Cohort selection
Consecutive patients aged ≤16 years who had no tracheostomy when admitted to ICUs or PICUs during the study period, between April 2014 and March 2017, were included in this study. Patients with tracheostomies before admittance to the ICU or PICU were excluded. Patients were not directly involved in the design of this study. All data were anonymized prior to their availability for this study by JaRPAC. This study was approved by the Institutional Review Board (30-025, in the Juntendo University Urayasu Hospital, Chiba, Japan), which waived the need for informed consent.
The JaRPAC is a multicentre clinical database of ICU and PICU paediatric patients that was founded by the Japanese Society for Emergency Medicine. It was initiated in April 2014, with the aim of evaluating critically ill paediatric patients and reducing their mortality rate. The JaRPAC database contains anonymized information regarding patient demographics, admissions, treatment, and outcomes, as well as scoring systems for severity and mortality [15]. Paediatric patients ≤16 years old in ICUs or PICUs are eligible for inclusion in this registry, and data are available on a per capita basis. The data were collected from admission until discharge from the ICU or PICU. The national Center for Child Health and Development is the primary institute managing this registry data, and hospitals that are affiliated with this institute are selected to participate in the registry. This includes twelve PICUs at children’s hospitals and eleven ICUs at critical care centres.
Design
This was an epidemiologic study based on JaRPAC data. Data concerning patients who had not undergone tracheostomy when admitted to the ICU or PICU were extracted from the database. These patients were divided into two groups: those who received tracheostomies while in ICU or PICU (tracheostomy group) and patients without tracheostomies (no-tracheostomy group). Risk factors for tracheostomy were evaluated using the JaRPAC data. The cause of admission was divided into six categories: respiratory failure, circulatory failure, neurological dysfunction, post-operative care, tight observation, and recovery from cardiopulmonary arrest (CPA). The final diagnosis for each patient was registered and assigned as either an intrinsic or an extrinsic cause. Intrinsic disease was coded based on the International Classification of Diseases v. 10 (ICD-10) and categorized into one of ten groups (cardiovascular, respiratory, neuromuscular, gastrointestinal/hepato-biliary-pancreatic, haematologic/oncologic, renal, sepsis, metabolic/endocrinologic, allergic groups, and others) in order to ensure sufficient patients for analysis.
We used the Paediatric Index of Mortality 2 (PIM2) as a measure of severity for patients. The PIM2 score is calculated from various coefficients determined by Slater et al.[15] The values used to calculate PIM2 result from the first face-to-face contact between patients and physicians at ICUs or PICUs. Data for some factors were not obtained for all cases; these factors were not included in the PIM2 calculations in these cases. Patient survival was defined as discharge from an ICU or PICU.
Post-operative care admission was considered as elective admission. Admissions from general wards or transportation from other hospitals due to rapid deterioration or from the emergency department (ED) were considered urgent admissions. The duration of interventions performed in the ICU or PICU were compared between the groups. Interventions included continuous mechanical ventilation (CMV), central venous access catherization (CV), peripherally inserted central catheterization (PICC), and arterial line catherization (A-line).
We also evaluated complications, such as acute respiratory distress syndrome (ARDS) and ventilator associated pneumonitis (VAP). ARDS was defined by definition of Berlin criteria, and VAP was considered as a pneumonitis associated with a mechanical ventilation period lasting over 48 hours[16, 17].
We defined chronic conditions according to Feudtner et al.’s definition, which states that a chronic condition ‘involves either several different organ systems or one organ system severely enough to require specialty paediatric care and probably some period of hospitalization in a tertiary care centre.’ [18] Chronic conditions were grouped into eight systems (cardiovascular, respiratory, neuromuscular, congenital/genetic abnormalities, gastrointestinal, renal, metabolic/endocrinologic and hematologic/immunologic), based on Feudtner’s complex chronic conditions. Children with multiple chronic conditions were counted multiple times, in each group corresponding to their conditions, for specific analysis, but were only counted once in the overall analysis. A clinically dominant chronic condition was defined as ‘the medical condition which carried the greatest morbidity for the child.’[19]
Statistical Analysis
Data regarding age, length of PICU or ICU stay, PIM2, and length of interventions from JaRPAC were clearly skewed, so medians with interquartile ranges were used for numerical variables. Numerical variable differences between the two groups were compared using a Mann-Whitney U test. The chi-square test was used to compare sex distribution as well as frequencies of urgent admission, chronic conditions, chromosomal anomalies and complications. Data management and statistical analyses were undertaken using EZR software (Y Kaneda, Saitama, Japan). A p-value of < 0.05 was considered statistically significant.