Page 27 - Screening for Cervical Cancer: Systematic Evidence Review
P. 27
Chapter I. Introduction
evaluate the effectiveness of cervical cancer screening systems for improving these outcomes.
Given the rarity of cervical carcinoma among women who receive any screening, much of this
literature examines surrogate endpoints such as the diagnosis of dysplasia by colposcopy and/or
cervical biopsies. The outcomes we have been able to assess were determined by the content of
the literature. Outcome(s) of interest were tailored to the related key questions in the following
fashion.
Outcomes for Key Question 1A and 1B, which address candidates for screening, include
diagnosis of dysplasia, severity of dysplasia, progression of dysplasia, diagnosis of carcinoma,
stage at diagnosis, and mortality. We retained cost data, if relevant to care in the United States,
for background information. We also sought harms related to screening.
For Key Questions 2 and 3, which examine the test characteristics of new methods of
preparing and evaluating cervical cytology and the contribution of HPV testing, the relevant
outcomes include sensitivity, specificity, positive and negative predictive value, and diagnostic
yield when compared to a gold standard. The standard used and the quality of evaluation of
normal screening results both vary. Much of this literature is limited by failure to apply the gold
standard evaluation to normal subjects, which precludes definitive assessment of specificity. We
sought to evaluate the influence of these tools on the diagnoses and features listed above for Key
Question 1, and we retained literature relevant to US costs for background.
Chapter II provides an overview of our methods for producing the SER. Chapter III
presents the results of our literature search and synthesis organized by the key questions. We
discuss the results and the limitations of the literature in Chapter IV with attention to
ramifications for future research. Appendix A contains acknowledgments, Appendix B provides
17
