This appendix documents procedures that the RTI-UNC Evidence-based Practice Center (EPC) staff used to develop this report on screening for prostate cancer. During preparation of the evidence report, we collaborated with two current members of the U.S. Preventive Services Task Force (USPSTF) who served as liaisons to the EPC topic team. We first document the analytic framework and key questions developed at the beginning of the review. We then describe the inclusion and exclusion criteria for admissible evidence, our strategy for literature search and synthesis, and our approach to developing the final summary of the evidence.
The analytic framework (Appendix Figure 1, 9 KB) describes the relationship between screening and treating patients in a clinical setting and reduced morbidity or mortality from prostate cancer. The arrows with superscripts in the analytic framework represent steps in the chain of logic connecting screening with reduced morbidity and/or mortality from prostate cancer; the superscripts refer specifically to nine key questions that guided our literature searches and synthesis of the evidence. We examined one overarching question (Key Question 1, linking screening and ultimate health outcomes) and eight additional questions pertaining to specific links in the analytic framework. The key questions were as follows:
Key Question 1: What are the health outcomes (both type and magnitude) of screening a defined population for prostate cancer compared to not screening?
Key Question 2: What is the yield of screening for prostate cancer (i.e., accuracy and reliability of screening tests, prevalence of undetected cancer in various populations)?
Key Question 3-6: What are the health outcomes associated with treating clinically localized prostate cancer with radical prostatectomy, external beam radiation therapy or brachytherapy, androgen deprivation, or watchful waiting?
Key Question 7: What harms are associated with the treatments of clinically localized prostate cancer?
Key Question 8: What costs are associated with screening for and early treatment of prostate cancer? Have studies modeled the potential benefits of screening? What is the cost-effectiveness of screening for prostate cancer?
Key Question 9: What harms are associated with screening for prostate cancer?
Because we found little evidence about Key Question 9, the harms of screening (one article with inconclusive results), we did not discuss this issue in the article submitted to the Annals of Internal Medicine.
The EPC staff and Task Force liaisons developed eligibility criteria for selecting the evidence relevant to answer the key questions (Appendix Table 1). We first searched for evidence from randomized controlled trials (RCTs) for the efficacy of screening. Because we found no well-conducted and well-analyzed RCT of screening, we then examined case-control and ecologic evidence regarding the overarching key question (Key Question 1).
For Key Question 2, concerning the operating characteristics of screening tests, we examined well-conducted systematic reviews and individual studies that started with a primary care or unselected population without prostate cancer and that compared the findings of one or more screening tests with an adequate reference standard. We also examined evidence of the yield of screening from well-conducted screening programs. For Key Questions 3 through 6, concerning the effectiveness of various therapies, we required evidence from RCTs. For Key Questions 7 and 9, concerning the harms of screening or treatment, we required either RCTs or well-controlled studies that included patient reports and the use of a valid measurement instrument. Finally, for Key Question 8, we searched for evidence of the costs and cost-effectiveness of screening, including models of potential benefits, that considered all appropriate costs and estimates of effectiveness supported by reasonable assumptions based on good evidence.
The analytic framework and key questions guided our literature searches. We examined the critical literature described in the review by the USPSTF (published in 1996)4 and searched the reference lists of systematic reviews (including Cochrane Library reviews) published since 1993. We then used our eligibility criteria to develop search terms and searched the MEDLINE® database for relevant articles concerning humans in the English language published between January 1, 1994, and September 15, 2002. We especially looked for articles involving patients whose experience is clearly generalizable to a primary care U.S. population.
The search strategy and results are given in Appendix Table 1 and Appendix Figure 2. All searches started with the term "prostate neoplasm" and then proceeded by adding further terms as shown in Appendix Table 1.
The first author reviewed abstracts of all articles found in the searches to determine which met eligibility criteria. Other EPC authors of the full systematic evidence review (SER) reviewed all abstracts excluded by the first reviewer. We retrieved the full text of all articles not excluded by both reviewers (next to last column in Appendix Table 1).
One reviewer then examined the full text of all retrieved articles against the eligibility criteria and discussed all excluded articles with one of the other reviewers. We included any article that either reviewer judged had met eligibility criteria (see last column in Table 1). Three of the authors of the SER then divided the articles and abstracted data from them, entering the relevant data into pre-designed evidence tables (Appendix B to the SER, Screening for Prostate Cancer). The abstracting author also graded the articles using the criteria established by the Methods Work Group of the USPSTF.5 The first author read all articles, checked the grading, and discussed the crucial ones with a second reviewer. The authors also discussed key articles with the Task Force liaisons.
We presented an initial work plan, including a provisional analytic framework and key questions, to the entire Task Force in September 2000; we also presented interim reports on results of the literature search and the early results of the synthesis of information in December 2000 and March 2001. A draft of the SER was submitted for broad-based external peer review in May 2001; the peer review involved individual experts in the field, representatives of relevant professional organizations, and representatives of organizations and federal agencies that serve as liaisons to the USPSTF. We revised the evidence review as appropriate after receiving peer review comments. The Task Force reviewed all information and voted on a recommendation in June 2001, revising the Rationale and Recommendation Statement in the spring of 2002 after review by involved professional associations and agencies. We then updated the searches, finalized the review, and shortened it for publication.