Pediatric Critical Care

Standardized Mortality Ratio, Unplanned Readmissions, and Length of Stay Ratio

2014 - 2018

The standardized mortality ratio (SMR) is the ratio of actual deaths to predicted deaths. The number of predicted deaths is calculated using the Pediatric Risk of Mortality III (PRISM III) score or the Pediatric Index of Mortality 2 (PIM 2) score — physiology-based mortality risk tools widely used to predict the intensive care outcomes of children. An SMR of 1 indicates that observed mortality is equal to predicted mortality. An SMR > 1 suggests that more patients died than predicted using PRISM III or PIM 2, whereas an SMR < 1 indicates that fewer patients than predicted died.

Quality Measures 2014 2015 2016 2017 2018
SMR based on PRISM III
Cleveland Clinic 0.66 0.97 1.29 1.47 0.93
Reference groupᵃ ᵇ 0.95 1 1.05 1.03 0.95
SMR based on PIM 2
Cleveland Clinic 0.36 0.67 0.99 0.90 0.63
Reference groupᵃ 0.82 0.92 0.98 0.95 0.89
Unplanned readmissions within 24 hours
Cleveland Clinic 0.73 1.4 0.85 1.31 1.14
Reference groupᵃ ᵇ 0.92 0.88 0.88 0.83 0.88
Standardized length of stay ratio based on PRISM III
Cleveland Clinic 1.1 0.96 0.91 0.97 1.02
Reference groupᵃ 1.1 1 1.01 0.99 0.99

PIM = Pediatric Index of Mortality, PRISM = Pediatric Risk of Mortality, SMR = standardized mortality ratio

ᵃData obtained from Virtual PICU Systems, LLC.

ᵇThere is no difference in standardized mortality ratio between Cleveland clinic's Pediatric Intensive Care Unit and the reference group.

PIM 2 uses physiologic data from the first hour of admission to the pediatric intensive care unit (PICU), though data from the immediate pre-ICU time frame may be used if the intensivists treated the patient before arrival in the PICU. The use of mechanical ventilation, whether the admission was elective, whether the patient was in the PICU primarily for postoperative recovery, the use of cardiac bypass, and the presence of certain high- and low-risk diagnoses are some other variables used to calculate the PIM 2 score.

PRISM III uses data from the first 12 hours of admission to the PICU to predict risk of mortality, using the most extreme values documented in the data collection time frame. Compared with PRISM III, PIM 2 is less likely to be biased by the quality of treatment after admission to the PICU but may be subject to bias from different intervention thresholds.

Using PRISM III as an indicator of severity of illness, standardized length-of-stay ratios are calculated by comparing the observed with the expected length of stay.