A b s t r a c tWe rescreened Papanicolaou smear slidesfrom 40,245 women, which In 1987, articles in the Wall Street Journal questioned the compétence of laboratories that read Papanicolaou (Pap) smears and attributed misdiagnosed cases of cancer to "excessive workloads of cytotechnologists, lack of quality control procédures, and poorly educated personnel." 1 2 Congressional hearings on the matter followed, and in 1988, the Clinical Laboratory Improvement Amendments (CLIA) were enacted to require ail people screening or interpreting Pap smears to take a periodic proficiency test, with such testing to take place under normal working conditions. The régulations to implement CLIA specified that a glass-slide test should be used for this proficiency test. 3 It has not been possible to implement such a program nationwide because of the resources needed to assemble and maintain the necessary number of high-quality glass slides in the required diagnostic caté-gories for the test. [4][5][6] This has renewed a longstanding discussion about the validity of glass-slide proficiency testing in Pap smear quality assurance and stimulated a search for alternative testing stratégies. 7 -8 State operated and approved proficiency testing programs hâve been available in the United States for more than 20 years, and thèse hâve been studied in détail, 5 -9-19 but the validity of the glass-slide proficiency test for measuring screening skills has not been established firmly.
20The goals of the présent study were to détermine whether the performance of screeners who screen Pap smears (determined by rescreening their previous work) correlates with screeners' performance on a glass-slide proficiency test, as required by the CLIA régulations, and to evaluate a computer-based proficiency test recently developed at the