Feasibility Study for Neurodevelopment Follow-up Study in PICU

The majority of human cohort investigations of neurotoxicity have studied general anesthesia for surgery. However, many animal studies have involved exposure to agents for significantly longer than the average duration of a general anesthetic in infants, some even for 24 hours (8).This is closer in exposure characteristics to prolonged sedation in PICU: days as compared to a few hours. However, although sedation in the Pediatric Intensive Care Unit (PICU) may continue for several days, there are no studies of its effects on subsequent neurodevelopment in previously healthy young infants. The only intensive care studies thus far have examined high risk subgroups, including the EPIPAGE study (20) in premature newborns and studies of infants after congenital heart surgery. The latter studies have shown a small but significant association between cumulative benzodiazepine dose and Visual-Motor integration at 4 years of age (21), and an association between higher exposure to volatile anesthetic agents and lower cognitive scores on the Bailey Scales of Infant Development III at 12 months (22), respectively.

It is essential that Pediatric Intensivists start investigating the effect of our commonly used sedation protocols, which frequently include benzodiazepines, on our patients' neurodevelopment. The standard sedative regimen in use for many years at Montreal Children's PICU is midazolam infusion (benzodiazepine) plus opioid (commonly fentanyl or morphine) infusion. Other agents including ketamine may be added. This is not an easy field of study since we cannot randomize patients to receive sedation or not.

Therefore this programme of research aims to:

1. Identify a suitable group of PICU patients who were exposed to PICU sedation, along with a comparable group who were not exposed to sedation during their PICU admission who will act as controls; b) Assess the feasibility of recruiting these retrospectively identified patients and performing the requisite neurodevelopmental assessments around 6 years after PICU admission; c) Expand to a multicentre study, if feasibility is confirmed, to allow the recruitment of sufficient numbers of cases to determine if PICU sedation has a significant effect on long-term neurodevelopment; d) Investigate alternative sedation protocols and/or interventions to minimize the adverse effects on development.
2. The initial aim of this study programme has been achieved by means of a Chart Review completed in March 2017. We hypothesized that previously healthy infants requiring ventilation for an acute respiratory illness would divide into two groups depending on their respiratory management: intubation with sedation versus non-invasive ventilation without sedation, and that these infants, with a unifying pathological diagnosis, could act as cases and controls for an investigation of PICU sedation. In order to optimize the validity of our comparison, patients must be as similar as possible, therefore many exclusion criteria and also matching are required. The chart review has determined that the two described groups are similar in terms of severity of illness scores (Pediatric Risk of Mortality [PRISM] and Pediatric Logistic Organ Dysfunction [PELOD]) therefore making them suitable subjects for a follow-up study (23). This phase of the programme has been presented and discussed at Canadian Critical Care Trials Group (CCCTG) meetings and endorsed by CCCTG.
3. Having identified an appropriate group of patients, the next phase of the research programme is to perform a feasibility study to determine if our methodology will lead to adequate recruitment to move forward to a large multicentre study. This is the subject of this protocol.

The proposed methodology is similar to that used successfully in the PANDA study (16) where study participants were identified using billing records (hernia surgery at up to 36 months of age). Their records were checked for exclusion criteria based on age, gestational age at birth and general health. Parents were contacted and invited to participate. Further exclusions were based on more detailed medical history from the parents and the existence of a sibling within 3 years of the index child's age who had not had any anesthetic exposure and who was also healthy. Another general anesthesia study with similar methodology is currently analysing data (24). This group from the Mayo Clinic used population-based controls for anesthesia-exposed index cases identified from a previous cohort study (10). Our participant identification will also be retrospective, from medical records to identify sedation-exposed and non-exposed subjects from within the PICU population as described below.

Hypothesis: feasibility study Patients identified from the PICU admission records and Medical Records search can be suc-cessfully recruited and assessed by standard Neurodevelopmental testing 6 years after their admission.

Hypothesis: definitive study We hypothesize that benzodiazepines alone or with opioid infusion during mechanical ventilation of previously healthy infants with respiratory diagnoses would decrease Wechsler Preschool and Primary Scale of Intelligence III (WPPSI III) score at 6 yrs, compared to infants who were treated with non-invasive ventilation without sedation.

Aims for feasibility study

1. Identify patients from PICU database and from previous Chart Review Database (see below).
2. Recruit 20 patients each from the sedated/intubated/ventilated and non-sedated/non-invasive ventilation groups (total 40).
3. Perform pediatric and neurodevelopmental assessments on patients at 5 years 8 months to 7 years 4 months of age using a battery of tests with Wechsler Pediatric and Primary Infant scales of Development (IV) as the primary outcome.

Primary outcome Successful identification and recruitment of patients and performance of testing.

Secondary outcomes Performance of Wechsler Pediatric and Primary Infant scales of Development IV (WPPSI IV) and NEPSY-II™ tests at 5 years 8 months to 7 years 4 months .

Completion of Childhood Memory Score (CMS™) and Child Behaviour Checklist (CBCL)™ questionnaires Pediatric examination: incidence of neurological abnormalities including cerebral palsy, hearing or visual deficits, seizure disorder, Attention Deficit disorder, Autism spectrum disorder.