Post-intensive care syndrome (PICS) is the development of new or worsening psychological, neurologic, and/or functional morbidities arising after critical illness. Primarily researched in adults, an increasing body of literature describes these deficits in children. One such study demonstrated that approximately 10% of critically ill children admitted to a pediatric intensive care unit (PICU) had an unfavourable outcome and were dependent on caregivers for daily care. In this study, 29% of the children discharged from the PICU had at least one disability (Taylor A, et al. Intensive Care Med. 2003;29:795-800).
Shein et al (Pediatr Crit Care Med. 2017;18:1106-1113) sought to characterize the acute and more long-term neurofunctional outcomes of a specific population of critically ill children—those younger than 2 years old with acute bronchiolitis. While mortality rates for bronchiolitis are low, it remains one of the most common admitting diagnoses to PICUs, with approximately 20% of admitted patients needing intensive care.
In order to evaluate the incidence of neurologic and functional morbidity (NFM) following the PICU admissions for acute bronchiolitis of previously healthy children, the authors looked at two separate databases: the Pediatric Health Information System (PHIS) and the Virtual Pediatric ICU Systems (VPS). PHIS is an administrative database for more than 45 pediatric hospitals in the United States that can provide information on patient encounters (both in- and outpatient), demographics, diagnoses, procedures, and resource utilization. PHIS was used to flag any encounters that contained a new code for developmental delay, brain imaging, neurologist evaluation, physical therapy, etc. by a patient who had a history of admission to a PICU for bronchiolitis. VPS is a prospectively made database containing demographics, diagnoses, Pediatric Index of Mortality (PIM) 2 scores, and Pediatric Overall Performance Category (POPC) scores of children admitted to a participating PICU. The authors included only children with acute bronchiolitis who had POPC scores of 1 (“good overall performance of activities of daily life and normal cognitive function”) on admission. They then compared the admission POPC scores of enrolled children to the discharge scores.
The authors analyzed more than 13,000 PICU admission for acute bronchiolitis (approximately 9,500 from the PHIS database and approximately 3,700 from the VPS database). They found that 1,104 children from the PHIS database had a subsequent hospital encounter suggesting an NFM. In the VPS database, the authors saw evidence of NFM at PICU discharge in 707 patients (18.6%). Of note, only three patients saw their POPC scores increase by 2 points or more.
The authors also sought to identify risk factors for the development of NFM. The use of mechanical ventilation was the main independent risk factor for NFM in both databases, although there was no difference in PIM 2 scores between ventilated and nonventilated subjects. Certain medications also increased the development of NFM, including diuretics, fentanyl, midazolam, morphine, and neuromuscular blockers. Lorazepam had no effect on NFM, and dexmedotomidine was associated with a decreased rate of NFM. Finally, among the VPS patients, lower pH, higher Pco2, and higher sodium levels were also associated with NFM.
The authors note several limitations of this study, including the ability to identify NFM using only variables available in each database. For example, the use of the POPC score may overestimate the neurologic impairment in discharged PICU patients (Baudin F, et al. Pediatric Crit Care Med. 2017;18:1178-1179), but other scales that may be more useful were not available. Another limitation is that patients may have used hospital resources from non-PHIS institutions.
The authors note that this work is important primarily because it is hypothesis-generating. Most PICU physisicans who have cared for critically ill children with bronchiolitis who seem to completely recover probably do not appreciate the long-term morbidities associated with the PICU experience and so have not considered some of the risk factors detailed in this work. The relationship between pH, Pco2, and NFM gives us pause, when permissive hypercapnia is a mainstream strategy for caring for ventilated patients. The association of NFM with various medications—including opioids and midazolam—adds more evidence for potential neurologic injury and the development and repercussions of delirium.
Author of this installment of Concise Critical Appraisal:
Daniel E. Sloniewsky, MD, is as associate professor in the Division of Pediatric Critical Care Medicine in the Department of Pediatrics at Stony Brook Long Island Children’s Hospital. Dr. Sloniewsky is an editor of Concise Critical Appraisal.