Pediatric Readmission Prevalence and Variability Across Hospitals
Berry JG, Toomey SL, Zaslavsky AM, et al
Reducing hospital readmissions through financial disincentives is a cost-saving mechanism proposed under the current healthcare reform efforts. Few data exist on the nature of pediatric rehospitalizations. This study used approximately 2 years of data from 72 children's hospitals in the United States to describe the nature of pediatric readmissions, focusing on specific conditions, and to identify where significant variation in readmission occurs. The ultimate goal was to identify probable targets for intervention.
Methods. The study children were aged 18 years or younger and were enrolled at 1 of the 72 hospitals. For each patient, the investigators began with an index admission and looked for the first (if any) readmission within 30 days of the index admission. Accepted approaches for classifying the primary diagnosis of each admission according to diagnostic-related groups were used.
Groups that might be routinely expected to return for subsequent admission, such as patients receiving chemotherapy, were excluded, as were other specialty patients who might be seen for subsequent, planned procedures. The analyses accounted for age, the presence of chronic medical conditions, insurance source, race and ethnicity, length of stay of the index hospitalization, and the characteristics of the individual children's hospitals.
Findings. More than 568,000 index admissions met the inclusion criteria, and 35% of the patients had at least 1 chronic medical condition. Of all index admissions, 6.5% were readmitted within 30 days of discharge. Of all readmitted children, 39% were readmitted within 7 days of discharge, and 61% within 14 days of discharge. About one third of the readmitted children had at least 1 chronic condition; congenital anomalies were the most prevalent chronic condition, closely followed by respiratory diseases and neurologic diseases. In adjusted analyses, children younger than 1 year to 4 years of age were more likely than older children to experience an unplanned readmission.
The number of chronic conditions per child was positively correlated with odds of readmission. Children with 4 or more chronic conditions had an adjusted readmission odds of 5.27 (95% confidence interval, 5.05-5.40). Patients who had publicly funded insurance were slightly more likely to be readmitted than patients who were privately insured. Both black and Latino children had greater odds of being readmitted (adjusted readmission odds, 1.06 and 1.09, respectively). The type of children's hospital (freestanding or not) was not associated with odds of readmission, nor was region of the country.
Ten chronic medical conditions accounted for 27.7% of pediatric readmissions. The most frequent conditions are those that many clinicians might guess: anemia, neutropenia, ventricular shunt, sickle cell disease, seizure disorder, and asthma. Other chronic conditions associated with a higher readmission frequency included episodic conditions, such as gastroenteritis, upper respiratory tract infection, pneumonia, appendectomy, and bronchiolitis.
Among the 10 conditions with the highest readmission prevalence, 3 had strikingly high readmission rates. Children with anemia or neutropenia experienced a 22.5% frequency of readmission, compared with 18.1% of children with a ventricular shunt and 16.9% of children admitted for a sickle cell anemia crisis. All other conditions had readmission frequencies of 7% or less. The researchers estimated that 48.3% of the readmissions were for a diagnosis related to the index admission.
Considerable variation was identified among the hospitals for overall readmission frequency, as well as for readmission frequency by condition. Berry and colleagues concluded that the overall frequency of unplanned pediatric readmissions within 30 days was 6.5%, and they emphasized the considerable variation in readmission rates among the hospitals.