Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Recorded presentation

Scott A. Lorch, MD, MSCE¹, Gabriel Escobar, MD², Susan Bakewell-Sachs, PhD, RN, APRN³, Barbara Medoff-Cooper, PhD, RN⁴, and Jeffrey H. Silber, MD, PhD¹. (1) Pediatrics, The Children's Hospital of Philadelphia, 3535 Market Street, Suite 1029, Philadelphia, PA 19104, 215-590-1714, lorch@email.chop.edu, (2) Division of Research, Kaiser Permanente, Northern California, 2000 Broadway Ave, Oakland, CA 94612, (3) School of Nursing, The College of New Jersey, P.O. Box 7718, Ewing, NJ 08628, (4) School of Nursing, University of Pennsylvania, 420 Guardian Drive, Philadelphia, PA 19104

Objective: Pay-for-performance has increased the interest in severity adjustment for outcomes in outpatient pediatrics, such as rehospitalizations of premature infants. Traditional severity adjustors, such as gestational age (GA) and the score of neonatal acute physiology (SNAP), may not adjust for later-occurring differences in neonatal practice. The objective of this study was to (1) evaluate the discrimination of various severity adjustment tools for the rehospitalization of premature infants within 2 and 4 weeks of discharge and (2) determine how the addition of physiologic maturity at discharge, a measure of later-occurring differences in the condition of neonates, influences the rank ordering of 31 outpatient pediatric practices.

Study Design: As part of the Infant Functional Status Study, we determined the physiologic maturity at discharge using the number of days prior to discharge that an infant was observed without supplemental heat or oxygen, apnea or bradycardia, methylxanthines, naso-gastric feedings, or weight loss. Eligible infants were born at a GA ≤ 32 weeks at 5 Kaiser Permanente hospitals (N=780) from 1998-2001. Logistic models were constructed to predict 2-week and 4-week rehospitalizations using various predictors including (1) GA and race; (3) GA, race, SNAP, and complications; and (4) GA, race, complications, SNAP, and physiologic maturity at discharge. These models were then used to rank order outpatient facilities.

Results: 280 of the 780 infants had a GA ≤ 28 weeks. The median number of days from attainment of each skill to discharge ranged from 7 days (interquartile range 5-11 days) for feeding to 20 days (interquartile range 14-27 days) for methylxanthines. Compared to GA and race (c-statistic 0.71), the addition of maturity variables resulted in the highest c-statistic for both the 2-week (c-statistic 0.84, p<0.02 versus all other models) and 4-week rehospitalization models (c-statistic 0.78). The addition of the maturity variables to the severity adjustment model with GA, SNAP, race, and complications resulted in 26% and 19% of the outpatient facilities changing rank order quintiles for 2-week and 4-week rehospitalizations, respectively. 11 of the 14 hospitals that changed quintiles moved at least 2 ranks. When the model accounted for physiologic maturity, 26% and 13% of the facilities in the 2-week and 4-week rehospitalization models, respectively, moved at least 10% in the final rank order.

Conclusions: The type of severity adjustment can significantly alter the ranked quality of care provided to premature infants by outpatient pediatric practices. Clinical information around discharge may improve the adequacy of severity adjustment.

Learning Objectives:

(
• Understand the factors that affect the ranked quality of care provided by outpatient pediatric practices (
• Analyze how early and late-occurring differences in the condition of premature infants can affect the measurement of illness severity. (
• Recognize the issues that need clarification to institute pay-for-performance programs in pediatrics

Keywords: Quality of Care, Peer Counselors

Presenting author's disclosure statement:

Any relevant financial relationships? No

Recorded presentation