Recognizing Differences in Hospital Quality Performance for Pediatric Inpatient Care

Abstract

Background: Hospital quality-of-care measures are publicly reported to inform consumer choice and stimulate quality improvement. The number of hospitals and states with enough pediatric hospital discharges to detect worse-than-average inpatient care remains unknown.

Methods: This study was a retrospective analysis of hospital discharges for children aged 0 to 17 years from 3974 hospitals in 44 states in the 2009 Kids' Inpatient Database. For 11 measures of all-condition or condition-specific quality, we assessed the number of hospitals and states that met a "power standard" of 80% power for a 5% level significance test to detect when care is 20% worse than average over a 3-year period. For this assessment, we approximated volume as 3 times actual 2009 admission volumes.

Results: For all-condition quality, 1380 hospitals (87% of all pediatric discharges) and all states met the power standard for the family experience-of-care measure; 1958 hospitals (95% of discharges) and all states met the standard for adverse drug events. For condition-specific quality measures of asthma, birth, and mental health, 203 to 482 hospitals (52%-90% of condition-specific discharges) met the power standard and 40 to 44 states met the standard. One hospital and 16 states met the standard for sickle cell disease. No hospital and ≤27 states met the standard for the remaining measures studied (appendectomy, cerebrospinal fluid shunt surgery, gastroenteritis, heart surgery, and seizure).

Conclusions: Most children are admitted to hospitals in which all-condition measures of quality have adequate power to show modest differences in performance from average, but most condition-specific measures do not. Policies regarding incentives for pediatric inpatient quality should take these findings into account.

FIGURE 1

Patients discharged from hospitals with enough discharges to detect worse-than-average performance on inpatient quality. The figure presents the extent to which children are discharged from hospitals that had a sufficient number of discharges over 3 years to detect significant differences in performance worse than average. For example, of all children admitted with sickle cell crisis, 5% were admitted to a hospital that was able to detect a 20% difference in performance on 30-day readmission. The asterisk indicates that no hospitals had enough discharges to detect a significant difference in performance worse than average.

FIGURE 2

States with enough discharges to detect worse-than-average performance on inpatient quality. The figure presents the extent to which states had sufficient number of discharges over 3 years across all of their hospitals to detect significant differences in performance worse than average. For example, 42% of states had enough hospital discharges over 3 years to detect a 20% difference in performance on 30-day readmission for seizure. KID 2009 contains hospitalizations from 44 states.