A chart audit was conducted at 28 intensive care nurseries in the Philadelphia region. All infants studied were insured by a medical assistance health maintenance organization. Hospitals were grouped by level of care and location. A total of 3 academic affiliate (A), 13 community (C), 6 rural (R) and 6 university (U) hospitals were audited for a recent 1 year period. Each audited hospital's charts were analyzed for length of hospital stay (LOS), and infants were placed in one of 30 groups as defined as an identified category of patients utilizing weight and disease process. LOS for each infant was compared against the National Perinatal Information Center (NPIC) data base. NPIC is a diverse payor mix, and severity of illness is not controlled. The difference between LOS and that predicted by the NPIC data are potential days saved (days saved). Data for charts reviewed (n), hospital days (hosp days), average LOS (ALOS), NPIC ALOS, and days saved are: Table Days saved also varied markedly between hospitals within the same classification. This analysis reveals the unwarranted variation in length of stay characteristic of unmanaged systems of care. A management plan is now in process to define the causes of variation. Supported by Paidos Health Management Systems, Inc.

Table 1
Full size table