⁎ Corresponding author. Department of Emergency Medicine, Mike O'Callaghan Federal Medical Center, 4700 N Las Vegas Blvd, Nellis AFB, NV 89191, USA. Tel.: +1 702 653 2660. E-mail address: [email protected] (J.S. Sontgerath).
rates at different cannulation sites [8,9], few have focused on the IO cannulation practices of practitioners. Studies have suggested that intramedullary pressure changes during intraosseous procedures may be implicated in the intravasation of bone marrow fat [10-15] and with pain reported during IO infusion in conscious patients [16-18]. It is common practice to flush the IO cannula with normal saline prior to establishing infusions, and there is consensus that flush maintains IO flow [18,19]. This practice is not universal, and has not been proven to increase rates of flow. Additionally, there is limited specific physiologic rationale available regarding optimal flush techniques and limited guidance is provided by IO manufacturers. Whereas Rubal et al. had used an isolated swine femur to assess intramedullary flush forces by laboratory personnel, no studies have explored the IO flush practices of clinical practitioners at sites commonly used in humans—the tibia and humerus [15]. The objective of this study was to use a human cadaver model to examine the variability of IO flush practices among emergency physicians and to assess the magnitude of the intramedullary forces typically generated during IO flush procedures.
http://dx.doi.org/10.1016/j.ajem.2014.03.001 0735-6757/Published by Elsevier Inc.
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
2
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
2. Methods This was a prospective, observational study of a convenience sample of physicians from a group of emergency medicine residents and staff using an adult human cadaver model. Cadavers have been used to define IO cannulation sites, compare speed and ease of IO placement, and train physicians, medical students and paramedics in advanced emergency procedures [6,20,21]. Cadavers were employed in this study to provide a controlled model for repeated infusion and permitted the direct recording of intraosseous pressures to assess variability in flush procedures among providers. All physicians had been trained in IO placement during residency or subsequently with a classroom lecture and hands-on training with mannequin, human cadaver or both. A brief survey was administered verifying prior training and querying the self-reported number of prior clinical IO insertions. The study was reviewed and approved by the institutional review board at Brooke Army Medical Center. Two cadavers (cadaver 1 = thawed, cadaver 2 = fresh) were used for placement of intraosseous cannulas. We used cadavers made available from another study being run concurrently in the same facility. Our data were obtained prior to any alteration of the anatomy by the concurrent study. Both cadavers had died of complications from brain tumors. IO cannulas were placed in the proximal tibiae (25 mm) and proximal humeri (45 mm) (EZ IO, Vidacare Corp, San Antonio, TX) using a rotary power drive, for a total of 4 IO cannulas per cadaver. A second IO cannula was placed in the diaphysis of the long bones to monitor intraosseous pressures. The medullary compartment was then cleared of clotted blood by gentle flush with normal saline. Intraosseous pressures were recorded from the second cannula using a high fidelity catheter (3F Mikro-tip, Millar Instruments, Houston, TX). IO compartment pressures were recorded using pressure amplifiers (Grass Model LP122, Astro-Med, Warwick, RI) calibrated for a full scale recording from − 100 to 3500 mm Hg. A 10 mL saline flush syringe was connected to the proximal infusion cannula, and the preparation was draped. Fifteen (N = 15) emergency physicians experienced in IO cannulation procedures were directed to “flush” the IO cannulas with 10 mL of normal saline (cadaver #1, N = 9; cadaver #2, N = 6). Participants were instructed to assume they were resuscitating an adult patient with a properly inserted IO cannula and to flush as they had been previously trained. Participants were blinded to the purpose of the study, to the intraosseous pressures that they generated during
flush, and to the flush procedures of the others. Each provider was assigned a number, and those numbers were randomly drawn to determine the order of providers delivering flushes at each individual site. Intraosseous pressure (IOP) and the derivative of pressure were continuously recorded over time. Peak intraosseous pressure, flush duration (t), average flush IO pressure, and the maximum rate of increase (max + dP/dt) and decrease (max –dP/dt) in medullary pressures were measured for each IO flush. 3. Data analysis In this study data are presented as mean ± SD and medians with interquartile ranges. Variability within measurements is also expressed as the coefficient of variation. A 2-factor analysis of variance was employed to compare IO flushes between cadaver preparations and IO flush sites. A Freidman repeated measures analysis of variance was used to assess differences in IO flush practices among physicians. A nonlinear regression model (y = A e −bt; Microsoft Excel 2013; Microsoft Corporation, Redmond, WA) was employed to characterize the relationships between the duration of flush and peak IO pressure for all flush procedures (N = 60) and for flushes by multiple providers at each infusion site. P b .05 was considered significant. Statistical analyses were performed using commercially available software (IBM SPSS Statistics, version 19; Armonk, NY, or Sigma Stat, version 3.11; Systat Software, San Jose, CA). 4. Results In addition to prior training, 60% of the study participants had inserted IO cannulas within a hospital emergency department environment or during care for combat injuries. The median IOP generated (N = 60) by providers (N = 15) at all sites was 903 mm Hg (range, 83-2941 mm Hg). In 25% (15/60) of the IO flush procedures, peak IO pressure was ≥1500 mm Hg; and in 15% of flushes, the peak IO pressure exceeded 2000 mm Hg. The median flush duration was 5.12 seconds (range, 0.95-13.4 s). Fig. 1 shows an example of intraosseous pressure changes during a 10 mL IO flush into the left proximal humerus. In this example the peak IOP is 1259 mm Hg and the flush duration was 5.5 s, the average IOP during flush was 1076 mm Hg and maximum positive and negative dP/dt were 1346 and − 4278 mm Hg/s, respectively. Table 1 illustrates the variability in the intramedullary pressures generated by physicians. The coefficient of
Fig. 1. Example of 10 mL IO flush.
