Rozmiar 98 KB
The use of intraosseous devices during cardiopulmonary resuscitation: Is this the answer for which we have been searching?
UC San Diego Emergency Medicine, 200 West Arbor Drive, #8676, San Diego, CA 92103, United States
This issue of Resuscitation contains several outstanding articles addressing the use of intraosseous (IO) devices for vascular access during resuscitation of adults.1, 2, 3 The volume and quality of investigations and analyses concerning adult IO in recent years has been extraordinary, with the current works as outstanding examples. Those advocating the use of IO have focused on the technical advantages, such as ease of insertion and time to drug administration, while citing generally favourable pharmacodynamic comparisons with peripheral intravenous (IV) and central venous routes.4, 5, 6 When considering the impact of these efforts, it is worth stepping back to consider their broader context as well as the limitations of our current understanding regarding optimal resuscitation.
From its inception, resuscitation science has been challenged by the use of surrogate markers as research outcomes. This is understandable, given the difficulties in retrieving definitive outcome measures and the sequential nature of resuscitation, with each intervention somewhat dependent on those that come before and after to achieve a successful outcome. Unfortunately, the over-reliance on surrogate markers and the acceptance of unproven assumptions has repeatedly led us astray. For example, the observation that elevated intracranial pressure (ICP) is a poor prognostic indicator following traumatic brain injury resulted in the routine use of hyperventilation as a treatment. Only later did we recognise that the resultant ischaemia worsened outcomes via hypocapnoeic vasoconstriction.7, 8 Similarly, traumatic hypotension is associated with high mortality, but rapidly restoring normal haemodynamics with aggressive fluid therapy may worsen outcome by increasing hemorrhage and promoting reperfusion injury.9, 10
The current model of cardiopulmonary resuscitation defines increasing perfusion as a core objective. The most visible manifestation of this focus has been the emphasis on high quality chest compressions with minimal interruptions as a primary means to maintain coronary perfusion pressure (CPP) and optimise the potential for return of spontaneous circulation (ROSC). Administration of vasopressor drugs, such as adrenaline (epinephrine) or vasopressin, is recommended to increase mean arterial pressure (MAP) and further raise CPP.11, 12 Unfortunately, clinical evidence supporting this approach is lacking. While animal models suggest benefit with intra-arrest administration of vasopressors, most clinical investigations have failed to demonstrate their effectiveness, whether in combination or compared to placebo.13, 14, 15, 16, 17, 18 One randomised, placebo-controlled trial has shown increased ROSC rates, but not long-term survival, with adrenaline.19 This apparent inconsistency between animal and human studies may reflect one of the following: methodological issues, such as a lack of statistical power, that obscure a true benefit of vasopressor use in human studies (Type II error); inherent physiological differences between animals and humans, including the presence of comorbidities and underlying vascular disease that may alter arrest pathophysiology; or important differences in the therapeutic approach used in animal versus human studies. This last explanation is worth exploring to better understand the implications of the IO device studies reported here.
One of the most important differences between experimental and clinical cardiac arrest concerns the quality of CPR performed during the resuscitation. The use of mechanical devices that produce consistent chest compressions and the elimination of CPR interruptions during animal models may not accurately reflect `real life' CPR, either in or out of the hospital.20, 21 Such differences could dramatically influence the effectiveness of vasopressors, since their primary action appears to be the augmentation of CPP during the performance of high quality chest compressions. Similarly, the presence of an advanced airway in the animal arrest models enables continuous compressions with asynchronous ventilations from the onset of CPR. The timing of vasopressor administration may also be an important consideration. In animal models drugs are generally given soon after CPR is initiated. However, tiered emergency medical service (EMS) systems or the presence of a single advanced life support (ALS) practitioner may result in delays to vascular access and vasopressor administration. Indeed, a 10-min EMS response followed by 15 additional minutes to administration of adrenaline or vasopressin may limit the effectiveness of these drugs, particularly when preceded by mediocre CPR that could amplify rather than reverse the ischaemic insult.
These factors may be critical in defining the role of vasopressors and route of vascular access in human cardiac arrest. Leidel et al. conducted an elegant study that randomised patients requiring resuscitation to undergo central venous versus IO placement as a salvage approach following failed IV catheter insertion.2 Not surprisingly, IO insertion was achieved 6
min faster than central venous access. It is understandable that providers were given up to 3 attempts or 2
min for IV attempts, as both central venous and IO routes are considered more invasive and most subjects were not in cardiac arrest. However, one could argue that all three approaches should be initiated simultaneously, particularly with cardiac arrest, and that medications should be administered via the first successful route.
It is notable that their study design prioritised the proximal humerus site for IO attempts and used commercially available mechanical devices to assist insertion.2 While IO insertion has been considered homogeneously, without differentiating between insertion site and device, these considerations may be important with regard to both technical and pharmacodynamic issues. As the methodological review by Weiser et al. suggests, the use of electrical insertion devices results in higher success rates, greater user satisfaction, and faster insertion times when compared with other mechanical or manual insertion techniques.1
The site of insertion may be extremely important in the interpretation of IO literature and defining the optimal approach to injecting intra-arrest medication. As data from Hoskins et al. suggest, tibial IO administration of tracer dye results in delays to peak blood levels of almost a minute and only 65% of drug absorption as compared to administration through a sternal IO device.3 In fact, pharmacodynamic profiles were similar when the authors compared central venous and sternal IO administration. Thus, it is clear that not all approaches to IO insertion are equivalent, with device and site variability in success rates, time to insertion, and time to peak drug concentration. Even less clear is the impact of IO insertion or other approaches to vascular access on the performance of other resuscitation techniques, such as chest compressions, airway management, rhythm analysis, and defibrillation. It is not difficult to imagine that diverting the resuscitation team leader's attention toward vascular access could adversely affect the performance of skills with equal or greater relative importance.
Lastly, it is worth noting that the optimal pharmacodynamic approach to intra-arrest medication administration remains unclear, both experimentally and clinically. Whether achieving high peak levels is preferable over strategies in which a lower drug concentration is sustained over a longer period of time is not known. It is also possible that the delay to peak concentration and lower absorption associated with tibial IO administration of medications can be overcome by higher doses. Such considerations are clearly important with administration of antibiotic therapy but have not been explored adequately in cardiac arrest. Furthermore, the relative advantages of speed and high success with peripheral IO insertion versus the pharmacodynamic profile of more central approaches are completely unknown.
In summary, a growing body of literature suggests IO as a viable alternative to IV or central venous injection during resuscitation. The pharmacodynamics of IO medication administration suggest that delays to peak serum drug levels, lower peak drug concentrations, and an overall decrease in total serum drug are predictable, although the clinical implications are unclear. This may warrant an increase in IO drug dosing, and the faster time to vascular access and medication administration with use of IO devices may offset these pharmacodynamic issues. It is also important to acknowledge the heterogeneity across various IO devices and insertion sites in respect of insertion times, overall success rates, and pharmacodynamic profiles. Most important, however, is the recognition that our assumptions about optimal vasopressor dosing and the general importance of medications during cardiopulmonary resuscitation may not be valid and require additional investigation.
Rozmiar 98 KB