Algorithm Helps Track Hospital Use Among Ventilator-Dependent Children
Published July 2018 | Pediatric Pulmonology
Children with established tracheostomy and ventilator dependence are known to be high users of healthcare resources. However, keeping track of their care over time in ways that provide useful data for analysis has been a challenge.
That overwhelming task has been greatly simplified thanks to an algorithm developed and tested at Cincinnati Children’s. A transdisciplinary team led by Barbara Giambra, PhD, RN, who is mentored by Maria Britto, MD, MPH, used the new tool to accurately pull data from the massive Pediatric Health Information System (PHIS). This system contains deidentified billing information for more than six million patient hospitalizations from 49 hospitals affiliated with the Children’s Hospital Association.
After sorting through more than 20,000 Cincinnati Children’s discharge records for 2014, the algorithm identified 152 unique patients with tracheostomy and ventilator dependence. That matched up well with the 153 patients listed in a hand-curated registry within the Division of Pulmonary Medicine. Overall, the study reports a sensitivity rate of 91% for the algorithm and a specificity rate of 99%.
The findings suggest that the algorithm can be useful to any pediatric pulmonologist working with PHIS data.
“For instance, discovering the most frequent cause for preventable readmissions among these children might lead to improvements in family education initiatives provided prior to discharge or in the clinic setting,” the co-authors state. “Future studies could also use the algorithm to identify variation in treatment, healthcare utilization and resource use among children’s hospitals and regions across the United States.”