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
3
Table 1 Descriptive statistics (N = 60)
Mean ± SD
a
CV
Range
Percentiles
b
c
25th
Median
75th
83 47 142 −21022 0.95
2942 2329 21084 −110 13.4
279 176 568 −9093 2.9
904 465 1407 −4681 5.2
1497 1006 2983 −1513 7.6
Min
Peak IOP (mm Hg) Average IOP (mm Hg) Max + dP/t (mm Hg/s) Max - dP/t (mm Hg/s) Duration flush (s) a b c
1021 674 2953 −6042 5.4
± 826 ± 609 ± 4249 ± 5487 ± 3.0
0.81 0.90 1.44 0.91 0.57
Max
CV, coefficient of variation. Min, minimum range value. Max., maximum range value.
variation was ≥80.9% for all pressure parameters and 56.6% for flush duration. When peak intraosseous pressure was plotted (Fig. 2) as a function of flush duration for all injections, a statistically significant nonlinear relationship was observed (R 2 = 0.470, P b .001). When the peak IO pressures generated by providers were stratified by flush sites to assess the effects of the cannulation site on provider flush practice, a significant (P b .001) interaction (IO site*cadaver) was noted. Table 2 presents the results of post hoc tests, indicating that the IO compartment force providers generated differed with flush sites within and between cadavers. We therefore examined the relationship between peak IOP and flush duration at each cannulation site. Fig. 3 demonstrates that each IO site flushed by multiple physicians could be characterized by a unique relationship between flush time and peak intramedullary pressure. Good curve fits (P b .001) were noted using an exponential decay model in spite of the relatively small number of data points for each injection site (N = 9 cadaver #1 and N = 6 cadaver #2). These data suggest that intraosseous compartment forces can be predicted at a given IO site when flush times are known. It was also observed that providers who delivered high pressures at one flush site consistently delivered high pressures at other sites, and providers that delivered low pressures at one site delivered low pressures at all sites. When providers were ranked by their median delivered pressures for each cadaver preparation, a significant rank order was noted (cadaver 1: P b .001, cadaver 2: P = .012). These results suggest that there was a consistency with which each provider executed flush procedures. Although the stability of cadaveric preparations employed in the present study was not directly assessed by controlled flushes using an infusion pump, no significant trend for change in the preparation was observed over the course of repeated flushes by provider.
5. Discussion It is common practice to flush the IO after initial cannulation and after medication infusion in order to facilitate flow. Although the training adage “No Flush = No Flow” is relevant for IO use, how IO cannulas are flushed in common practice has not been previously examined. This study is the first to demonstrate the wide range of variability among emergency physicians in the forces employed to flush IO cannulas. There was a 35-fold difference among providers in the peak IO flush pressures generated. The median peak IO pressure during flush was N900 mm Hg with several physicians generating IO pressures near 3000 mm Hg. Among our providers nearly a 150-fold difference was noted in the rate of rise in medullary pressures (14221,084 mm Hg/s) and greater than a 14-fold difference in the duration of time providers flushed IO cannulas. The practice of flushing IO cannulas following insertion was well documented by early clinical proponents for IO infusions [22]. It is generally recognized that the insertion of an IO cannula produces bone fragments and disrupts the integrity of medullary tissue at the insertion site. Local debris and thromboplastin release from damaged tissue activates the clotting cascade and promotes clot formation [23]. Flush clears debris and clots and temporarily retards clot formation by diluting clotting factors. Once flow paths are cleared, the outflow resistance of the bone is relatively constant [24]. Under normal physiological conditions IO pressures are small (20%-30% arterial pressure); however, at high IO flush rates, the IO compartment pressures rise rapidly due to fixed outflow resistance of bone and the rigid medullary compartment. These flush forces are rapidly transmitted through medullary structures. Pressure transients within the IO compartment have been associated with shear strains within the bone marrow [25]. In the conscious patient, the pressure transient associated with IO flush is associated with pain [16,17,19,22]. In this setting, it is common practice for providers to titrate flush rates to the level patients find more tolerable. Our results suggest that each IO cannulation site has a unique outflow resistance given the differences observed between flush sites and cadavers, and that at each site there is an exponential relationship between flush time and peak force generated within the IO compartment. Results suggest that the intramedullary pressure forces associated with IO flush may be
Table 2 Post hoc analysis comparing peak IOP (mm Hg) using the Student-Newman-Keuls method LPH Cadaver 1 Mean (SD) Cadaver 2 Mean (SD)
Fig. 2. Peak IOP and flush duration.
a,b
a,b
765 (429)
1459 (760)
RPH
LPT
RPT
a,b
b
a,b
b
388 (402)
1195 (919)
1474 (850) b
614 (422)
1856 (801)
a,b
213 (80)
LPH, left proximal humerus; RPH, right proximal humerus; LPT, left proximal tibia; RPT, right proximal tibia. a IO sites within a cadaver (P b .05). b Compares the same IO sites between cadavers (P b .05).
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
4
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
Fig. 3. Peak IOP and flush duration by IO site.
