Study Highlights Importance of Different-Hospital Readmission in Measuring Pediatric Care Quality

A study published in JAMA Pediatrics examined different-hospital readmission (DHR) as a tool to measure quality of care, considering that while hospitals can track readmissions to their own site, they lack information when their patients are admitted to a different hospital. So the 30-day readmission data commonly used by payers and hospitals themselves to assess healthcare quality may not be the most accurate.

Among pediatric patients, 6.5% have been shown to have unplanned readmissions within 30 days of an acute care incident. While DHR statistics exist for adult readmissions, data for pediatric patients are scarce. The authors therefore estimated the prevalence of 30-day pediatric DHR, assessed their influence on hospitals’ estimated readmission performance, and listed patient and hospital characteristics associated with DHR.

The data set that was evaluated for the study included hospital discharges from January 1, 2005 to November 30, 2009, in the Agency for Healthcare Research and Quality’s New York State Inpatient Database. These data were associated with patients aged 0 to 17 years when admitted. Exclusions included healthy newborns, obstetric care, mental health conditions, conditions ending in death or departure against medical advice, incomplete records, Readmissions for planned procedures and chemotherapy were also excluded. Outcomes measured were the proportion of DHR versus same-hospital readmission (SHR), all-hospital readmission (AHR), and SHR rates.

The authors found that nearly 14% of the 31,325 AHRs were DHRs. For each individual hospital, the median percent DHR was 21.6%, median adjusted AHR was 3.4%, and the median adjusted SHR was 2.5%. SHR-based excess readmission rates were inaccurate for 11.3% of the 177 hospitals included in the study, all of which were nonchildren’s hospitals and 90% were nonteaching hospitals. Higher odds ratios of DHR were associated with being younger, being white, having private insurance, having a chronic condition indicator for a mental disorder, disease of the nervous system or circulatory system.

The authors conclude that SHRs are not an appropriate surrogate for AHRs and that failing to incorporate DHRs into readmission measurement may impede quality assessment, anticipation of penalties, and quality improvement.