Page 37 - Screening for Cervical Cancer: Systematic Evidence Review

Chapter II. Methods

Literature Synthesis and Preparation
of Systematic Evidence Review

Data Abstraction and Development of Evidence Tables

We abstracted information about study objective, design, population, conduct, outcomes,

and quality into designated sections and positions within evidence tables created in Microsoft

Excel and Word. Two readers, a methodologist and a clinician-researcher, reviewed each article

in an evidence table. The order of review by each pair of readers was not mandated, and both

parties checked calculations of summary data, such as test sensitivity, that was generated for the

tables.

To assess systematically comparable features of included articles and assure consistency,

we used a checklist of potential indicators of study quality for the literature related to each key

question. For Key Questions 2 and 3, we provide scores using the system designed for the

Evaluation of Cervical Cytology evidence report, which fully documents development of the

32
scoring system. For Key Question 1, we incorporated indicators more relevant to cohort

research, eliminating those items related purely to evaluation of diagnostic tests. Scores were

assigned separately by two individuals and discussed as a group in the rare cases of substantial

differences of opinion. These scores and a global categorization of the internal and external

validity of the reviewed research contributed to grading of individual articles and the body of

33
relevant literature consistent with USPSTF methods.

34