Page 33 - Screening for Cervical Cancer: Systematic Evidence Review

Chapter II. Methods

Key Question 2: New Technologies for Cytology

Although our search strategy identified articles since 1995 relevant to Key Question 2 on

new methods for preparing or evaluating cytology, we conducted full abstraction of only those

publications that had appeared since completion of the AHCPR Evidence Report/Technology

Assessment, Number 5: Evaluation of Cervical Cytology prepared by a team at the Duke Medical

32
Center EPC. Working with colleagues at the Duke EPC, we applied criteria and data extraction

techniques that result in comparable reporting formats between the two evidence reviews. This

approach allows us to streamline preparation of this report while capitalizing on the data

previously collected to describe the performance characteristics of new cytology technologies.

The Evaluation of Cervical Cytology evidence report had several goals, including

developing a model to estimate the costs associated with cervical cancer screening, evaluation,

treatment, and follow-up of cervical cytologic abnormalities and the costs of treatment and

follow-up of cervical cancer. To accomplish their modeling goals and make the best use of the

existing literature, the Duke EPC included studies in which relative performance of new

technologies could be assessed (e.g., use of computerized rescreening compared to conventional

cytology alone) through use of expert adjudication of cytology findings with clinical

confirmation of at least 50% of high-grade lesions.

Our team elected in advance of the literature review to focus primarily on studies that

used clinical confirmation of cytology by colposcopy, cervical biopsy, or both (see Key Question

2 Screener in Appendix B). Specifically, we required that the new method being evaluated be

• Obtained as a screening test or adjunct to screening (i.e., not as follow-up of

documented disease);

30