Upstairs downstairs: Vertical integration of a pediatric service

Background. The combined effects of recent changes in health care financing and training priorities have compelled academic medical centers to develop innovative structures to maintain service commitments yet conform to health care marketplace demands. In 1992, a municipal hospital in the Bronx, New York, affiliated with a major academic medical center reorganized its pediatric service into a vertically integrated system of four interdependent practice teams that provided comprehensive care in the ambulatory as well as inpatient settings. One of the goals of the new system was to conserve inpatient resources. Objective. To describe the development of a new vertically integrated pediatric service at an inner-city municipal hospital and to test whether its adoption was associated with the use of fewer inpatient resources. Design. A descriptive analysis of the rationale, goals, implementation strategies, and structure of the vertically integrated pediatric service combined with a before-and-after comparison of in-hospital resource consumption. Methods. A before-and-after comparison was conducted for two periods: the period before vertical integration, from January 1989 to December 1991, and the period after the adoption of vertical integration, from July 1992 to December 1994. Four measures of inpatient resource use were compared after adjustment for case mix index: mean certified length of stay per case, mean number of radiologic tests per case, mean number of ancillary tests per case, and mean number of laboratory tests per case. Difference-in- differences-in-differences estimators were used to control for institution- wide trends throughout the time period and regional trends in inpatient pediatric practice occurring across institutions. Results. In 1992, the Department of Pediatrics at the Albert Einstein College of Medicine reorganized the pediatric service at Jacobi Medical Center, one of its principal municipal hospital affiliates, into a vertically integrated pediatric service that combines ambulatory and inpatient activities into four interdependent practice teams composed of attending pediatricians, allied health professionals, house officers, and social workers. The new vertically integrated service was designed to improve continuity of care for patients, provide a model of practice for professional trainees, conserve scarce resources, and create a clinical research infrastructure. The vertically integrated pediatric service augmented the role of attending pediatricians, extended the use of allied health professionals from the ambulatory to the inpatient sites, established interdisciplinary practice teams that unified the care of pediatric patients and their families, and used less inpatient resources. Controlling for trends within the study institution and trends in the practice of pediatrics across institutions throughout the time period, the vertical integration was associated with a decline in 0.6 days per case, the use of 0.62 fewer radiologic tests per case, 0.21 fewer ancillary tests per case, and 2.68 fewer laboratory tests per case. Conclusions. We conclude that vertical integration of a pediatric service at an inner-city municipal hospital is achievable; conveys advantages of improved continuity of care, enhanced opportunities for primary care training, and increased participation of senior clinicians; and has the potential to conserve significant amounts of inpatient resources.