Bibliometric analyses along cumulative meta-analyses : the Life Cycle of Evidence in Cancer Care (LIFE CYCLE)

Background: The LIFE CYCLE study aims to identify the key milestones in the life cycle of evidence of 9 novel chemotherapy agents, approved in Canada between 1992 and 2002 for metastatic breast and non small cell lung (NSCLC) cancers, using cumulative meta-analyses and bibliometric analyses. Objective: Bibliometrics is used to describe patterns of publications. This analysis has been used to complement the statistical and clinical data of the larger study. Methods: Several information sources were searched: i) Science Citation Index (SCI) for cited references of all included studies; ii) MEDLINE for editorials, commentaries and narrative reviews for information from opinion leaders; iii) Internet and MEDLINE for clinical guidelines; and iv) hand-search of general and cancer specific textbooks in oncology. We present analyses of vinorelbine in NSCLC as a case-study. Results: Nine studies were located for the meta-analysis pertaining to vinorelbine therapy published between 1994 and 2002, two studies presenting duplicate publications. 756 articles in SCI cited the 9 studies above. In MEDLINE, editorial, comment and narrative review articles were located, identifying 1 comment and 145 narrative reviews. Important recommendations about the use of vinorelbine were identified in nine textbooks from 1992-2002. Preliminary analysis identified the Le Chevalier study, first published in Feb 1994 (2 publications, 1994 and 1996), significantly supporting vinorelbine use, accounted for 509 (67%) of the cited references dominating the other trials in terms of research payback. Heath Canada and FDA approvals for vinorelbine were May and Dec 1994 respectively. In 1996 the citations were 10% (53/509) and peaked in 2001 (14%) when the Le Chevalier six year follow-up results were published. From 1996 onwards majority of textbooks reported vinorelbine therapy as standard first line treatment for NSCLC. Conclusion: This case study indicates that the life cycle of evidence in metastatic lung cancer may evolve more rapidly compared to Antman et als findings in treatments for myocardial infarction (1992). New agents are adopted rapidly and recommendations are published within 2 years of major research results. Bibliometric analyses alongside cumulative meta-analyses are useful in understanding the time it takes for research to feed into clinical recommendations.