Factors in Pediatric Isotonic Fluid Resuscitation Efficiency
Factors in Pediatric Isotonic Fluid Resuscitation Efficiency
This trial demonstrates a significant impact of syringe size on fluid administration time in a study setting involving health care provider subjects and a non-clinical pediatric fluid resuscitation model. Our results suggest that the use of larger syringe sizes (30 mL or 60 mL) is most efficient and dissuades the use of 10 mL syringes in situations where rapid pediatric fluid resuscitation is required. While the 20 mL syringe size was not statistically inferior to the 30 and 60 mL sizes, there was a trend towards inferiority and the 20 mL group results did not statistically differ from the 10 mL group.
We had hypothesized that HCPs would objectively fatigue over the course of performing the intervention as borne out by differences in the administration times of boluses 1, 2, and 3. We were unable to confirm or refute this hypothesis due to the presence of an interaction that precluded assessment of the main effects in this analysis. There was, however, a statistically significant increase in HCP self-reported fatigue with successive fluid boluses. Examining our results graphically by syringe group, it appears as if there is a trend toward increased self-reported fatigue among providers using the 60 mL syringe size (Figure 6). This makes sense with what we know about the physics and physiology: it is physically more difficult for providers to depress the syringe plunger of fluid filled syringes of a larger diameter.
While the presence of an interaction prevented us from assessing an impact on fluid administration time by bolus number, the interaction graph is itself interesting and somewhat instructive (Figure 5). The interaction appears to have occurred because the 10 mL group seemed to "speed up" with time, while the other 3 syringe size groups appear to have slowed down with ongoing fluid resuscitation. One hypothesis that may be generated from this finding is that individuals in the 10 mL group may have become more efficient at connecting and disconnecting the syringes over the course of the intervention. Because the 10 mL group had the greatest number of syringes to connect and disconnect for each bolus, proportionally speaking, the time allocated to disconnecting and reconnecting syringes was greatest for this group. In contrast, the observation that the other syringe size groups appeared to slow down with time would fit with our a priori hypothesis of provider fatigability.
Our finding of progressive subjective fatigue among trial participants is certainly noteworthy and not previously reported in the literature. In other physically strenuous resuscitative tasks, such as the performance of chest compressions (CPR), current best practices involve frequent provider switches to avoid performance decay. We suggest that given how fatiguing rapid manual fluid administration can be, perhaps routine provider switches are warranted for this resuscitation task as well. This issue is not addressed in current resuscitation guidelines. A logical time for provider switches would be between 20 mL/kg boluses.
The finding that a number of our study participants believed that regular infusion pumps were an adequate pediatric fluid resuscitation method underscores that more education is needed for HCPs regarding optimal fluid resuscitation performance. We still encounter standard IV pumps being inappropriately utilized in the setting of shock. Such pumps provide a maximum fluid delivery rate of 999 mL/hr, which in almost all cases is insufficient to achieve ACCM benchmarks. For example, for a 15 kg child, as simulated by our model, a 20 mL/kg bolus would take 18 minutes to infuse with use of a regular IV pump. As such pumps are often the most convenient means to operationalize a fluid bolus order, it is imperative for the physicians writing such orders to be explicit regarding the intended time frame and method of administration.
There are several limitations to our trial that warrant mention. Firstly, in a real resuscitation, syringes are not neatly prepared as was the case in our trial. However, with use of the 'disconnect-reconnect' method, syringes are typically prepared quickly by one or more HCPs resulting in a similarly "limitless" supply - that often accumulates. Second, while use of a crossover design may have been preferable, with 4 intervention groups, we felt that use of this design would negatively impact on feasibility and increase the risk of participant dropout. We are satisfied that participants' characteristics appear well balanced across the groups in our study. Thirdly, no catheter dislodgement events were recorded in our trial. It is possible that features attributable to our model had a protective effect against catheter dislodgement, although this was indeed possible and occurred during pilot testing. In this context, it is notable that only 1/57 subjects in the clinical trial performed by Stoner et al. experienced a catheter failure. Finally, our trial protocol did not strictly adhere to ACCM guideline insofar as "patient" reassessments between each 20 mL/kg bolus are recommended. In our experience, these reassessments often do not slow HCPs from administering fluid where ongoing resuscitation in required and such assessments are often done concurrently.
Although our study was conducted in the non-clinical setting, we had typical health care providers perform rapid fluid administration as they would under resuscitative conditions. The model incorporated an IV catheter and so resistance to fluid flow was as it would be in the clinical setting. Further, infants and children with decompensated shock, as in our clinical vignette are typically lethargic and so patient movement may not be all that dissimilar to our model. We therefore believe that our findings can likely be cautiously extrapolated to the clinical setting.
Our conclusions, and any other optimizations to be made in rapid fluid resuscitation relate to statistically significant differences in the order of seconds to minutes. Therefore, ultimately demonstrating whether improvements in pediatric fluid resuscitation performance have an impact on patient important outcomes like morbidity and mortality may be difficult. Nonetheless, observational studies have provided the basis for current goal-directed ACCM benchmarks, and subsequent prospective studies have shown morbidity and mortality benefit with adherence to these.
Morbidity and mortality related to pediatric septic shock has dropped significantly in recent decades – owing in part to improved recognition and aggressive management, of which fluid resuscitation is currently considered a critical component. While studies such as the FEAST trial have begun to raise questions regarding the role and extent of fluid resuscitation in the treatment of septic shock, the purpose of our study was not to challenge current ACCM guidelines, for which support has recently been reaffirmed. Instead, we sought to evaluate and improve upon implementation, which is known to be problematic. Furthermore, the need to increase clarity and pragmatic instruction for health care providers regarding how best to perform fluid resuscitation is relevant to the management of all forms of non-cardiogenic shock.
