In children requiring emergency vascular access, are intraosseous lines associated with significant long-term complications?
Date First Published:
April 22, 2026
Last Updated:
April 22, 2026
Report by:
Emma Carley, Medical student (University of Nottingham medical school)
Search checked by:
Gregory Yates, Academic emergency medicine resident doctor
Three-Part Question:
In [children requiring vascular access in emergency settings] are [IO lines] associated with [complications including infection, compartment syndrome, venous thrombosis, or fat embolism?]
Clinical Scenario:
You attend a cardiac arrest in an 8-year-old child. As part of the resuscitation, you place a tibial intraosseous line (IO) and administer drugs including adrenaline. ROSC is achieved and you arrange for transfer to PICU. Before the patient is moved, however, the PICU consultant requests that the IO catheter be removed, due to the risk of complications. You wonder if there is evidence to support this concern and decide to find out more.
Search Strategy:
Our inclusion criteria were as follows:
1. Children (<16 years) undergoing IO line insertion
2. Procedure undertaken by emergency medicine (EM) or pre-hospital emergency medicine (PHEM) clinician
3. Data reported on mid/long-term complications, including osteomyelitis, fat embolism, and compartment syndrome
1. Children (<16 years) undergoing IO line insertion
2. Procedure undertaken by emergency medicine (EM) or pre-hospital emergency medicine (PHEM) clinician
3. Data reported on mid/long-term complications, including osteomyelitis, fat embolism, and compartment syndrome
Search Details:
EMBASE (1980-2025) and MEDLINE (1946-2025) databases were searched using the Ovid interface and the following keyword / Medical Subject Headings (MeSH) strategy:
1. exp infusions, intraosseous/
2. intraosseous.mp
3. 1 or 2
4. osteomyelitis.mp
5. exp osteomyelitis/
6. fat embolism.mp
7. exp embolism, fat/
8. compartment syndrome.mp.
9. exp compartment syndromes/
10. 4 or 5 or 6 or 7 or 8 or 9
11. 3 and 10
Conference abstracts, case reports, and review articles were filtered out. No date or language filters were used.
The Google Scholar ‘cited by’ function was then used to find studies that had referenced the papers we identified as relevant in our database searches. Finally, the reference lists of relevant papers were screened for studies missed by our search process.
1. exp infusions, intraosseous/
2. intraosseous.mp
3. 1 or 2
4. osteomyelitis.mp
5. exp osteomyelitis/
6. fat embolism.mp
7. exp embolism, fat/
8. compartment syndrome.mp.
9. exp compartment syndromes/
10. 4 or 5 or 6 or 7 or 8 or 9
11. 3 and 10
Conference abstracts, case reports, and review articles were filtered out. No date or language filters were used.
The Google Scholar ‘cited by’ function was then used to find studies that had referenced the papers we identified as relevant in our database searches. Finally, the reference lists of relevant papers were screened for studies missed by our search process.
Outcome:
282 records were identified using our search strategy. 264 studies were excluded during title/abstract review. Sixteen studies underwent full text review and five were excluded for the following reasons: mostly included hospital inpatients (n=3), unverified survey data (n=1), IO not inserted by EM/PHEM clinicians (n=1).
Eleven papers remained for final analysis: eight retrospective and three prospective observational studies. In studies including both adult and paediatric patients, authors were contacted to provide separate datasets. One (Joerck et al., 2023) did so.
Eleven papers remained for final analysis: eight retrospective and three prospective observational studies. In studies including both adult and paediatric patients, authors were contacted to provide separate datasets. One (Joerck et al., 2023) did so.
