ICU Liberation Bundles and The Legend of Three Arrows*

In Japan, there is an old legend called "The Three Arrows": An ageing Samurai passed each of his three sons an arrow and asked them to snap it. They did so easily. He then handed them each three arrows and asked his sons to snap them. They could not. The Samurai explained that one arrow can be broken easily, but three arrows held together cannot. In critical care, applying multiple evidence-based interventions together may be more effective than applying each one individually. They may be stronger together than each on their own. The Society of Critical Care Medicine developed an ABCEDF bundle (Assessing Pain, Both Spontaneous Awakening and Breathing Trials, Choice of Sedation, Delirium Monitoring/Management, Early Exercise/Mobility, and Family Engagement/Empowerment) (1) to facilitate liberation from ICU-associated interventions in order to decrease ICU-acquired morbidities (2). In adult critical care, it has been reported that the application of the ABCDEF bundle is significantly associated with decreases in mechanical ventilation (MV) duration, coma, delirium, restraint, ICU readmission, ICU length of stay (LOS), and mortality (3). However, the efficacy in pediatric critical care has yet to be elucidated. In this issue of Pediatric Critical Care Medicine, Lin et al (4) present a prospective, multicenter, cohort study to assess clinical outcomes following an application of a Pediatric ABCDEF Bundle (5). They assessed compliance of bundle utilization in children greater than 2 months with expected PICU stay greater than 2 days and need for mechanical ventilation. They were unable to demonstrate an association between bundle utilization and MV duration, PICU LOS, or the incidence of delirium. After adjustment, they unexpectedly found a potential relationship between the subject-specific or day-specific utilization of the bundles and altered probability of death. The results of this study provide us with several points of reflection regarding the application of the ABCDEF bundle into pediatric critical care practice. In line with previous international multicenter, cross-sectional surveys of unit-based practice (6), the full compliance of the bundle was still as low as 64%, even after interventions for practice changes, specifically due to the low adaptations for the B, D, and E components. Although the intervention has shown statistically significant increases in utilization rates (0–13%, 22–61%, 50–79% in B, D, E, respectively), those were not sufficient to achieve high overall compliance. The results suggest fundamental difficulties in applying the A–F bundle, specifically for the least commonly practiced but important components of B, D, and E in pediatric settings. What should we learn from the results and where should we direct further efforts in this regard? One approach would be promoting quality improvement programs. Identifying barriers to implementation at a patient, clinician, protocol, or institutional level might help improve compliance (7). This approach; however, is time- and resource-consuming. Another arguably more practical approach would be to develop and validate a modified pediatric ABCDEF bundle (8). As suggested previously, the approach to element B is different between adults and children. Spontaneous waking trials (daily sedation interruption) in combination with spontaneous breathing trials could be discouraged because it is not recommended for use in the PICU (9). Use of nurse-driven analgesia and sedation titration, with or without protocols (10), to be included in integrated elements A and C may be sufficient to titrate sedative and analgesic use, leading to earlier extubation. Furthermore, the addition of assessment and interventions for iatrogenic withdrawal syndrome in addition or instead of element D could also be worth considering (8). The authors observed an associated reduction in the probability of death, rather than actual deaths, which was not associated with corresponding changes in MV duration, ICU LOS, and delirium incidence. It is difficult to imagine that the significant but small changes in bundle compliance could directly affect mortality and the underlying reason for this observation requires further study. It could be due to chance alone or perhaps related to a form of the legendary Hawthorne Effect, where the attempt at quality improvement leads to improved outcomes irrespective of specific bundle elements. Regardless, the authors have conducted an important study that aimed to improve the outcomes of critically ill children. Further research can be directed in refining specific elements of the bundle and better characterizing the relationship between the ABCDEF bundle application and outcomes in PICUs.