Is the intraosseous access route fast and efficacious compared to conventional central venous catheterization in adult patients under resuscitation in the emergency department? A prospective observational pilot study

Success rate and time need for vascular access is crucial in the emergency patient under resuscitation. However, peripheral intravenous (IV) access might be difficult especially in the dehydrated or hemodynamic unstable, injured or critically ill patient with collapsed peripheral veins. Also chemotherapy and long-term IV drug addiction can lead to inaccessible peripheral veins. Failure rates of IV access in emergency situations are described between 10-40% [1‐3]. Reattempts to gain vascular access lead to valuable time loss with potential subsequent influence of patients' safety. The average time necessary for peripheral IV catheterization is reported to add up to 2.5-13 min, and sometimes even up to 30 min in patients with difficult to access peripheral veins [1‐5]. This can lead to a delay of necessary treatment and longer on-scene times [6]. Longer on-scene times might be followed by additional delay in the emergency department, when reattempting vascular access. Also isolated prolongation of on-scene time delays definitive treatment. Time lag of necessary diagnostic and treatment procedures consecutively compromises the emergency patient [7, 8]. Therefore, current evidence based practice management guidelines for the prehospital fluid resuscitation in injured patients of the Eastern Association for the Surgery of Trauma (EAST) recommends not to perform vascular access on scene, if it delays patient transport to definitive care [9].

Alternative ways of drug and fluid administration are sublingual, endotracheal, subcutaneous and intramuscular application. However, these options do not really reflect a reasonable possibility, due to its uncertain and uncontrolled administration of substance dosages with incalculable pharmacodynamic and pharmacokinetic effects. Furthermore, only small amounts of certain applicable drugs can be given and volume resuscitation or transfusion of blood products are impossible at all [10, 11].

If vascular access is necessary in the acute setting of unstable patients admitted to the emergency department, and peripheral IV cannulation is impossible, central venous catheterization (CVC) is a common alternative procedure [10, 11]. Besides providing vascular access for fluid resuscitation, CVC also allows hemodynamic monitoring [12‐15]. However CVC is relatively time-consuming and associated with relevant risks for the patient, especially in the emergency setting. Most frequent complications include venous thrombosis, catheter related infections, arterial puncture and pneumothorax [12, 14‐17].

Consequently, a different vascular access technique may be reasonable to increase patients' safety, at least as a bridging procedure during ongoing resuscitation efforts until the patient is in a more stable condition. In this respect, intraosseous (IO) vascular access is an option, already established in paediatric patients for decades [18, 19]. Its importance in adults is less propagated, especially for in-hospital use. In general, IO cannulation of the non-collapsible and highly vasculated intramedullary venousplexus of the concellous bone marrow can provide a rapid, safe and easy vascular access to administer drugs, volume and blood products to the emergency patient. All metaphyseal segments of long bones are filled with highly perfused bone marrow, which are able to transport applied drugs and fluids to the central vascular system rapidly. Transport times are reported to be 1-2 min, even during cardiopulmonary resuscitation with chest compressions [10, 11]. Therefore the aim of the present study was to compare the IO access versus CVC regarding success rate of the procedure on first attempt and procedure time needed in adult patients under trauma or medical resuscitation admitted to our emergency department with impossible peripheral IV access. We hypothesised, that CVC is faster and more efficacious compared to IO access.

Four women and six men, ranging in age from 18 to 70 (on average 39.7 ± 18.2) years were included. The IO insertion site was proximal tibial in four patients, and humeral in six patients. CVC was achieved in three internal jugular veins, and in seven subclavian veins.

The success rate on first attempt was 90% for IO access versus 60% for CVC. One IO cannulation failed due to operator mishandling by not selecting the correct insertion site at the proximal humerus. The IO catheter did not reach the bone marrow because of the overlying soft tissue at the incorrect insertion site. Four CVC procedures failed at first attempt, requiring at least one more attempt. In all unsuccessful CVC attempts the guidewire was unable to be inserted or advanced into the vessel.

The mean time required for the vascular access procedure was significantly shorter (p < 0.001) for IO cannulation (2.3 min ± 0.8; time range 1-3 min) compared to CVC (9.9 min ± 3.7; time range 5-17 min) (Fig. 3).

