Meta_Analysis在肿瘤循证医学中的应用_英文(2)

Chin J Clin Oncol (2008) 5: 1~9updated collection of evidence-based medicine da-tabases, including The Cochrane Database of Sys-tematic Reviews. The Cochrane Library interface is provided by Wiley InterScience. The Cochrane Li-brary collected 5,053 systematic reviews and 522,340 randomized controlled clinical trials. Among those, 373 systematic reviews and 28,743 trials in the field of oncology were collected up to November 2007[4]. Evidence-based medicine and meta-analy-sisof the breast, lung, colon, liver, stomach and cervix uteri were reviewed in three fields, viz., cancer etio-logic research, screening and therapy. The selection categories of publications in this review were based on up-to-date, new problems and inconsistent results obtained from different studies. EtiologyMeta-analysis is the quantitative assessment of the pooled results of multiple clinical trials. It is ideally suited for use in assessing results of trials which, in oncology in particular, are often under-powered to de-tect small differences in primary endpoints, let alone subgroup analyses. There are 3 levels of evidence when we look at evidence-based medicine. The first level, the highest level, is data from randomized controlled trials, or meta-analyses of randomized controlled trials. The second level is data from case control studies or co-hort studies, and the third level of evidence is from case studies or single case reports. Based on PubMed Medline searching, the publica-tions for researches on evidence-based oncology by using meta-analysis increased from 107 in 1997 to 372 in 2006.

Basic methods of meta-analysis[5]

Meta-analysis has become a clearly defined tech-nique, with reporting standards for both randomized controlled trials and observational studies. There is a clear order for proceeding with a meta-analysis beginning with selection of the subject fol-lowed by retrieval of primary studies, evaluation of study quality and selection of those studies to be used. Thereafter, an appropriate statistical model must be selected, and the studies evaluated for heterogene-ity (qualitative and quantitative). The actual meta-analysis is then performed, and finally the results are evaluated for reproducibility (sensitivity testing) to ensure that bias does not influence the result. There are a number of freeware programs for meta-analysis. RevMan (Review Manager) developed by the Cochrane Collaboration can be downloaded from /RevMan. Another freeware, EPIMETA developed by CDC/USA can be down-loaded from http://ftp.cdc.gov/pub/Software/epimeta.Evidence-Based Oncology for Breast Cancer

Recent findings by using meta-analysis for cancers Breast cancer, as one of the known hormone related cancers, is a common malignant female disease. In recent decades, the associations between genes and breast cancer have been studied. Chen, S. and G. Par-migiani[6] analyzed 10 studies with a meta-analysis method, and reported that the cumulative breast can-cer risks at age 70 years were 57% (95% confidence interval [CI]=47%~66%) for BRCA1 and 49% 95% CI=40%~57%) for BRCA2 mutation carriers. An-other BRCA gene related disease, ovary cancer risk of 40% (95% CI=35%~46%) for BRCA1 and 18% (95% CI=13%~23%) for BRCA2 mutation carri-ers. The authors concluded that the risk estimates for BRCA1 and BRCA2 mutation carriers can be used by counselors and clinicians. The association between diabetes mellitus and risk of breast cancer has been summarized by Larsson, et al.[7] by using a random-effects model of meta-analysis including 20 studies (5 case-control and 15 cohort studies). All 20 studies showed that women with (versus without) diabetes had a statistically sig-nificant 20% increased risk of breast cancer (relative risk [RR], 1.20; 95% CI=1.12~1.28). The summary estimates were similar for case-control studies (RR, 1.18; 95% CI=1.05~1.32) and cohort studies (RR, 1.20; 95% CI=1.11~1.30). This meta-analysis indi-cates that diabetes is associated with an increased risk of breast cancer. It has being asked for decades whether oral contra-ceptives (OC) increase breast cancer risk. Kahlenborn et al.[8] performed a meta-analysis of case-control studies published after 1980, and examined the as-sociation between prior OC use and premenopausal breast cancer. Thirty-four studies were identified. Use of OCs was associated with an increased risk of pre-menopausal breast cancer in general (odds ratio [OR], 1.19; 95% CI=1.09~1.29) and across various patterns of OC use. OC use was associated with breast cancer risk in both parous (OR, 1.29; 95% CI=1.20~1.40) and nulliparous (OR, 1.24; 95% CI=0.92~1.67) wom-en. Use of OCs is associated with an increased risk of premenopausal breast cancer, especially with use before first full-term pregnancy in parous women. Green tea is another topic of great interest for the public as well as for medical professionals. Sun et al.[9] concluded that meta-analysis indicated a lower

百度搜索“77cn”或“免费范文网”即可找到本站免费阅读全部范文。收藏本站方便下次阅读，免费范文网，提供经典小说教育文库Meta_Analysis在肿瘤循证医学中的应用_英文(2)在线全文阅读。