predicted and controlled by understanding the nonlinear relationship between flush duration and IO pressures. The median IO compartment pressure observed in our study is greater than the maximum pressures observed during interventional orthopedic procedures [26] which are known to be high risk procedures for fat embolism [27]. Orloswski et al reported that fat embolism is a universal finding with IO infusions [28] and others have shown that fat emboli remain within lungs 72 hours following IO infusions [14]. In contrast, Fiallos et al found no increase in pulmonary fat embolism with IO infusion in a hypovolemic resuscitation mode [29]. Only recently have the contributions of initial flush practices been implicated as a significant contributor to the fat intravasation with IO procedures [15]. Although pulmonary fat embolism with IO use is generally considered a subclinical event [28,30], the impact of persistent fat embolism in the management of critically ill patients is unresolved [31]. The very large pressures observed in our study likely cause significant shear forces within the medullary cavity and could result in greater downstream effects than previously appreciated, although further research is required to determine the degree to which IO pressures generated during flushing contributes to increased amounts and size of pulmonary fat emboli. By way of example, our study revealed that many providers generated IO pressures that exceeded the IO pressures associated with death and fat embolism syndrome during intraosseous phlebography [32-34]. In these radiographic procedures, peak IO pressures have been estimated at 2069 mm Hg [35]. In our study, 15% of flushes (N = 9) generated pressures above 2000 mm Hg. Although the delivered volumes of contrast media in these case reports were greater than the volume of saline delivered in the IO flushes in our study, the relative risk of flushing IO cannulas with flow rates that induce such high intramedullary pressures has not been assessed. Of interest was the relative consistency with which providers applied flush pressures (same rank order) in spite of the randomization of providers to different flush sites. The implication is that certain providers will always generate higher pressures, and other providers will typically generate lower pressures. Providers likely have no concept of the intraosseous pressures they are generating with a 10 mL flush. Our study gives context to the processes that underlie the clinical presentation of pain and fat embolization associated with IO flush. Our study suggests that IO compartment pressures may be titrated by
controlling flush duration. By incorporating feedback into IO training, providers can be made aware of the IO pressures that they are generating in flush procedures, and can learn to regulate their IO pressures by controlling flush duration. 6. Limitations Although our study describes procedural aspects of IO use and does not include measures of clinical outcomes, it does provide information unlikely to be obtained during prehospital care or in an emergency department setting. Cadavers were employed to assess flush practices of providers because this model provided a controlled preparation that allowed repeated flushes. In this model, post mortem intraosseous clots and debris were removed prior to the flushes of providers. Therefore, the provider variability assessed in this study is attributed the effects IO flush rates have on intramedullary pressures rather than flush pressures which may be required to clear clots. Since the providers flushed each IO site in the cadaver only once, we did not examine the intra-provider repeatability in our study. We also report results from a convenience sample of providers, so it is unclear if the sample represents a typical provider who would be flushing the IO in practice. Compared to a live model, the lack of physiologic blood flow through the IO compartment and the inability to produce active clotting likely reduced the complexity of the exponential model used to describe the relationship between flush duration and peak IOP. Table 2 illustrates the unique flush-IO pressure relationship of each individual site, which is likely due to minor variations in placement of the IO needle and individual bone structure and anatomic variation. It appears that pressures are higher in the humeri of cadaver 1, but higher in the tibiae of cadaver 2. We believe that if more cadavers had been examined, we would likely see similar variation at different sites within subjects. This experiment only tested the pressures generated by the EZ IO (Vidacare Corp, San Antonio, TX) intraosseous needle and does not necessarily infer similar results from other manufacturers. We were unable to compare flush practices observed in our study to the specifics of the provider’s prior training. 7. Conclusion Although it accepted that flushes maintain IO flows, optimal practices for flush procedures have not been established. The IO
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
compartment pressures generated by providers demonstrated significant inter-operator variability with greater than 35-fold difference in flush forces, and an inverse relationship between flush duration and intraosseous pressures. It may be prudent practice for providers to extend the flush over several seconds, thus limiting maximal pressures. Further studies are needed, however, to determine the specific flush practices that will reduce potential risks associated with high intraosseous pressures. Acknowledgment Funding was provided by the US Army Tele-medicine and Advanced Technology Research Center. References [1] Miller LJ, Montez DF, Puga TA, Philbeck TE. Intraosseous vascular access in the 21st century: improvements further reduced complication rates. Ann Emerg Med 2012;60(48):S112. [2] Voigt J, Waltzman M, Lottenberg L. Intraosseous vascular access for in-hospital emergency use: a systematic clinical review of the literature and analysis. Pediatr Emerg Care 2012 Feb;28(2):185–99. [3] Miller L, Kramer GC, Bolleter S. Rescue access made easy. JEMS 2005;30(Suppl. 8– 18) quiz suppl 19. [4] LaRocco BG, Wang HE. Intraosseous infusion. Prehosp Emerg Care 29930(23): 280–5. [5] Leidel BA, Kirchhoff C, Braunstein V, Bogner V, Biberthaler P, Kanz KG. Comparison of two intraosseous access devices in adult patients under resuscitation in the emergency department: a prospective, randomized study. Resuscitation 2010; 81:994–9. [6] Levitan RM, Bortle CD, Snyder TA, Nitsch DA, Pisaturo JT, Butler KH. Use of a battery-operated needle driver for intraosseous access by novice users: skill acquisition with cadavers. Ann Emerg Med 2009;54:692–4. [7] Brenner T, Bernhard M, Helm M, et al. Comparison of two intraosseous infusion systems for adult emergency medical use. Resuscitation 2008;78:314–9. [8] Johnson DL, Findlay J, Macnab AJ, Susak L. Cadaver testing to validate design criteria of an adult intraosseous infusion system. Mil Med 2005;170:251–7. [9] Lairet J, Bebarta V, Lairet K, 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;17:280–4. [10] Pfeifer R, Sellei R, Pape HC. The biology of intramedullary reaming. Injury 2010;41 (Suppl. 2):S4–8. [11] Wenda K, Runkel M, Degreif J, Ritter G. Pathogenesis and clinical relevance of bone marrow embolism in medullary nailing–demonstrated by intraoperative echocardiography. Injury 1993;24(Suppl 3):S73–81. [12] Lairet J, Bebarta V, Mathis D, et al. Comparison of intraosseous infusion rates of blood under high pressure in adult hypovolemic swine model in three different limb sites. Ann Emerg Med 2012;60(4S):S75. [13] Wile UJ, Schamberg IL. Pulmonary embolism following infusion via the bone marrow. J Invest Dermatol 1942;5:173–7. [14] Plewa MC, Kaplan RM, LaCovey D, McPherson J, Stapczynski JS Klein E. Fat embolism following intraosseus infusion. Ann Emerg Med 1988;17:407.
