[Risk adjusted assessment of quality of perinatal centers - results of perinatal/neonatal quality surveillance in Saxonia].

Background: The weak point of the country-wide perinatal/neonatal quality surveillance as a tool for evaluation of achievements of a distinct clinic, is the ignorance of interhospital differences in the case-mix of patients. Therefore, that approach can not result in a reliable bench marking.

Objective: To adjust the results of quality assessment of different hospitals according to their risk profile of patients by multivariate analysis.

Method: The perinatal/neonatal data base of 12.783 newborns of the saxonian quality surveillance from 1998 to 2000 was analyzed. 4 relevant quality indicators of newborn outcome -- a) severe intraventricular hemorrhage in preterm infants < 1500 g, b) death in hospital of preterm infants < 1500 g, c) death in newborns with birth weight > 2500 g and d) hypoxic-ischemic encephalopathy -- were targeted to find out specific risk predictors by considering 26 risk factors. A logistic regression model was used to develop the risk predictors.

Results: Risk predictors for the 4 quality indicators could be described by 3 - 9 out of 26 analyzed risk factors. The AUC (ROC)-values for these quality indicators were 82, 89, 89 and 89 %, what signifies their reliability. Using the new specific predictors for calculation the risk adjusted incidence rates of quality indicator yielded in some remarkable changes. The apparent differences in the outcome criteria of analyzed hospitals were found to be much less pronounced.

Conclusion: The application of the proposed method for risk adjustment of quality indicators makes it possible to perform a more objective comparison of neonatal outcome criteria between different hospitals or regions.