Beside the above mentioned unsuccessful access procedures on first attempt following IO cannulation or CVC, no further complications were detected. Especially no malposition, dislodgment, bleeding, compartment syndrome, arterial puncture, haematothorax, pneumothorax, venous thrombosis and vascular access related infection was observed.

Present guidelines of the Eastern Association for the Surgery of Trauma (EAST) and the National Association of EMS Physicians (NAEMSP) advocate IO access in the prehospital setting, if peripheral venous catheterization is impossible [9, 28]. Because lacking peripheral IV access is also an issue in adults under in-hospital resuscitation, we were interested whether IO approach could be an alternative in the emergency department. To our knowledge, no other study has prospectively compared IO access versus standard alternative vascular access procedures in a real scenario for in-hospital patient care yet. Therefore, we investigated the in-hospital IO approach versus CVC in adult patients under resuscitation lacking peripheral IV access. IO cannulation was performed as a bridging procedure and was compared to CVC regarding success rate on first attempt and procedure duration. Our results show that IO vascular access was significantly more successful and required significantly less time when compared to CVC.

Today, there are different techniques of IO access available, also for the adult. Several out-of-hospital trials documented, that IO approach is consistent, rapid, and safe for drug delivery and fluid resuscitation in adults. Different investigators reported successful vascular access within 1-2 min in 72-100% of patients [21‐24, 26, 27]. In two prospective out-of-hospital trials including more than 500 adults, IO cannulation was even successful within 20 sec in 95-97% of patients [21, 26].

Only a small number of trials evaluated the isolated utilization of in-hospital IO access in severely injured or critically ill adults. Ong et al. published their experience of 24 tibial and 11 humeral successful IO insertions in altogether 24 adults. Success rate at first attempt was 97% (34/35); a second attempt was necessary for one tibial access. All insertions were achieved within 20 sec, relevant complications did not occur [25]. Valdes already reported in 1977 about IO access in 15 critically ill adults (18-86 years of age) with a success rate of 87% (13/15). In these patients, an average of 4 litres were administrated for an average of 5.4 days. Complications like infection or emboli did not occur. Application times for IO access were not available [29]. Iserson described the successful application of IO access in 22 adults (36-84 years of age) suffering cardiac arrest. Application time was <1 min and complications were not reported [30]. Cooper et al. published their experience of 22 successful inserted IO needles in adults in the military combat environment. Success rate was 97%, relevant complications did not occur, especially no infection [31]. In line with these findings, our results show also a high success rate of 90% and a low mean procedure time of 2.3 min. However, our recorded procedure time included not only the IO insertion itself, but also the preparation of the insertion site, the device and its assembling until the administration of drugs or fluids.

The majority of insertion sites for IO access in the adult include the sternum, medial clavicle, proximal humerus, distal radius, proximal tibia, distal tibia and distal fibula [21‐27, 29‐31].

Following an instruction course of a maximum of 2 hours, the safe and rapid use of diverse IO access devices are demonstrated by different studies. Regarding learnability and handling, data reveal success rates of 93-100% in IO access performance within 2 min [32‐37].

Most drugs can be administered IO in equivalent dosage and with the same time effect compared to IV. Pharmacodynamic and pharmacokinetic effects of IO applied drugs and infusions are well described in the literature [38‐41]. Currently no resuscitation drugs are contraindicated for IO administration. In our study, we mainly administered common resuscitation drugs, beside crystalloid and colloid solutions, red packed cells, and fresh frozen plasma.

Over all, flow rates of IO vascular access are lower than large bore peripheral IV catheters, and depend on patients' age, site of insertion and cannula size. Most IO cannulas for adults are 15-gauge needles, and enable flow rates comparable to a 20-gauge peripheral IV catheter. IO Flow rates on gravidity account between 10-34 ml/min and can be increased up to 80-165 ml/min using a pressure bag [25, 42‐45]. Therefore IO rapid volume resuscitation is limited, however two or more IO cannulations in the same patient may facilitate fluid therapy [25].

Limitations and risks of IO access have to be considered. According to manufacturers' recommendations, fractured bones should not be cannulated due to the risk of extravasation at the fracture site. Furthermore, to minimize infection risks, areas of limb or joint prosthesis should not be selected as insertion site. IO needles should be removed within 24 hours after insertion and no further IO cannulation should be performed at the same site within 24 hours. Obvious infection at the insertion site contraindicates IO access. However, in life threatening conditions, these contraindications are relative.