5
[15] Rubal BJ, McKay K, Armstrong KR, Rubal MP, Marbach MJ. Variability in intraosseous pressure induced by saline flush of an intraosseous cannula by multiple practitioners. Lab Anim (NY) 2012;20;41:224–9. Erratum in: Lab Anim (NY) 2012;41:349. [16] Philbeck TE, Miller LJ, Montez D, Puga T. Hurts so good. Easing IO pain and pressure. JEMS 2010;35:58–62. [17] Schalk R, Schweigkofler U, Lotz G, Zacharowski K, Latasch L, Byhahn C. Efficacy of the EZ-IO needle driver for out-of-hospital intraosseous access—a preliminary, observational, multicenter study. Scand J Trauma Resusc Emerg Med 2011; 26:19:65. [18] Tarrow AB, Turkel H, Thompson MS. Infusions via the bone marrow and biopsy of the bone and bone marrow. Anesthesiology 1952;13:501–9. [19] Davidoff J, Fowler R, Gordon D, et al. Clinical evaluation of a novel intraosseous device for adults: prospective, 250-patient, multi-center trial. JEMS 2005;30 (Suppl 20–23). [20] Tabas JA, Rosenson J, Price DD, Rohde D, Baird CH, Dhillon N. A comprehensive, unembalmed cadaver-based course in advanced emergency procedures for medical students. Acad Emerg Med 2005;12:782–5. [21] Hubble MW, Trigg DC. Training prehospital personnel in saphenous vein cutdown and adult intraosseous access techniques. Prehosp Emerg Care 2001; 5:181–9. [22] Tocantins LM, O’Neil JF, Jones HW. Infusions of blood and other fluids via the bone marrow: application in pediatrics. JAMA 1941;117:1229–34. [23] Plewa MC, King RW, Fenn-Buderer N, Gretzinger K, Renuart D, Cruz R. Hematologic safety of intraosseous blood transfusion in a swine model of pediatric hemorrhagic hypovolemia. Acad Emerg Med 1995;2:799–809. [24] Watson WC, Ryan DM, Dubick MA, Simmons DJ, Kramer GC. High pressure delivery of resuscitation fluid through bone marrow. Acad Emerg Med 1995;2:403 (Abstract). [25] Rubal BJ, Gerhardt RT, Sartin CW, Neal CL, DeLorenzo RA. Medullary shear and pressure changes associated with high intraosseous infusion rates in an Isolated hind limb preparation. Ann Emerg Med 2010:56:S113, 2010. [26] Kröpfl A, Berger U, Neureiter H, Hertz H, Schlag G. Intramedullary pressure and bone marrow fat intravasation in unreamed femoral nailing. J Trauma 1997;42: 946–54. [27] Aoki N, Soma K, Shindo M, Kurosawa T, Ohwada T. Evaluation of potential fat emboli during placement of intramedullary nails after orthopedic fractures. Chest 1998;113:178–81. [28] Orlowski JP, Julius CJ, Petras RE, Porembka DT, Gallagher JM. The safety of intraosseous infusions: risks of fat and bone marrow emboli to the lungs. Ann Emerg Med 1989;18:1062–7. [29] Fiallos M, Kissoon N, Abdelmoneim T, et al. Fat embolism with the use of intraosseous infusion during cardiopulmonary resuscitation. Am J Med Sci 1997; 314:73–9. [30] Hasan MY, Kissoon N, Khan TM, Saldajeno V, Goldstein J, Murphy SP. Intraosseous infusion and pulmonary fat embolism. Pediatr Crit Care Med 2001;2:133–8. [31] Koessler MJ, Fabiani R, Hamer H, Pitto RP. The clinical relevance of embolic events detected by transesophageal echocardiography during cemented total hip arthroplasty: a randomized clinical trial. Anesth Analg 2001;92:49–55. [32] Thomas ML, Tighe JR. Death from fat embolism as a complication of intraosseous phlebography. Lancet 1973;22(2):1415–6. [33] Young AE, Evans IL, Irving D, Hanning CD. Fat embolism after pertrochanteric venography. Br Med J 1973;4:592. [34] Gildenhorn HL, Gildenhorn VB, Amromin G. Marrow embolism and intraosseous contrast radiography. J Am Med Assoc 1960;173:758–60. [35] Young AE, Thomas ML, Browse NL. Intraosseous phlebography and fat embolism. Br Med J 1976;2:89–90.
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Brief Report
Variability in intraosseous flush practices of emergency physicians☆,☆☆,★ Joseph S. Sontgerath, MD a,⁎, Bernard J. Rubal, PhD b, Robert A. DeLorenzo, MD c, Trent L. Morgan, MD a, John A. Ward, PhD b a b c
Department of Emergency Medicine, Fort Sam Houston, TX 78234 Cardiology Service, San Antonio Military Medical Center, Fort Sam Houston, TX 78234 US Army Institute of Surgical Research, Fort Sam Houston, TX 78234
a r t i c l e
i n f o
Article history: Received 26 November 2013 Received in revised form 4 March 2014 Accepted 4 March 2014 Available online xxxx
a b s t r a c t Objective: Intramedullary pressure changes during intraosseous (IO) procedures have been implicated in the intravasation of bone marrow fat and with pain in conscious patients. The objective of this study was to demonstrate inter-provider variability in pressures generated during initial flush procedures. Methods: IO cannulas were inserted into the proximal tibiae and humeri by study personnel. A second cannula was placed in the mid diaphysis of each bone to record intramedullary pressures. Fifteen emergency physicians performed 60 flushes in random order in two cadavers while flush duration and IO pressure were continuously recorded. Providers were blinded to the flush pressures they generated and the flush techniques of others. Results: The median IO pressure (IOP) generated by providers was 903 mm Hg (range, 83-2941 mm Hg) and the median flush duration was 5.2 seconds (range, 1.0-13.4 seconds). Significant differences were noted among providers in peak IOP generated (analysis of variance P b .001). Providers were consistent in the forces they generated relative to each other. An inverse, nonlinear relationship was observed between flush duration and the peak IOP generated. Significant differences were noted in intramedullary flush pressures at flush sites within cadavers (analysis of variance P: cadaver #1 P b .001; cadaver #2 P = .012). Conclusions: The IO compartment pressures generated by physicians demonstrated significant interoperator variability with greater than 35-fold difference in flush forces, and an inverse relationship between intraosseous pressure and flush duration. It may be prudent practice for providers to extend the flush over several seconds, thus limiting maximal pressures. Published by Elsevier Inc.