Relevant Paper(s):
| Study Title | Patient Group | Study type (level of evidence) | Outcomes | Key results | Study Weaknesses |
|---|---|---|---|---|---|
| An assessment of long-term complications following prehospital intraosseous access: A nationwide study Petersen LB et al 05/12/2024 Denmark | 5387 prehospital patients receiving IO access. 375 were un-identified or lost to follow up. All patients received IO access in pre-hospital setting. Of the remaining 5012 cases: -237 (4.73%) children. -4775 (95.3%) adults 3034 (60.6%) males. | Retrospective observational study (Level 3b) |
Complications within 180 days | 0 (0%) children had long-term complications. | High mortality means that the potential complication rate may be under- estimated. Study relied on retrospective prehospital data and accurate ICD coding of complications, both of which may be incomplete or inaccurate. Large proportion of the population were adults (95.3%) – results may not be as reliable for paediatric population. |
| No cases of osteomyelitis, osteonecrosis or compartment syndrome reported 175 days after IO access across the 237 children. | |||||
| Integration of Intraosseous Approach Method in Georgia Kikodze et al 16/01/2024 USA | 26 patients undergoing emergency IO insertion in the ED | Prospective observational study (Level 2b) |
Incidence of IO related complications | No osteomyelitis. | Unknown follow-up period Very small sample size – likely under-powered to detect complications of IO. |
| In the hospital using IO approach, 35% of these patients continued treatment in ICU, compared to 100%, 97% and 100% of patients in hospitals that don’t use IO approach. | |||||
| No local infection from IO approach. | |||||
| Use of Intraosseous Access in Neonatal and Pediatric Retrieval-Neonatal and Pediatric Emergency Transfer Service, New South Wales Joerk C. 01/07/2023 Australia | 467 paediatric patients who received IO access between 2006-2020. 102 (21.8%) neonatal (≤44 weeks). 365 (78.2%) paediatric (>44 weeks, <16 years). 281 (60.2%) males. Common IO indications: -cardiac arrest -sepsis -respiratory distress -encephalopathy Setting: newborn and paediatric emergency transport services (NETSNSW). | Retrospective observational study (Level 3b) |
IO access related complications. | IO access related injury in 14.7% paediatric and 10.8% neonatal patients. | Retrospective nature of this review with potential measurement bias due to manual data entry Evidence of inconsistent data entry in free text fields. |
| IO related complications: | |||||
| -40 (8.6%) extravasations | |||||
| -6 (1.3%) dislodgements | |||||
| -3 (0.6%) compartment syndrome | |||||
| -1 (0.2%) complete bone penetration | |||||
| -1 (0.2%) fracture. | |||||
| Cutaneous Complications Associated With Intraosseous Access Placement Konopka E et al 08/07/2021 USA | 113 patients who received IO line placement over an 18-month period. Age range: 10 days – 94 years. | Retrospective observational study (Level 3b) |
IO access related complications. | 3 (2.7%) of patients experienced a complication. | Exact number of paediatric patients not reported – likely to be a small sample Inconsistent follow up periods for patients in this study. |
| Of these, IO insertion related problems: | |||||
| -1 cutaneous complication | |||||
| -1 compartment syndrome | |||||
| -1 needle breakage. | |||||
| Semi-automatic intraosseous device (EZ-IO) in a paediatric emergency department Mori et al 01/06/2020 Japan | 72 children in Japanese PED undergoing IO placement. Children <16 years old. Median age was 9 months. 38 (52.8%) make. Most common reasons for insertion: Cardiopulmonary arrest (54.2%), shock (22.2%), iv access failure (12.5%), seizure (9.7%) and trauma (1.4%). Outcomes following IO access: admission (59.7%), death at ED (37.5%) or transfer to another hospital (2.8%) | Retrospective observational study (Level 3b) |
Complication rate following use of EZ-IO device. | No cases of osteomyelitis or compartment syndrome. | Variable follow up rate (0-8 years from insertion). Retrospective design. Median age was 9 months – may not be as representative of the wider child age group population. Selection bias as this is a single-centre study, assessing patients from one tertiary children’s hospital. |
| 17 (17.5%) extravasations. | |||||
| 4 (4.1%) dermal abrasions. | |||||
| Observational review of paediatric intraosseous needle placement in the paediatric emergency department Pifko et al 10/11/2017 USA | 50 PED patients undergoing IO placement. Mean age of 3.95 years (median = 0.88). Reasons for IO access: 26 (52%) in cardiac arrest. | Retrospective observational study. (Level 3b) |
Extravasation, fracture, compartment syndrome, fat embolism, or | 2 extravasations following IO placement – both resolved unaided, without long-term effects. | Retrospective design. Variable/unknown follow up period. Low overall success rates following manual IO access (58%) and EZ-IO device access (69%) – may underestimate number of potential complications. Small sample size, only 50 patients across an 8.5-year period. |
| thrombosis following both manual IO and EZ-IO access. | |||||
| No other complications described. | |||||
