Analysis of the effects of applying federal diagnosis-related grouping (DRG) guidelines to a population of high-risk newborn infants.

Length of stay data collected for high-risk newborn infants admitted to a tertiary care children's hospital neonatal unit over a 6-year period were compared with mean and outlier lengths of stay published in the Federal Register as part of a proposed system for prospective payment of hospital cost by diagnosis-related groupings (DRGs). We found that the classification system for newborns markedly underestimated the number of days required for the treatment of these infants. The use of the geometric mean instead of the arithmetic mean as the measure of central tendency was a significant contributor to the discrepancy, especially in those subgroups with bimodal frequency distributions of lengths of stay. Another contributor to the discrepancy was the lack of inborn patients in the children's hospital cohort. The system of prospective payments, as outlined, does not take into account several factors that have a strong influence on length of stay such as birth weight (which requires more than three divisions to serve as an effective predictor), surgery, outborn status, and ventilation. Implementation of the system described in the Federal Register would severely discourage tertiary care referral hospitals from providing neonatal intensive care.