1. Introduction Intraosseous (IO) cannulation is an accepted means to achieve vascular access when peripheral venous access is not available. With the advent of semi-automated cannulation systems more than a million IO cannulas have been utilized in the past decade [1] with only rare complications reported [2-4]. Although numerous studies have demonstrated the success of IO placement, the speed and efficacy of IO cannulation [5-7], or compared infusion
☆ Presentation: Variability in Intraosseous Flush Practices by Trained Emergency Physicians, Oral Presentation at the SAEM 2013 Annual Meeting in Atlanta, GA, May 14 to 18, 2013. ☆☆ Financial Disclosures: None ★ Disclaimer. The opinions or assertions contained herein are the private views of the authors and are not to be construed as reflecting the views of the Department of the Army or the Department of Defense. ⁎ Corresponding author. Department of Emergency Medicine, Mike O'Callaghan Federal Medical Center, 4700 N Las Vegas Blvd, Nellis AFB, NV 89191, USA. Tel.: +1 702 653 2660. E-mail address: [email protected] (J.S. Sontgerath).
rates at different cannulation sites [8,9], few have focused on the IO cannulation practices of practitioners. Studies have suggested that intramedullary pressure changes during intraosseous procedures may be implicated in the intravasation of bone marrow fat [10-15] and with pain reported during IO infusion in conscious patients [16-18]. It is common practice to flush the IO cannula with normal saline prior to establishing infusions, and there is consensus that flush maintains IO flow [18,19]. This practice is not universal, and has not been proven to increase rates of flow. Additionally, there is limited specific physiologic rationale available regarding optimal flush techniques and limited guidance is provided by IO manufacturers. Whereas Rubal et al. had used an isolated swine femur to assess intramedullary flush forces by laboratory personnel, no studies have explored the IO flush practices of clinical practitioners at sites commonly used in humans—the tibia and humerus [15]. The objective of this study was to use a human cadaver model to examine the variability of IO flush practices among emergency physicians and to assess the magnitude of the intramedullary forces typically generated during IO flush procedures.
http://dx.doi.org/10.1016/j.ajem.2014.03.001 0735-6757/Published by Elsevier Inc.
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
2
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
2. Methods This was a prospective, observational study of a convenience sample of physicians from a group of emergency medicine residents and staff using an adult human cadaver model. Cadavers have been used to define IO cannulation sites, compare speed and ease of IO placement, and train physicians, medical students and paramedics in advanced emergency procedures [6,20,21]. Cadavers were employed in this study to provide a controlled model for repeated infusion and permitted the direct recording of intraosseous pressures to assess variability in flush procedures among providers. All physicians had been trained in IO placement during residency or subsequently with a classroom lecture and hands-on training with mannequin, human cadaver or both. A brief survey was administered verifying prior training and querying the self-reported number of prior clinical IO insertions. The study was reviewed and approved by the institutional review board at Brooke Army Medical Center. Two cadavers (cadaver 1 = thawed, cadaver 2 = fresh) were used for placement of intraosseous cannulas. We used cadavers made available from another study being run concurrently in the same facility. Our data were obtained prior to any alteration of the anatomy by the concurrent study. Both cadavers had died of complications from brain tumors. IO cannulas were placed in the proximal tibiae (25 mm) and proximal humeri (45 mm) (EZ IO, Vidacare Corp, San Antonio, TX) using a rotary power drive, for a total of 4 IO cannulas per cadaver. A second IO cannula was placed in the diaphysis of the long bones to monitor intraosseous pressures. The medullary compartment was then cleared of clotted blood by gentle flush with normal saline. Intraosseous pressures were recorded from the second cannula using a high fidelity catheter (3F Mikro-tip, Millar Instruments, Houston, TX). IO compartment pressures were recorded using pressure amplifiers (Grass Model LP122, Astro-Med, Warwick, RI) calibrated for a full scale recording from − 100 to 3500 mm Hg. A 10 mL saline flush syringe was connected to the proximal infusion cannula, and the preparation was draped. Fifteen (N = 15) emergency physicians experienced in IO cannulation procedures were directed to “flush” the IO cannulas with 10 mL of normal saline (cadaver #1, N = 9; cadaver #2, N = 6). Participants were instructed to assume they were resuscitating an adult patient with a properly inserted IO cannula and to flush as they had been previously trained. Participants were blinded to the purpose of the study, to the intraosseous pressures that they generated during
flush, and to the flush procedures of the others. Each provider was assigned a number, and those numbers were randomly drawn to determine the order of providers delivering flushes at each individual site. Intraosseous pressure (IOP) and the derivative of pressure were continuously recorded over time. Peak intraosseous pressure, flush duration (t), average flush IO pressure, and the maximum rate of increase (max + dP/dt) and decrease (max –dP/dt) in medullary pressures were measured for each IO flush. 3. Data analysis In this study data are presented as mean ± SD and medians with interquartile ranges. Variability within measurements is also expressed as the coefficient of variation. A 2-factor analysis of variance was employed to compare IO flushes between cadaver preparations and IO flush sites. A Freidman repeated measures analysis of variance was used to assess differences in IO flush practices among physicians. A nonlinear regression model (y = A e −bt; Microsoft Excel 2013; Microsoft Corporation, Redmond, WA) was employed to characterize the relationships between the duration of flush and peak IO pressure for all flush procedures (N = 60) and for flushes by multiple providers at each infusion site. P b .05 was considered significant. Statistical analyses were performed using commercially available software (IBM SPSS Statistics, version 19; Armonk, NY, or Sigma Stat, version 3.11; Systat Software, San Jose, CA). 4. Results In addition to prior training, 60% of the study participants had inserted IO cannulas within a hospital emergency department environment or during care for combat injuries. The median IOP generated (N = 60) by providers (N = 15) at all sites was 903 mm Hg (range, 83-2941 mm Hg). In 25% (15/60) of the IO flush procedures, peak IO pressure was ≥1500 mm Hg; and in 15% of flushes, the peak IO pressure exceeded 2000 mm Hg. The median flush duration was 5.12 seconds (range, 0.95-13.4 s). Fig. 1 shows an example of intraosseous pressure changes during a 10 mL IO flush into the left proximal humerus. In this example the peak IOP is 1259 mm Hg and the flush duration was 5.5 s, the average IOP during flush was 1076 mm Hg and maximum positive and negative dP/dt were 1346 and − 4278 mm Hg/s, respectively. Table 1 illustrates the variability in the intramedullary pressures generated by physicians. The coefficient of
Fig. 1. Example of 10 mL IO flush.