| Intraosseous line use, complications, and outcomes among a population-based cohort of children presenting to California hospitals Hansen M et al 27/10/2011 USA | 291 children who received IO line access between 2005-2007. Ages 0-18; mean age: 1.7 years. 46% male. 239 (82%) ED patients. 52 (18%) inpatients. Common reasons for insertion: Cardiac arrest (34%). Acute respiratory failure (6%). Trauma (19%). | Retrospective observational study (Level 3b) |
IO access related complications. | 0 (0%) of paediatric patients experienced any long-term complications from IO access. | Low data availability (15%) for long-term complications. Reliance on retrospective review of potentially incomplete medical records. Inclusion of 16–18-year-olds may limit generalisability of findings to centres who do not see this age group. |
| Including: | |||||
| Compartment syndrome. | |||||
| Osteomyelitis. | |||||
| Fractures. | |||||
| Emergency intraosseous access in a helicopter emergency medical service: a retrospective study Sunde G 07/10/2010 Norway | 70 adult and paediatric emergency patients who received IO access by HEMS unit between May 2003 and April 2010. 18 (25.7%) aged 0-2 years. 0 (0%) aged 3-6 years. 5 (7.14%) aged 7-17 years. 47 (67.1%) aged 18-78 years. IO locations: 46 (59.0%) proximal tibia. 3 (3.8%) proximal humerus. 29 (37.2%) unknowns. | Retrospective observational study (Level 3b) |
Complications of IO insertion | 81% overall IO success rate. | Period of follow up not defined in this study. Retrospective nature and notes review may have missed long term complications. Includes comparison of different IO devices used across different time periods – potential temporal bias. |
| 1 case of bone fracture at insertion site. | |||||
| 1 needle dislocation. | |||||
| 1 case of extravasation. | |||||
| No medium or long term complications reported. | |||||
| 0 cases of osteomyelitis or other serious IO related complications. | |||||
| Powered intraosseous insertion provides safe and effective vascular access for pediatric emergency patients Horton et al 24/06/2008 USA | 95 patients undergoing IO in a tertiary PED. 53 (56%) males. 35 (36.8%) females. 7 unknown gender. Mean patient age was 5.5 +/- 6.1. Classification of cases: 55 (57.9%) medical, 30 (31.6%) trauma, 10 (10.5%) unclassified. 48 (50.5%) IO placements made in ED. 45 (47.4%) in prehospital setting. 2 (2.2%) in ICU. | Prospective observational study (Level 2b) |
Complication rate | No osteomyelitis reported. | Unknown follow up period. Inclusion of a small number of non-ED patients. Selection bias as this is a single-centre study, assessing patients from one tertiary children’s hospital. |
| No complications leading to risk to patient safety. | |||||
| 6 (6.3%) insertion failures. | |||||
| 4 (4.2%) complications: | |||||
| - 1 extravasation | |||||
| - 1 needle dislodged | |||||
| - 1 failure of IO device | |||||
| - 1 failure of infusion tubing | |||||
| Long-term effects on tibial growth after intraosseous infusion: a prospective, radiographic analysis Claudet et al 19/12/2003 France | 78 paediatric patients undergoing IO placement in PED. Of the 78 initial patients: 41 (52.6%) males. 27 (34.6%) females. Of the 23 final participants: 14 (60.9%) males. 9 (39.1%) females. Mean age of 23 participants: 18.6 months. Reasons for IO insertion: 12 (52.2%) hypovolemic shock, 4 (17.4%) status epilepticus, 2 (8.7%) near-drowning, 2 (8.7%) septic shock, 1 (4.35%) traffic accident, 1 (4.35%) cardiac insufficiency and 1 (4.35%) haemorrhagic shock. Of the 23 children access site: proximal tibia | Prospective observational study (Level 2b) |
Tibial growth effects post IO | No difference in growth as compared to other tibia. | High mortality means that the potential complication rate may be under- estimated. Small sample of children (23) included in the final analysis. |
| Within the initial population of 78 children: | |||||
| 1 (1.28%) osteomyelitis, 20 (25.6%) subcutaneous diffusions and 1 (1.28%) articular puncture. | |||||
| Emergency department utilization and success rates for intraosseous infusion in paediatric resuscitations Njissen-Jordan 02/01/2000 Canada | 57 children undergoing IO insertion in the emergency department. 15 excluded as IO access was performed outside ED – left 42 children in final study sample. Patient age range: 4-10 years. 42 (98%) less than 3 years old. | Retrospective observational study | Incidence of complications | 2 fractures (on in proximal tibia, one in distal femur), both in the same 10-day old infant. | Indeterminate follow up period. Retrospective observational design. Children who were discharged were deemed to have survived to hospital discharge– potential missed complication data due to loss of follow up. |
| No complications documented in patients with a successful IO placement. |
Author Commentary:
Our key results are summarised in Table 1. When inserted by emergency medicine (EM) trained clinicians, IO lines were associated with a low rate of adverse events (0-4%) of which few were severe. The two studies of highest methodological quality (Claudet et al., 2003; Horton et al., 2008) reported no complications at all, including osteomyelitis.
