[Evaluating hospital appropriateness with different tools: administrative data versus analytic review]

PRUO, a modified version of AEP, is a widely used clinical-based tool to evaluate hospital appropriateness in Italy. We developed the APPRO method for assessing organizational appropriateness using administrative data. APPRO estimates the amount of inappropriate hospitalisation, giving consideration to severity of illness through APR-DRG classification system. The aims of the study were to: measure the agreement between evaluators using PRUO; investigate the relation between APR-DRG severity subgroups and PRUO assessment; asses the validity of APPRO method comparing its performance to PRUO results. We selected 361 hospital episodes assigned to DRG 39 ("lens procedures with or without vitrectomy") and 242 hospital episodes assigned to DRG 183 ("miscellaneous of digestive disorders, age > 17") from three hospitals in 2000. Clinical records were independently evaluated by two pairs of physicians using PRUO. Proportions of inappropriate episodes by hospital and DRG were also estimated through APPRO using data from Lazio regional hospital information system. The agreement between the two pairs of evaluators was high (k=0.93; p<0.0001). We observed no statistically significant association between APR-DRG severity subgroups and inappropriate hospitalisation found by PRUO. APPRO underestimates rates of inappropriate hospitalisation compared to PRUO. It depends on the different characteristics of the tools and particularly on the caution of APPRO in performing the evaluation using routine data.