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
3
Table 1 Descriptive statistics (N = 60)
Mean ± SD
a
CV
Range
Percentiles
b
c
25th
Median
75th
83 47 142 −21022 0.95
2942 2329 21084 −110 13.4
279 176 568 −9093 2.9
904 465 1407 −4681 5.2
1497 1006 2983 −1513 7.6
Min
Peak IOP (mm Hg) Average IOP (mm Hg) Max + dP/t (mm Hg/s) Max - dP/t (mm Hg/s) Duration flush (s) a b c
1021 674 2953 −6042 5.4
± 826 ± 609 ± 4249 ± 5487 ± 3.0
0.81 0.90 1.44 0.91 0.57
Max
CV, coefficient of variation. Min, minimum range value. Max., maximum range value.
variation was ≥80.9% for all pressure parameters and 56.6% for flush duration. When peak intraosseous pressure was plotted (Fig. 2) as a function of flush duration for all injections, a statistically significant nonlinear relationship was observed (R 2 = 0.470, P b .001). When the peak IO pressures generated by providers were stratified by flush sites to assess the effects of the cannulation site on provider flush practice, a significant (P b .001) interaction (IO site*cadaver) was noted. Table 2 presents the results of post hoc tests, indicating that the IO compartment force providers generated differed with flush sites within and between cadavers. We therefore examined the relationship between peak IOP and flush duration at each cannulation site. Fig. 3 demonstrates that each IO site flushed by multiple physicians could be characterized by a unique relationship between flush time and peak intramedullary pressure. Good curve fits (P b .001) were noted using an exponential decay model in spite of the relatively small number of data points for each injection site (N = 9 cadaver #1 and N = 6 cadaver #2). These data suggest that intraosseous compartment forces can be predicted at a given IO site when flush times are known. It was also observed that providers who delivered high pressures at one flush site consistently delivered high pressures at other sites, and providers that delivered low pressures at one site delivered low pressures at all sites. When providers were ranked by their median delivered pressures for each cadaver preparation, a significant rank order was noted (cadaver 1: P b .001, cadaver 2: P = .012). These results suggest that there was a consistency with which each provider executed flush procedures. Although the stability of cadaveric preparations employed in the present study was not directly assessed by controlled flushes using an infusion pump, no significant trend for change in the preparation was observed over the course of repeated flushes by provider.
5. Discussion It is common practice to flush the IO after initial cannulation and after medication infusion in order to facilitate flow. Although the training adage “No Flush = No Flow” is relevant for IO use, how IO cannulas are flushed in common practice has not been previously examined. This study is the first to demonstrate the wide range of variability among emergency physicians in the forces employed to flush IO cannulas. There was a 35-fold difference among providers in the peak IO flush pressures generated. The median peak IO pressure during flush was N900 mm Hg with several physicians generating IO pressures near 3000 mm Hg. Among our providers nearly a 150-fold difference was noted in the rate of rise in medullary pressures (14221,084 mm Hg/s) and greater than a 14-fold difference in the duration of time providers flushed IO cannulas. The practice of flushing IO cannulas following insertion was well documented by early clinical proponents for IO infusions [22]. It is generally recognized that the insertion of an IO cannula produces bone fragments and disrupts the integrity of medullary tissue at the insertion site. Local debris and thromboplastin release from damaged tissue activates the clotting cascade and promotes clot formation [23]. Flush clears debris and clots and temporarily retards clot formation by diluting clotting factors. Once flow paths are cleared, the outflow resistance of the bone is relatively constant [24]. Under normal physiological conditions IO pressures are small (20%-30% arterial pressure); however, at high IO flush rates, the IO compartment pressures rise rapidly due to fixed outflow resistance of bone and the rigid medullary compartment. These flush forces are rapidly transmitted through medullary structures. Pressure transients within the IO compartment have been associated with shear strains within the bone marrow [25]. In the conscious patient, the pressure transient associated with IO flush is associated with pain [16,17,19,22]. In this setting, it is common practice for providers to titrate flush rates to the level patients find more tolerable. Our results suggest that each IO cannulation site has a unique outflow resistance given the differences observed between flush sites and cadavers, and that at each site there is an exponential relationship between flush time and peak force generated within the IO compartment. Results suggest that the intramedullary pressure forces associated with IO flush may be
Table 2 Post hoc analysis comparing peak IOP (mm Hg) using the Student-Newman-Keuls method LPH Cadaver 1 Mean (SD) Cadaver 2 Mean (SD)
Fig. 2. Peak IOP and flush duration.
a,b
a,b
765 (429)
1459 (760)
RPH
LPT
RPT
a,b
b
a,b
b
388 (402)
1195 (919)
1474 (850) b
614 (422)
1856 (801)
a,b
213 (80)
LPH, left proximal humerus; RPH, right proximal humerus; LPT, left proximal tibia; RPT, right proximal tibia. a IO sites within a cadaver (P b .05). b Compares the same IO sites between cadavers (P b .05).
