Intraosseous (IO) Placement: Sites and Devices

Intraosseous placement is the insertion of a needle into the marrow cavity of a bone to deliver fluids, medications, or draw blood samples when traditional intravenous access is unavailable or too slow. First described in 1922, the technique has become standard practice in emergency resuscitation, taught across Advanced Cardiac Life Support, Pediatric Advanced Life Support, and military trauma programs. The procedure works because the bone marrow contains a dense network of tiny blood vessels that drain directly into the central circulation, essentially turning a bone into a non-collapsible vein that stays accessible even during cardiac arrest or severe shock.

Where the Needle Goes

Three anatomic regions account for the vast majority of IO insertions: the proximal tibia (the flat area just below the kneecap), the proximal humerus (the upper arm near the shoulder), and the sternum (the breastbone). Each has different advantages depending on the patient’s age, body size, and clinical scenario.

The proximal tibia is by far the most commonly used site. In a combat trauma registry analysis, roughly three-quarters of all IO insertions were placed in the proximal tibia.¹ Anatomical studies in adults show that the thinnest cortical bone and largest marrow cavity sit about half a centimeter below the tibial tuberosity, on the flat medial surface of the shin. That spot offers the easiest puncture and the most room inside the bone for fluid to flow.² In neonates, the anatomy is tighter but the recommendation is similar: at least 10 mm below the tibial tuberosity, where the cortex is thinnest (around 1.3 mm) and the medullary space is widest (around 4.5 mm).³ Going too high risks piercing the growth plate or entering the knee joint, and going too low runs into thick cortical bone that resists needle entry.⁴

The proximal humerus has gained popularity, particularly in military and prehospital settings. A study comparing humeral IO catheter placement to standard peripheral IV and central venous catheterization found that humeral IO was significantly faster: about 1.5 minutes on average, compared with roughly 3.6 minutes for a peripheral IV and over 15 minutes for a central line.⁵ For trauma patients wearing body armor or those whose legs are trapped, the humerus can be the only accessible site. Combat medics have increasingly adopted it alongside the sternum and tibia as a frontline option.⁶

The sternum, specifically the manubrium (the thick upper portion), is the third major site. It has an anatomic advantage the other two lack: blood from the sternal marrow flows into the internal thoracic vein, which empties directly into the subclavian vein and central circulation. The humeral and tibial routes drain through more peripheral veins before reaching the heart.⁷ That directness translates into faster drug delivery to the heart in time-critical situations.

How Flow Rates Differ by Site

Not all bones push fluid at the same speed. In a cadaver study comparing high-pressure infusions at all three sites, the sternum delivered fluid at about 94 mL per minute, the humerus at about 57 mL per minute, and the tibia at roughly 31 mL per minute over a five-minute bolus.⁸ Those differences matter when you need to push large volumes fast, such as during hemorrhagic shock.

Applying external pressure with a pressure bag or syringe can increase flow, but the effect varies by site and device. One study found that a sternal IO device showed a strong, reliable increase in flow rate as external pressure rose, while the tibial route responded only weakly to added pressure. The humerus fell somewhere in between, responding moderately after several minutes of sustained pressure.⁹ In animal models of hemorrhagic shock, a manual push-pull syringe technique outperformed both pressure bags and rapid infuser devices, achieving rates close to 97 mL per minute compared to roughly 60-73 mL per minute for the mechanical alternatives.¹⁰

The practical takeaway: if rapid large-volume resuscitation is the priority and the sternum is accessible, it will deliver fluid fastest. When the tibia is the only available site, adding pressure helps only modestly, so a manual push-pull technique or switching to a higher-flow site when possible may be more effective.

Drill-Powered Devices Versus Manual Needles

The shift from manual bone marrow needles to battery-powered drill devices has been one of the biggest practical advances in IO access. In a swine model tested by resident physicians, a drill-assisted IO needle reached the marrow space in under 4 seconds, while manual insertion averaged over 33 seconds. The drill achieved a 100% success rate versus about 76% for the manual needle, and the most common manual complication was a bent needle, which happened in a third of attempts.¹¹

When emergency technicians tested the EZ-IO (a widely used powered drill device) against a manual needle in an adult cadaver model, insertion times were comparable at about 32-33 seconds, but the EZ-IO’s first-attempt success rate was significantly higher: about 98% versus 80%.¹² The gap widens dramatically in adult simulations that include chest compressions. One study found that using a manual bone marrow needle during CPR, only 3 of 22 participants could successfully maintain an IO route, while all participants succeeded with a mechanical device.¹³ In pediatric and infant simulations, both manual and drill needles achieved 100% success, reflecting the thinner and softer cortical bone in children.

These findings explain why powered IO drills have largely replaced manual needles in prehospital and emergency department protocols for adults. Manual needles still perform well in very young patients, but the drill offers a meaningful safety margin in adults, where thicker cortical bone and the chaotic environment of a resuscitation combine to make manual insertion less reliable.

