Time delay and risk of toxicity of intraosseous anaesthesia use for awake intraosseous access in children

Intraosseous access (IO) is a crucial, life-saving alternative vascular access in paediatric emergency medicine. The technique is reasonably safe and has high success rates of up to 98.3%. IO is the quickest vascular access method in small children when peripheral venous access is difficult to obtain, or the provider has little experience [1‐4].

The pain level of drilling the needle into the bone is reported to be comparable to that of a peripheral or central venous access placement. In contrast, the pain caused by infusion (i.e. the first flush of the medullary cavity) is reported to be significantly higher. Furthermore, in recent years, the use of anaesthetic agents to reduce the perceived endostal pain caused by applied volume or drugs has been recommended in guidelines and procedure manuals. Therefore, it is understandable that some providers are worried about pain caused by intraosseous access and tend to consider some form of analgesia, be it local, systemic or intraosseous [3‐8].

Due to the relatively low maximum dose of lidocaine in small children (of 5 mg/kg) with a narrow therapeutic range and the rare use in paediatric emergency medicine, the risk of dosing error is significantly high. In recent years, several adverse events due to medication errors with local anaesthetics, of which some had lethal outcomes, were observed. In Germany, this led the BfArM (German Federal Institute for Drugs and Medical Devices) to release a statement advising against intraosseous lidocaine use in paediatric IO placement [9‐11].

Data were collected and analysed as described in Pfeiffer et al. and Schwindt et al. [2, 12]: In brief, our prospective surveillance study collected all patients aged > 28 days to 18 years who received one or more intraosseous access attempt(s) in and out of the hospital setting in Germany from 1 July 2017 to 30 June 2019 via the reporting mechanism of the German Paediatric Surveillance Unit (GPSU).

Our prospective nationwide surveillance study in Germany analysed 234 children, aged > 28 days to 18 years, who received one or more intraosseous access attempt(s) in the context of a GPSU survey from July 2017 to June 2019. Of those, we excluded 160 patients who received cardiopulmonary resuscitation (CPR) or were unconscious during IO access attempts. This method identified 74 cases in which the patient received intraosseous access while awake. All patients were younger than 6 years old.

Almost every third child (31.6%) is awake during intraosseous access use. In 18.9% (n = 14) of all awake patients, an intraosseous anaesthetic was used before the IO was drilled or the marrow cavity was flushed. In our study, this leaves close to every fifth patient at risk of potentially life-threatening complications from lidocaine dosing errors and consecutive toxicity. In all 14 patients, peripheral venous access was attempted before IO was established (see Table 1).

Furthermore, intraosseous anaesthetics led to a significant time delay (Kruskal–Wallis test p = 0.001). It took, on average, more than 3 min longer to establish intraosseous access when intraosseous anaesthetics were used (median < 3 min vs > 6 min) (see Fig. 1). All 14 patients were younger than 6 years and survived to discharge. In addition, only 1 out of 14 provider’s institutions had a guideline or standardised operating procedure (SOP) for IO use. In all, 71.4% of providers received some training before IO use.

Intraosseous anaesthesia prolongs the establishment of working vascular access in an emergency. Furthermore, intraosseous lidocaine use in small children provides an inherent risk of drug toxicity. To prevent adverse events, particular emphasis must be placed on placement without intraosseous anaesthesia, and alternative pain management (intranasal) must be considered if necessary, for example, esketamine (2 mg/kg).

Furthermore, the topic should be addressed in guidelines and training courses dealing with vascular access establishment in critically ill children with a focus on alternative pain management strategies. Standard operating procedures should reflect the advised current practice. Further research is necessary to evaluate possible changes in local anaesthesia usage rates after the BfArM statement [9, 13].

Three patients in our study cohort received an intraosseous access during a seizure/status epilepticus. Meaning they were likely unconscious at the time of puncture and did not need any form of analgesia, especially, for the pain of the application of antiepileptic drugs with small volumes. Considering all these implications of lidocaine use, we propose a practical stepwise approach: 1) If a patient is awake and the physician is concerned about the pain of placement, re-evaluate the indication for IO use—is it a life-threatening situation? Are fluids or life-saving medications only available intravenously or intraosseously necessary to stabilise the patient? 2) If IO is deemed necessary and it is a life-threatening, urgent situation (in which the consciousness of the critically ill patient is likely impaired), potential pain is secondary to the primary goal of administering life-saving therapy. If analgesia is necessary and the clinical situation of the patient allows it, alternative anaesthetic methods should be used to administer a safer form of analgesia (e.g. intranasally) before IO is attempted (two-stage approach) [13, 14].

For example, the intranasal application of esketamine (2 mg/kg) or fentanyl (1.5–2 µg/kg). There is sufficient evidence in the literature supporting the use of ketamine and fentanyl with a sound safety profile. Multiple studies report a similar pain reduction efficacy of fentanyl and ketamine when used intranasally. Prescott et al. (2023) reports similar pain relief to intramuscular morphine and intravenous fentanyl. Minor transient adverse events (e.g. dizziness, sleepiness, nausea, bad taste in mouth) occur more frequently when using subdissociative dose ketamine. No severe adverse events were reported in Reynolds et al. (2017) and Graudins et al. (2013). Intranasal ketamine causes more sedation compared to intranasal fentanyl. Intranasal subdissociative dose ketamine should be preferred when an increased sedational component is necessary, or when there are contraindications for opioids. There is insufficient literature on the direct comparison of intranasal analgesia compared to intraosseous lidocaine in paediatric patients—and, with respect to the more recent warning letters and position papers, there hopefully never will be, as the use of lidocaine as anaesthetic for intraosseous access should be considered obsolete [14‐20].

Intraosseous anaesthesia prolongs the establishment of working vascular access in an emergency. Furthermore, intraosseous lidocaine use in small children provides an inherent risk of drug toxicity. To prevent adverse events, particular emphasis must be placed on placement without intraosseous anaesthesia, and alternative pain management (intranasal) must be considered if necessary. Furthermore, the topic should be addressed in guidelines and training courses dealing with vascular access establishment in critically ill children. Standard operating procedures should reflect the advised current practice. Further research is necessary to evaluate possible changes in local anaesthesia usage rates after the BfArM (German Federal Institute for Drugs and Medical Devices) statement.

This study has multiple limitations that apply. These are similar to the limitations in Pfeiffer et al. (2023) and Schwindt et al. (2022). In short, GPSU’s voluntary reporting mechanism, using only GPSU-reporting hospitals, introduces a significant underreporting and selection bias. Our study method might miss cases from nonreporting institutions, which could mean that certain practices or adverse events are underrepresented in our cohort. Due to our stringent exclusion criteria (CPR or unconscious) and the resulting (relatively) small study cohort, the broader applicability of our findings is limited. Since the use of intraosseous access in paediatric emergencies is a very rare event, further large registry studies (potentially in cohorts with mandatory instead of voluntary reporting) are needed to systematically evaluate larger cohorts of awake intraosseous access use in children. Larger cohorts would allow a more representative assessment of the frequency of use, the time difference and better control for confounding variables. The underrepresentation of older paediatric patients (> 6 years of age) further limits the generalisability of our results, especially for those older patients, who might possess different pain tolerances or requirements for anaesthetic management. Retrospective cognitive reconstruction of events, especially relating to information about procedure times (e.g., time to first IO access), is known to be inconsistent [2, 12].

Bild vergrößern