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
4
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
Fig. 3. Peak IOP and flush duration by IO site.
predicted and controlled by understanding the nonlinear relationship between flush duration and IO pressures. The median IO compartment pressure observed in our study is greater than the maximum pressures observed during interventional orthopedic procedures [26] which are known to be high risk procedures for fat embolism [27]. Orloswski et al reported that fat embolism is a universal finding with IO infusions [28] and others have shown that fat emboli remain within lungs 72 hours following IO infusions [14]. In contrast, Fiallos et al found no increase in pulmonary fat embolism with IO infusion in a hypovolemic resuscitation mode [29]. Only recently have the contributions of initial flush practices been implicated as a significant contributor to the fat intravasation with IO procedures [15]. Although pulmonary fat embolism with IO use is generally considered a subclinical event [28,30], the impact of persistent fat embolism in the management of critically ill patients is unresolved [31]. The very large pressures observed in our study likely cause significant shear forces within the medullary cavity and could result in greater downstream effects than previously appreciated, although further research is required to determine the degree to which IO pressures generated during flushing contributes to increased amounts and size of pulmonary fat emboli. By way of example, our study revealed that many providers generated IO pressures that exceeded the IO pressures associated with death and fat embolism syndrome during intraosseous phlebography [32-34]. In these radiographic procedures, peak IO pressures have been estimated at 2069 mm Hg [35]. In our study, 15% of flushes (N = 9) generated pressures above 2000 mm Hg. Although the delivered volumes of contrast media in these case reports were greater than the volume of saline delivered in the IO flushes in our study, the relative risk of flushing IO cannulas with flow rates that induce such high intramedullary pressures has not been assessed. Of interest was the relative consistency with which providers applied flush pressures (same rank order) in spite of the randomization of providers to different flush sites. The implication is that certain providers will always generate higher pressures, and other providers will typically generate lower pressures. Providers likely have no concept of the intraosseous pressures they are generating with a 10 mL flush. Our study gives context to the processes that underlie the clinical presentation of pain and fat embolization associated with IO flush. Our study suggests that IO compartment pressures may be titrated by
controlling flush duration. By incorporating feedback into IO training, providers can be made aware of the IO pressures that they are generating in flush procedures, and can learn to regulate their IO pressures by controlling flush duration. 6. Limitations Although our study describes procedural aspects of IO use and does not include measures of clinical outcomes, it does provide information unlikely to be obtained during prehospital care or in an emergency department setting. Cadavers were employed to assess flush practices of providers because this model provided a controlled preparation that allowed repeated flushes. In this model, post mortem intraosseous clots and debris were removed prior to the flushes of providers. Therefore, the provider variability assessed in this study is attributed the effects IO flush rates have on intramedullary pressures rather than flush pressures which may be required to clear clots. Since the providers flushed each IO site in the cadaver only once, we did not examine the intra-provider repeatability in our study. We also report results from a convenience sample of providers, so it is unclear if the sample represents a typical provider who would be flushing the IO in practice. Compared to a live model, the lack of physiologic blood flow through the IO compartment and the inability to produce active clotting likely reduced the complexity of the exponential model used to describe the relationship between flush duration and peak IOP. Table 2 illustrates the unique flush-IO pressure relationship of each individual site, which is likely due to minor variations in placement of the IO needle and individual bone structure and anatomic variation. It appears that pressures are higher in the humeri of cadaver 1, but higher in the tibiae of cadaver 2. We believe that if more cadavers had been examined, we would likely see similar variation at different sites within subjects. This experiment only tested the pressures generated by the EZ IO (Vidacare Corp, San Antonio, TX) intraosseous needle and does not necessarily infer similar results from other manufacturers. We were unable to compare flush practices observed in our study to the specifics of the provider’s prior training. 7. Conclusion Although it accepted that flushes maintain IO flows, optimal practices for flush procedures have not been established. The IO
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
J.S. Sontgerath et al. / American Journal of Emergency Medicine xxx (2014) xxx–xxx
compartment pressures generated by providers demonstrated significant inter-operator variability with greater than 35-fold difference in flush forces, and an inverse relationship between flush duration and intraosseous pressures. It may be prudent practice for providers to extend the flush over several seconds, thus limiting maximal pressures. Further studies are needed, however, to determine the specific flush practices that will reduce potential risks associated with high intraosseous pressures. Acknowledgment Funding was provided by the US Army Tele-medicine and Advanced Technology Research Center. References [1] Miller LJ, Montez DF, Puga TA, Philbeck TE. Intraosseous vascular access in the 21st century: improvements further reduced complication rates. Ann Emerg Med 2012;60(48):S112. [2] Voigt J, Waltzman M, Lottenberg L. Intraosseous vascular access for in-hospital emergency use: a systematic clinical review of the literature and analysis. Pediatr Emerg Care 2012 Feb;28(2):185–99. [3] Miller L, Kramer GC, Bolleter S. Rescue access made easy. JEMS 2005;30(Suppl. 8– 18) quiz suppl 19. [4] LaRocco BG, Wang HE. Intraosseous infusion. Prehosp Emerg Care 29930(23): 280–5. [5] Leidel BA, Kirchhoff C, Braunstein V, Bogner V, Biberthaler P, Kanz KG. Comparison of two intraosseous access devices in adult patients under resuscitation in the emergency department: a prospective, randomized study. Resuscitation 2010; 81:994–9. [6] Levitan RM, Bortle CD, Snyder TA, Nitsch DA, Pisaturo JT, Butler KH. Use of a battery-operated needle driver for intraosseous access by novice users: skill acquisition with cadavers. Ann Emerg Med 2009;54:692–4. [7] Brenner T, Bernhard M, Helm M, et al. Comparison of two intraosseous infusion systems for adult emergency medical use. Resuscitation 2008;78:314–9. [8] Johnson DL, Findlay J, Macnab AJ, Susak L. Cadaver testing to validate design criteria of an adult intraosseous infusion system. Mil Med 2005;170:251–7. [9] Lairet J, Bebarta V, Lairet K, 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;17:280–4. [10] Pfeifer R, Sellei R, Pape HC. The biology of intramedullary reaming. Injury 2010;41 (Suppl. 2):S4–8. [11] Wenda K, Runkel M, Degreif J, Ritter G. Pathogenesis and clinical relevance of bone marrow embolism in medullary nailing–demonstrated by intraoperative echocardiography. Injury 1993;24(Suppl 3):S73–81. [12] Lairet J, Bebarta V, Mathis D, et al. Comparison of intraosseous infusion rates of blood under high pressure in adult hypovolemic swine model in three different limb sites. Ann Emerg Med 2012;60(4S):S75. [13] Wile UJ, Schamberg IL. Pulmonary embolism following infusion via the bone marrow. J Invest Dermatol 1942;5:173–7. [14] Plewa MC, Kaplan RM, LaCovey D, McPherson J, Stapczynski JS Klein E. Fat embolism following intraosseus infusion. Ann Emerg Med 1988;17:407.
