Intraosseous (IO) injection of medications (as opposed to intravenous IV administration) has been used in recent years, primarily for antibiotic prophylaxis in total knee arthroplasty (TKA), with the aim of reducing the rate of periprosthetic joint infection (PJI). The study by Pekas et al. explored IO injection of morphine for perioperative pain management and demonstrated no difference in the effectiveness of pain management when adding IO morphine. Upon IO injection of a medication into the proximal tibia, the medullary cavity, as a non-collapsible venous plexus, rapidly drains the medication into the venous circulation. Therefore, medication administered intraosseously behaves similarly to that administered intravenously, albeit achieving >10 times the local subcutaneous tissue and osseous concentrations compared with IV application in patients with a tourniquet1. In contrast, without use of a tourniquet, IO injection has shown no significant differences in the average vancomycin concentrations in the distal femur, proximal tibia, or suprapatellar synovial tissue compared with IV administration2. Numerous studies have shown a significant difference in the rate of PJI associated with the use of IO antibiotics, although some patients in the IO study groups received a combination of IV and IO administration3,4. In the study by Pekas et al., all patients received a combination of 1,000 mg of acetaminophen, 15 mg of ketorolac, and 50 or 75 mg of pregabalin prior to surgery. The peak effect of acetaminophen, ketorolac, and pregabalin is expected to be around 1 to 2 hours after administration, and the duration of the analgesic effect is an estimated 6 to 8 hours. All patients underwent surgery with spinal anesthesia consisting of 0.5% isobaric bupivacaine (4 to 8 mg) with or without fentanyl (25 to 50 µg), which is intended to control pain for 3 to 4 hours after surgery. In addition, the surgeon performed an adductor canal block containing 15 mL of 0.5% ropivacaine, which augments pain control for approximately 10 to 14 hours. Both groups received these pain management modalities, which would have provided effective pain control for at least 4 hours after the initiation of the spinal anesthesia. The study then asked whether adding an IO injection of 10 mg of morphine in 110 mL of normal saline solution at the time of surgery would influence visual analog scale (VAS) pain scores at any point from the post-anesthesia care unit (PACU) time period to 14 days postoperatively. The study found no difference in pain or functional outcome between the treatment groups. To understand the lack of an effect on pain management, one might have to evaluate the pharmacokinetics of morphine. IV morphine has a half-life time of up to 2 hours and an average analgesic effect duration of approximately 3 to 4 hours. When morphine is given intraoperatively at the time of the adductor canal block and spinal anesthesia, its analgesic effect likely does not extend beyond the pain control provided by the other 2 modalities. In my mind, there is a considerable difference between perioperative PJI prophylaxis and perioperative pain control. Perioperative PJI prophylaxis aims to minimize bacterial contamination of the surgical field at the time of surgery. Achieving significantly higher concentrations of perioperative antibiotics by applying them intraosseously and containing them in the surgical field with use of a tourniquet appears to be highly effective in reducing the long-term risk of PJI3,4. In contrast, pain control in TKA requires a more long-term effect. Preoperative medication, as received by both groups in the Pekas et al. study, blocks the sensitization before it starts by providing preemptive analgesia (blocking nociceptive input before incision) and preventive analgesia (reducing sensitization throughout the perioperative period). Spinal anesthesia and the adductor canal block provide adequate perioperative pain control. Surgeons are most interested, however, in pain control after the patient has been weaned off spinal anesthesia. Given the relatively short half-life time of IV morphine, it is pharmacokinetically unlikely that morphine would provide additional pain relief when administered at the time of spinal anesthesia. It is therefore not surprising that almost no literature exists regarding an IV morphine bolus during spinal anesthesia. The effect of IV morphine is too short to have any meaningful benefit when combined with adequate spinal anesthesia, and its effect is gone by the time the spinal anesthesia and the block wear off. In light of the pharmacokinetics of IV morphine, it is unlikely that the study by Pekas et al. would have shown a positive effect of IO morphine. This has nothing to do with the fact that IO application in the tibia can achieve high local tissue levels around the knee; the lack of an effect is simply the result of morphine’s short half-life time, and its use in combination with spinal anesthesia is unlikely to have a positive impact on pain control from the PACU to 14 days after surgery.
Friedrich Boettner (2026) studied this question.