Discussion
This trial demonstrates a significant impact of syringe size on fluid administration time in a study setting involving health care provider subjects and a non-clinical pediatric fluid resuscitation model. Our results suggest that the use of larger syringe sizes (30 mL or 60 mL) is most efficient and dissuades the use of 10 mL syringes in situations where rapid pediatric fluid resuscitation is required. While the 20 mL syringe size was not statistically inferior to the 30 and 60 mL sizes, there was a trend towards inferiority and the 20 mL group results did not statistically differ from the 10 mL group.
We had hypothesized that HCPs would objectively fatigue over the course of performing the intervention as borne out by differences in the administration times of boluses 1, 2, and 3. We were unable to confirm or refute this hypothesis due to the presence of an interaction that precluded assessment of the main effects in this analysis. There was, however, a statistically significant increase in HCP self-reported fatigue with successive fluid boluses. Examining our results graphically by syringe group, it appears as if there is a trend toward increased self-reported fatigue among providers using the 60 mL syringe size (Figure 6). This makes sense with what we know about the physics and physiology: it is physically more difficult for providers to depress the syringe plunger of fluid filled syringes of a larger diameter.
While the presence of an interaction prevented us from assessing an impact on fluid administration time by bolus number, the interaction graph is itself interesting and somewhat instructive (Figure 5). The interaction appears to have occurred because the 10 mL group seemed to "speed up" with time, while the other 3 syringe size groups appear to have slowed down with ongoing fluid resuscitation. One hypothesis that may be generated from this finding is that individuals in the 10 mL group may have become more efficient at connecting and disconnecting the syringes over the course of the intervention. Because the 10 mL group had the greatest number of syringes to connect and disconnect for each bolus, proportionally speaking, the time allocated to disconnecting and reconnecting syringes was greatest for this group. In contrast, the observation that the other syringe size groups appeared to slow down with time would fit with our a priori hypothesis of provider fatigability.
Our finding of progressive subjective fatigue among trial participants is certainly noteworthy and not previously reported in the literature. In other physically strenuous resuscitative tasks, such as the performance of chest compressions (CPR), current best practices involve frequent provider switches to avoid performance decay. We suggest that given how fatiguing rapid manual fluid administration can be, perhaps routine provider switches are warranted for this resuscitation task as well. This issue is not addressed in current resuscitation guidelines. A logical time for provider switches would be between 20 mL/kg boluses.
The finding that a number of our study participants believed that regular infusion pumps were an adequate pediatric fluid resuscitation method underscores that more education is needed for HCPs regarding optimal fluid resuscitation performance. We still encounter standard IV pumps being inappropriately utilized in the setting of shock. Such pumps provide a maximum fluid delivery rate of 999 mL/hr, which in almost all cases is insufficient to achieve ACCM benchmarks. For example, for a 15 kg child, as simulated by our model, a 20 mL/kg bolus would take 18 minutes to infuse with use of a regular IV pump. As such pumps are often the most convenient means to operationalize a fluid bolus order, it is imperative for the physicians writing such orders to be explicit regarding the intended time frame and method of administration.
There are several limitations to our trial that warrant mention. Firstly, in a real resuscitation, syringes are not neatly prepared as was the case in our trial. However, with use of the 'disconnect-reconnect' method, syringes are typically prepared quickly by one or more HCPs resulting in a similarly "limitless" supply - that often accumulates. Second, while use of a crossover design may have been preferable, with 4 intervention groups, we felt that use of this design would negatively impact on feasibility and increase the risk of participant dropout. We are satisfied that participants' characteristics appear well balanced across the groups in our study. Thirdly, no catheter dislodgement events were recorded in our trial. It is possible that features attributable to our model had a protective effect against catheter dislodgement, although this was indeed possible and occurred during pilot testing. In this context, it is notable that only 1/57 subjects in the clinical trial performed by Stoner et al. experienced a catheter failure. Finally, our trial protocol did not strictly adhere to ACCM guideline insofar as "patient" reassessments between each 20 mL/kg bolus are recommended. In our experience, these reassessments often do not slow HCPs from administering fluid where ongoing resuscitation in required and such assessments are often done concurrently.
Although our study was conducted in the non-clinical setting, we had typical health care providers perform rapid fluid administration as they would under resuscitative conditions. The model incorporated an IV catheter and so resistance to fluid flow was as it would be in the clinical setting. Further, infants and children with decompensated shock, as in our clinical vignette are typically lethargic and so patient movement may not be all that dissimilar to our model. We therefore believe that our findings can likely be cautiously extrapolated to the clinical setting.
Our conclusions, and any other optimizations to be made in rapid fluid resuscitation relate to statistically significant differences in the order of seconds to minutes. Therefore, ultimately demonstrating whether improvements in pediatric fluid resuscitation performance have an impact on patient important outcomes like morbidity and mortality may be difficult. Nonetheless, observational studies have provided the basis for current goal-directed ACCM benchmarks, and subsequent prospective studies have shown morbidity and mortality benefit with adherence to these.
Morbidity and mortality related to pediatric septic shock has dropped significantly in recent decades – owing in part to improved recognition and aggressive management, of which fluid resuscitation is currently considered a critical component. While studies such as the FEAST trial have begun to raise questions regarding the role and extent of fluid resuscitation in the treatment of septic shock, the purpose of our study was not to challenge current ACCM guidelines, for which support has recently been reaffirmed. Instead, we sought to evaluate and improve upon implementation, which is known to be problematic. Furthermore, the need to increase clarity and pragmatic instruction for health care providers regarding how best to perform fluid resuscitation is relevant to the management of all forms of non-cardiogenic shock.
Source...