5
[15] Rubal BJ, McKay K, Armstrong KR, Rubal MP, Marbach MJ. Variability in intraosseous pressure induced by saline flush of an intraosseous cannula by multiple practitioners. Lab Anim (NY) 2012;20;41:224–9. Erratum in: Lab Anim (NY) 2012;41:349. [16] Philbeck TE, Miller LJ, Montez D, Puga T. Hurts so good. Easing IO pain and pressure. JEMS 2010;35:58–62. [17] Schalk R, Schweigkofler U, Lotz G, Zacharowski K, Latasch L, Byhahn C. Efficacy of the EZ-IO needle driver for out-of-hospital intraosseous access—a preliminary, observational, multicenter study. Scand J Trauma Resusc Emerg Med 2011; 26:19:65. [18] Tarrow AB, Turkel H, Thompson MS. Infusions via the bone marrow and biopsy of the bone and bone marrow. Anesthesiology 1952;13:501–9. [19] Davidoff J, Fowler R, Gordon D, et al. Clinical evaluation of a novel intraosseous device for adults: prospective, 250-patient, multi-center trial. JEMS 2005;30 (Suppl 20–23). [20] Tabas JA, Rosenson J, Price DD, Rohde D, Baird CH, Dhillon N. A comprehensive, unembalmed cadaver-based course in advanced emergency procedures for medical students. Acad Emerg Med 2005;12:782–5. [21] Hubble MW, Trigg DC. Training prehospital personnel in saphenous vein cutdown and adult intraosseous access techniques. Prehosp Emerg Care 2001; 5:181–9. [22] Tocantins LM, O’Neil JF, Jones HW. Infusions of blood and other fluids via the bone marrow: application in pediatrics. JAMA 1941;117:1229–34. [23] Plewa MC, King RW, Fenn-Buderer N, Gretzinger K, Renuart D, Cruz R. Hematologic safety of intraosseous blood transfusion in a swine model of pediatric hemorrhagic hypovolemia. Acad Emerg Med 1995;2:799–809. [24] Watson WC, Ryan DM, Dubick MA, Simmons DJ, Kramer GC. High pressure delivery of resuscitation fluid through bone marrow. Acad Emerg Med 1995;2:403 (Abstract). [25] Rubal BJ, Gerhardt RT, Sartin CW, Neal CL, DeLorenzo RA. Medullary shear and pressure changes associated with high intraosseous infusion rates in an Isolated hind limb preparation. Ann Emerg Med 2010:56:S113, 2010. [26] Kröpfl A, Berger U, Neureiter H, Hertz H, Schlag G. Intramedullary pressure and bone marrow fat intravasation in unreamed femoral nailing. J Trauma 1997;42: 946–54. [27] Aoki N, Soma K, Shindo M, Kurosawa T, Ohwada T. Evaluation of potential fat emboli during placement of intramedullary nails after orthopedic fractures. Chest 1998;113:178–81. [28] Orlowski JP, Julius CJ, Petras RE, Porembka DT, Gallagher JM. The safety of intraosseous infusions: risks of fat and bone marrow emboli to the lungs. Ann Emerg Med 1989;18:1062–7. [29] Fiallos M, Kissoon N, Abdelmoneim T, et al. Fat embolism with the use of intraosseous infusion during cardiopulmonary resuscitation. Am J Med Sci 1997; 314:73–9. [30] Hasan MY, Kissoon N, Khan TM, Saldajeno V, Goldstein J, Murphy SP. Intraosseous infusion and pulmonary fat embolism. Pediatr Crit Care Med 2001;2:133–8. [31] Koessler MJ, Fabiani R, Hamer H, Pitto RP. The clinical relevance of embolic events detected by transesophageal echocardiography during cemented total hip arthroplasty: a randomized clinical trial. Anesth Analg 2001;92:49–55. [32] Thomas ML, Tighe JR. Death from fat embolism as a complication of intraosseous phlebography. Lancet 1973;22(2):1415–6. [33] Young AE, Evans IL, Irving D, Hanning CD. Fat embolism after pertrochanteric venography. Br Med J 1973;4:592. [34] Gildenhorn HL, Gildenhorn VB, Amromin G. Marrow embolism and intraosseous contrast radiography. J Am Med Assoc 1960;173:758–60. [35] Young AE, Thomas ML, Browse NL. Intraosseous phlebography and fat embolism. Br Med J 1976;2:89–90.
Please cite this article as: Sontgerath JS, et al, Variability in intraosseous flush practices of emergency physicians, Am J Emerg Med (2014), http://dx.doi.org/10.1016/j.ajem.2014.03.001