Regarding complications following IO access, the rate of adverse events is described low in literature. Immediate complications include device failure, extravasation, fat embolism, fracture of the canullated bone and compartment syndrome of the concerning limb. The only prospective studies of IO access, including 553 adults, described none of these complications [21, 25, 26]. Late complications following IO access include osteomyelitis and bone marrow necrosis. Subsequent 4,270 IO cannulations of the sternum or tibia in children, infection rate was 0.6% [46]. Other authors found fewer or no infections at all. The risk for infection seems to correlate with exposure time of IO access exceeding several days, pre-existing bacteraemia, insufficient asepsis on insertion and usage as well as administration of hypertonic fluids [21, 26, 47]. Negative long-term effects for the cannulated bone, epiphysial plates or bone marrow in humans following IO infusion have not been demonstrated so far [48]. In our study population no IO related complication was detectable. Especially device-associated infection could be excluded by routine inspection of the insertion site three times daily and microbiological examination of each removed IO cannula.

Alternative vascular access techniques in the adult patient under resuscitation with impossible peripheral IV catheterization include CVC, ultrasound-guided catheterization of peripheral veins, and saphenous vein cutdown.

CVC provides vascular access for fluid resuscitation, vasoactive medication and antibiotics. Furthermore, it allows hemodynamic monitoring and cardiac pacing [12‐15]. However, CVC is relatively time-consuming and associated with complications, especially in the emergency setting. Complication rates for CVC are reported around 15%, including malposition, arterial puncture, hematoma, pneumothorax, venous thrombosis and catheter related infections [12, 14‐17]. The average risk of CVC related infections in the critical ill patient accounts for 5.3 per 1,000 catheter days and leads to 80,000 CVC-associated bloodstream infections and consecutively up to 28,000 deaths in U.S. intensive care units each year. Mortality for hospitalized patients with CVC related infection accounts for 12-25%. The attributable cost per infection is estimated USD 34,508-56,000. Furthermore, CVC is relatively contraindicated for fibrinolytic therapy [10, 49, 50]. In our study none CVC related complications were detectable.

Ultrasound-guided peripheral IV access enables success rates of 92%. However, the necessary time adds up to 2.5-4 min for catheter cannulation itself, and up to 13 min for the whole procedure. Furthermore, this approach requires the ultrasound device and an experienced operator [4, 5, 51].

Saphenous vein cutdown is slow, less successful and associated with risks. The reported time need for this invasive vascular access procedure was 2-7.6 min with a success rate of 69-94%, when performed by experienced personnel. Trauma to the lower extremities might preclude saphenous vein cutdown. Time effect of administered drugs and fluids to the saphenous vein may be delayed due to long distance between vein cutdown and the heart, especially in shock conditions with impaired circulation [52‐55].

Several potential limitations of this study have to be addressed. First, the sample size of this study is small due to its preliminary analysis of the first 10 consecutive patients. However, there was a statistically significant difference in procedure times of enabling vascular access, resulting in a time benefit for the IO approach (p < 0.001). Second, a potential bias of the investigators favouring towards the IO access was limited by performing both vascular access methods simultaneously by two independent operators. Third, differences in access success rates on first attempt and procedure times due to unequal experience in the applied techniques were limited by determining each participant to be an experienced consultant with long lasting expertise in resuscitation in the emergency department following standardized protocols.

As a consequence of the present study, we continued the IO vascular access protocol as a bridging procedure in adult patients under resuscitation in the emergency department with impossible peripheral IV access. To evaluate the handling of different IO access devices, we included different IO systems in our modified protocol of the ongoing trial.

IO vascular access is a safe, reliable and rapid option in the acute setting of adult patients under resuscitation with inaccessible peripheral veins in the emergency department. Compared to CVC, IO cannulation is more successful on first attempt and requires significantly less time. However, IO access is not a surrogate for CVC and cannot replace it. Complications following IO access are rare, providing correct indication, training skills and adequate handling. Therefore, a change in practice from CVC to immediate IO access for the initial emergency resuscitation should be strongly considered as a reasonable bridging technique to increase patients' safety in the emergency department, if peripheral IV access was attempted unsuccessful three times for a maximum duration of 2 min. These findings are in accordance with current guidelines of the American Heart Association and the European Resuscitation Council [10, 11]. Based on our findings, further prospective large-scale randomized trials are necessary to provide the treating physician with clear, evidence-based guidelines in the future.