Drugs Delivered Through Bone Work Like IV Drugs

A natural concern about infusing medications through bone rather than directly into a vein is whether the drugs arrive at the same concentration and speed. The evidence is reassuring. An early comparison study found that the IO route delivered emergency drugs including epinephrine, sodium bicarbonate, lidocaine, calcium chloride, dextrose, and hydroxyethyl starch with peak effect magnitudes and durations equivalent to both central and peripheral IV administration.¹⁴

More targeted pharmacokinetic work has confirmed this pattern for individual drugs. One study found that morphine administered through an IO line reached equivalent blood levels as IV morphine, supporting bioequivalence for pain management.¹⁵ Another demonstrated the same bioequivalence for tranexamic acid, a clotting drug increasingly used in trauma care.¹⁶ In practice, virtually any medication or fluid that can go through an IV line can be pushed through an IO line, making it a genuine alternative rather than a limited backup.

Using IO Blood Samples for Lab Analysis

IO lines can also be used to draw blood for basic lab testing when no other sample source is available. The agreement between IO and arterial blood samples is good for several common values, including glucose, lactate, and pH.¹⁷ Carbon dioxide levels run moderately higher and oxygen levels lower in IO samples compared with arterial blood, which makes sense: the marrow functions more like a venous compartment. Oxygen values can still help rule out severe hypoxia when pulse oximetry is unavailable.

Some values, however, are unreliable from an IO draw. Potassium runs consistently high in IO samples, and hemoglobin and hematocrit show too much variability to trust. Sodium and ionized calcium are slightly low but close enough to detect dangerous extremes.¹⁸ An animal study confirmed that while systematic differences exist between IO and arterial values for most variables, the differences trend in predictable directions, which allows experienced clinicians to interpret results with appropriate adjustment.¹⁹

In children, IO samples analyzed on a bedside cartridge-based device showed clinically acceptable agreement with venous blood for pH, base excess, sodium, ionized calcium, and glucose.²⁰ A practical benefit of these point-of-care cartridge analyzers is that they avoid the problem of bone marrow fragments clogging conventional laboratory equipment. Still, IO lab samples should be treated as a last resort, useful when no arterial or venous sample can be obtained, and interpreted cautiously rather than taken at face value.

Complications and How to Avoid Them

IO placement is remarkably safe for an invasive procedure, but complications do occur. Fluid extravasation, where infused fluid leaks out of the bone into surrounding soft tissue, is the most common problem, reported in roughly 12% of cases.²¹ Most extravasation is minor and results only in local swelling. In rare cases, however, it can escalate to compartment syndrome, a dangerous buildup of pressure inside a limb’s muscular compartments that threatens blood flow and requires emergency surgical release.

One published case involved a 30-year-old woman whose tibial IO catheter was inadvertently driven through both the front and back walls of the tibia, creating a direct path for blood products to leak into surrounding tissue. Imaging confirmed the needle had exited the far cortex. She developed acute compartment syndrome that required fasciotomy of all four compartments of her lower leg.²² Earlier reports documented similar outcomes in children, including a sudden infant death syndrome patient who developed severe tissue necrosis and a near-drowning victim who required fasciotomy after IO-related compartment syndrome.²³

Osteomyelitis, a bone infection, is feared but rare, occurring in fewer than 1% of IO placements.²⁴ Risk goes up when the IO line is left in place for extended periods or when sterile technique is compromised during a chaotic resuscitation. Guidelines generally recommend removing IO catheters as soon as alternative access is established, typically within 24 hours.

Most complications trace back to technique rather than inherent risk. Driving the needle too deep (through the far cortex), inserting into a fractured bone, or failing to recognize extravasation early are the preventable mistakes that lead to serious harm. Confirming correct placement by feeling the characteristic loss of resistance, checking that the needle stands upright without support, and watching for tissue swelling during initial flush are basic safeguards that catch most misplacements before they become dangerous.²⁵

IO Versus IV in Cardiac Arrest

Whether IO access produces outcomes as good as IV access during cardiac arrest has been hotly debated. A large randomized trial in Scandinavia assigned patients in out-of-hospital cardiac arrest to IO-first or IV-first strategies. Sustained return of spontaneous circulation occurred in about 30% of the IO group and 29% of the IV group. At 30 days, survival was 12% versus 10%, and favorable neurological outcomes were 9% versus 8%, with none of these differences reaching statistical significance.²⁶ This trial, published in the New England Journal of Medicine, was the strongest single piece of evidence suggesting IO and IV are roughly equivalent when each is used as the first-line approach.

A systematic review and meta-analysis pooling data from three randomized trials with over 9,000 participants found similar results: no meaningful difference in 30-day survival or neurological outcomes. The one wrinkle was that IO access was associated with slightly lower odds of achieving sustained return of spontaneous circulation, though the effect was small.²⁷

A broader meta-analysis that included observational studies alongside trials painted a somewhat less favorable picture for IO. When 17 studies were pooled, IO access was associated with lower odds of return of spontaneous circulation and less favorable neurological outcomes at hospital discharge.²⁸ The difference between these two analyses likely reflects selection bias in the observational data: patients who get IO access in real-world practice tend to be sicker or harder to access, which drags their outcomes down regardless of the route. The randomized trial data, which controls for this by randomly assigning the route, is more trustworthy for judging the IO route itself.

