Page 31 - Screening for Cervical Cancer: Systematic Evidence Review

Chapter II. Methods

Literature Reviewed

Citation Database

We used the search strategy in Table 4 to identify potentially relevant publications. From

our first search (December 1999) we imported references into a ProCite database, which

enumerates, stores, manages, and retrieves bibliographic citations; we also recorded the fate of

all identified publications as they were screened for inclusion. We repeated the search in June

2000, eliminating duplicate references and adding the new additions to the ProCite database.

Screening of Articles

Two EPC staff independently reviewed the titles and abstracts of the articles identified

and excluded those that did not meet eligibility criteria. If the reviewers disagreed, we carried

the article in question forward to the next stage in which we then reviewed the full article and

made a final decision about inclusion or exclusion. At each step we recorded the fate of the

article in the ProCite database. Table 5 presents the disposition of articles identified as

potentially relevant publications (for review of the full article), summarizing the number of

publications at each step and their categorization.

Limiting the exhaustive search (search 8, Table 4) to identify only articles that reported

on trials, we identified 57 articles. Of these, 25 are primary reports of randomized trials: 15

address methods to promote uptake and continuance of appropriate screening; 3 examine

methods to improve follow-up of abnormal screening findings; 3 compare tools for collecting

cytologic samples (i.e., type of spatula, brush or swab); 3 investigate patient education and

28