Overall, this review should encourage EM clinicians towards an "early IO" strategy in critically unwell children. Our data may help to remove a psychological barrier to performing this procedure, much like the use of local anaesthetic prior to drilling (Ilicki & Scholander, 2016). Additionally, our data may dissuade clinicians from the routine administration of antibiotics to children after IO insertion, given the negligible rate of osteomyelitis reported.
A randomised trial is unlikely to be conducted with this patient group, although it has been attempted in adults (Ko et al., 2024; Reades et al., 2011). Future prospective studies should therefore aim to recruit a comparator group: children who have received multiple unsuccessful attempts at intravenous cannulation when emergency access is required. We would expect the 'complications' in such a cohort to also include thrombosis and infection -- possibly at a higher rate than we have found with IO.
Overall, this review should encourage EM clinicians towards an "early IO" strategy in critically unwell children. Our data may help to remove a psychological barrier to performing this procedure, much like the use of local anaesthetic prior to drilling (Ilicki & Scholander, 2016). Additionally, our data may dissuade clinicians from the routine administration of antibiotics to children after IO insertion, given the negligible rate of osteomyelitis reported.
A randomised trial is unlikely to be conducted with this patient group, although it has been attempted in adults (Ko et al., 2024; Reades et al., 2011). Future prospective studies should therefore aim to recruit a comparator group: children who have received multiple unsuccessful attempts at intravenous cannulation when emergency access is required. We would expect the 'complications' in such a cohort to also include thrombosis and infection -- possibly at a higher rate than we have found with IO.
Bottom Line:
Severe complications are rarely reported with paediatric IO lines and ED clinicians should not be deterred from inserting or retaining them if definitive intravenous access has yet to be established.
Level of Evidence:
Level 2: Studies considered were neither 1 or 3
References:
- Petersen LB et al. An assessment of long-term complications following prehospital intraosseous access: A nationwide study
- Kikodze et al. Integration of Intraosseous Approach Method in Georgia
- Joerk C.. Use of Intraosseous Access in Neonatal and Pediatric Retrieval-Neonatal and Pediatric Emergency Transfer Service, New South Wales
- Konopka E et al. Cutaneous Complications Associated With Intraosseous Access Placement
- Mori et al. Semi-automatic intraosseous device (EZ-IO) in a paediatric emergency department
- Pifko et al. Observational review of paediatric intraosseous needle placement in the paediatric emergency department
- Hansen M et al. Intraosseous line use, complications, and outcomes among a population-based cohort of children presenting to California hospitals
- Sunde G. Emergency intraosseous access in a helicopter emergency medical service: a retrospective study
- Horton et al. Powered intraosseous insertion provides safe and effective vascular access for pediatric emergency patients
- Claudet et al. Long-term effects on tibial growth after intraosseous infusion: a prospective, radiographic analysis
- Njissen-Jordan. Emergency department utilization and success rates for intraosseous infusion in paediatric resuscitations