The practical conclusion that most emergency medicine guidelines have landed on: IO access is an acceptable first-line alternative when IV access is difficult or time-consuming, and should not be withheld out of concern that it produces worse outcomes. If a skilled provider can obtain IV access quickly, that remains preferred, but seconds lost struggling with a peripheral IV in a pulseless patient are not worth the delay.

Special Considerations in Neonates

Neonates and especially premature infants pose a unique challenge for IO placement because their bones are tiny. The standard recommendation of inserting at least 10 mm below the tibial tuberosity, which works well in older children and adults, does not translate reliably to the smallest patients. An ultrasound-based study of term and preterm neonates found that inserting 10 mm below the tibial tuberosity violated the safety distance to the growth plate in over half of proximal tibial attempts and 85% of distal tibial attempts.²⁹ Even when experienced pediatricians chose the site by feel, they violated the safety margin about a third of the time.

Ultrasound guidance identified a safe insertion point in cases where both the standard measurement and clinical judgment fell short. This finding suggests that for very small neonates, especially those born prematurely, point-of-care ultrasound should be considered whenever time permits to confirm that the needle path avoids the growth plate and joint space.

Pain Management During IO Infusion

IO insertion itself is often performed on unconscious or critically ill patients, so pain during the needle puncture is rarely an issue. The real pain problem comes during infusion: conscious patients frequently report severe pain when fluid is pushed under pressure into the marrow cavity. The pressure distends the rigid bone, stimulating nociceptors in the periosteum and marrow space in a way that standard IV infusions never do.

Lidocaine injected slowly through the IO line before starting fluid infusion can reduce this pain.³⁰ The typical approach involves injecting a small bolus of 2% preservative-free lidocaine, allowing it to dwell in the marrow for about a minute, and then beginning the infusion. This step is easily forgotten during a high-pressure resuscitation, but it makes a meaningful difference for any patient who is awake enough to feel it. Given that IO access is increasingly used in battlefield and disaster settings where patients may be alert despite needing emergency vascular access, lidocaine pretreatment is worth building into standard IO protocols.

Military and Prehospital Use

IO access has become a cornerstone of tactical combat casualty care. When a combat medic is treating hemorrhagic shock in the field, peripheral veins are often collapsed, extremities may be wounded or tourniqueted, and conditions make sterile central line placement impossible. A retrospective registry analysis of prehospital combat trauma found that IO insertion succeeded on the first attempt in the large majority of cases, with an overall success rate of 85% and no major acute complications documented.³¹

The tactical evolution of IO access has moved from the tibia (still the most common site) toward increasing use of the humeral head and sternum, partly because body armor and protective equipment can make the tibial site harder to reach in the field.³² Spring-loaded injection devices and drill devices are compact enough to fit in a medic’s kit, and their high first-attempt success rate means less time spent on vascular access in an environment where speed is survival.

Training Makes a Measurable Difference

IO placement is considered a procedural skill with a relatively short learning curve, but structured training still improves outcomes. A simulation-based program for pediatric residents found that procedural skills scores jumped from baseline to near-maximum after a hands-on workshop, and knowledge scores improved significantly as well.³³ Another study tracking residents after simulation training found that confidence in IO placement and knowledge of indications and anatomical landmarks remained significantly elevated at a two-month follow-up, with competency assessments above 91% across all training levels.³⁴

Skill decay is a concern for any procedure performed infrequently. A neonatal simulation study showed that while procedural scores improved after training, they can drift downward over time without reinforcement.³⁵ Periodic refresher sessions, even brief ones using low-fidelity simulation models, help maintain the muscle memory and landmark recognition that make the difference between a smooth first-attempt insertion and a failed or misplaced one.

IO Access in Veterinary Medicine

The same principles that make IO access valuable in human emergencies apply to animals. Small animal patients, particularly cats and small dogs, often present the same challenge as pediatric patients: tiny, fragile veins that collapse during shock. A study comparing three IO insertion methods in cat cadavers found that all three achieved acceptable results, with a spring-loaded device performing significantly faster and with greater ease of use than manual approaches. Both the tibia and humerus worked equally well as insertion sites.³⁶

In dogs, a spring-loaded bone injection gun achieved IO access in about 22 seconds compared to 42 seconds for a manual needle, with comparable success rates of about 83% and 96%, respectively.³⁷ Veterinary emergency protocols now routinely include IO access for critically ill animals, using the same sites (proximal tibia, proximal humerus, and occasionally the femur) and following the same confirmation steps: loss of resistance on entry, stable needle position, and clean saline flush without soft tissue swelling.