ABSTRACT
Objective:
To determine the concordance of mortality risk classification through the use of the Pediatric Index of Mortality (PIM) 2 and 3.
Methods:
Through a retrospective cohort, we evaluated patients admitted to the pediatric intensive care unit between April 2016 and December 2018. We calculated the mortality risk with the PIM 2 and 3. Analyses were carried out to determine the concordance between the risk classification obtained with both scales using unweighted and linearly weighted kappa.
Results:
A total of 722 subjects were included, and 66.6% had a chronic condition. The overall mortality was 3.7%. The global kappa concordance coefficient for classifying patients according to risk with the PIM 2 and 3 was moderate at 0.48 (95%CI 0.43 - 0.53). After linear weighting, concordance was substantial at 0.64 (95%CI 0.59 - 0.69). For cardiac surgery patients, concordance for risk classification was fair at 0.30 (95%CI 0.21 - 0.39), and after linear weighting, concordance was only moderate at 0.49 (95%CI 0.39 - 0.59). The PIM 3 assigned a lower risk than the PIM 2 in 44.8% of patients in this subgroup.
Conclusion:
Our study proves that the PIM 2 and 3 are not clinically equivalent and should not be used interchangeably for quality evaluation across pediatric intensive care units. Validation studies must be performed before using the PIM 2 or PIM 3 in specific settings.
Keywords:
Pediatric Index of Mortality; Mortality; Infant, newborn; Cardiac surgery; Intensive care units, pediatric