The strengths and limitations of meta-analyses based on aggregate data

被引：179

作者：

Lyman G.H. ^{[1
]}

Kuderer N.M. ^{[1
]}

机构：

[1] Department of Medicine, University of Rochester Medical Center, Rochester

来源：

BMC Medical Research Methodology | / 5卷 / 1期

关键词：

Individual Patient Data; Preventive Service Task; Individual Investigator; Summary Effect Size; Survival Curve Estimate;

D O I：

10.1186/1471-2288-5-14

中图分类号：

学科分类号：

摘要：

Background: Properly performed systematic reviews and meta-analyses are thought by many to represent among the highest level of evidence addressing important clinical issues. Few would disagree that meta-analyses based on individual patient data (IPD) offer several advantages and represent the standard to which all other systematic reviews should be compared. Methods: All cancer-related meta-analyses cited in Medline were classified as based on aggregate or individual patient data. A review was then undertaken of all reports comparing the comparative strengths and limitations of meta-analyses using either aggregate or individual patient data. Results: The majority of published meta-analyses are based on summary or aggregate patient data (APD). Reasons suggested for this include the considerable resources, years of study and often, broad international cooperation required for IPD meta-analyses. Many of the most important features of systematic reviews including formal meta-analyses are addressed by both IPD and APD meta-analyses. The need for defining an explicit and relevant clinical question, exhaustively searching for the totality of evidence, meticulous and unbiased data transfer or extraction, assessment of between study heterogeneity and the use of appropriate statistical methods for estimating summary effect measures are essentially the same for the two approaches. Conclusion: IPD offers advantages and, when feasible, should be considered the best opportunity to summarize the results of multiple studies. However, the resources, time and cooperation required for such studies will continue to limit their use in many important areas of clinical medicine which can be meaningfully and cost-effectively approached by properly performed APD meta-analyses. APD meta-analyses continue to be the mainstay of systematic reviews utilized by the US Preventive Services Task Force, the Cochrane Collaboration and many professional societies to support clinical practice guidelines. © 2005 Lyman and Kuderer; licensee BioMed Central Ltd.

引用

共 20 条

[1] Tierney J.F., Clarke M., Stewart L.A., Is there bias in the publication of individual patient data meta-analyses?, Int J Technol Assess Health Care, 16, pp. 657-667, (2000)
[2] Khan K.S., Bachmann L.M., Ter Riet G., Systematic reviews with individual patient data meta-analysis to evaluate diagnostic tests, Eur J Obstet Gynecol Repro Biol, 108, pp. 121-125, (2003)
[3] Stewart L.A., Tierney J.F., To IPD or Not to IPD? Advantages and Disadvantages of Systematic Reviews Using Individual Patient Data, Eval Health Prof, 25, pp. 76-97, (2002)
[4] Sylvester R., Collette L., Duchateau L., The role of meta-analyses in assessing cancer treatments, Eur J Cancer, 36, pp. 1351-1358, (2000)
[5] Piedbois P., Buyce M., Meta-analyses based on abstracted data: A step in the right direction, but only a first step, J Clin Oncol, 22, pp. 3839-3841, (2004)
[6] Burdett S., Stewart L.A., A Comparison of the Results of Checked Versus Unchecked Individual Patient Data Meta-Analyses, Int J Technol Assess Health Care, 18, pp. 619-624, (2002)
[7] Steinberg K.K., Smith S.J., Stroup D.F., Olkin I., Lee N.C., Williamson G.D., Thacker S.B., A comparison of effect estimates from a meta-analysis using individual patient data for ovarian cancer studies, Am J Epidemiology, 145, pp. 917-925, (1997)
[8] Olkin I., Sampson A., Comparison of Meta-Analysis Versus Analysis of Variance of Individual Patient Data, Biometrics, 54, pp. 317-322, (1998)
[9] Ioannidis J.P.A., Contopoulos-Ioannidis D.G., Lau J., Recursive Cumulative Meta-analysis: A Diagnostic for the Evolution of Total Randomized Evidence from Group and Individual Patient Data, J Clin Epidemiol, 52, pp. 281-291, (1999)
[10] Trikalinos T.A., Ioannidis J.P., Predictive modeling and heterogeneity of baseline risk in meta-analysis of individual patient data, J Clin Epidemiol, 54, 3, pp. 245-252, (2001)

← 1 2 →