Validation of a centrally maintained computerized hospital database: comparison with operating room logbooks

Substantial interest exists in variations in the use of surgical procedures by specific populations. Studies of this issue are often based on routinely collected data that are maintained in central computer systems. In this study a method is presented for examining the validity of such a database, which is maintained by Kupat Holim, in terms of sensitivity and positive predictive value by comparing its data to data from other information sources, such as operating room log books and in-patient medical records. The validation process was performed in Israel for three surgical procedures: cholecystectomy and prostatectomy each in four hospitals and hysterectomy in two hospitals. The sensitivity of the computerized database ranged from 90% to 98% and the positive predictive value from 96% to 99%. We conclude that the centrally maintained computerized database is a reliable source of information, however, when extremely accurate information is needed the use of complementary sources of information, e.g., operating room logbooks, is recommended.