Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
The cost of incomplete research
- Hiten R Patel, Ciaran S. Hill, Senthy Sellaturay and Manit Arya (9 June 2009)
|
|
|||
|
Hiten R Patel, Consultant Laparoscopic Uro-Oncological Surgeon University College Hospital, London, NW1 2BU, Ciaran S. Hill, Senthy Sellaturay and Manit Arya
Send response to journal:
|
We read with interest the article on funding of medical research by Waters (BMJ2009:338:b2177). It is particularly important at this time of financial hardship that research is well designed and holds the promise of producing results that could benefit science and medicine. In the age of evidence based medicine and protocol/guideline driven medicine we noted that many articles concluded that “further or larger or randomized” studies were needed. The value of research is not to be found in either positive or negative findings, nor in the novelty of the study (although original results are likely to always outweigh repetition). Instead it is sine qua non that the research demonstrates something. A statement that a particular drug has or has not given effect has concrete value in intellectual progression but a statement that a study into a drug’s effect is inconclusive, neither providing the means to accept an association or reject it is nothing more than a cautionary tale of an inadequate study. From our observations, we reviewed the research findings in each of four major journals; New England Journal of Medicine (NEMJ), British Medical Journal (BMJ), Journal of the American Medical Association (JAMA) and The Lancet over a 12 month period in the year 2005 to 2006. Data was collected by our team who applied uniform criteria to determine the total number of new research articles presented and how many of these stated that further or repeat study was required to confirm their findings. This included suggestions that larger scale or randomized controlled trials were necessary. Statements that suggested further research to explore findings or investigate new avenues identified by the research were not included. Articles that suggested more research on the basis of assessing financial or costing issues were also excluded as these were deemed unrelated to intellectual advancement per se. Findings of our search: -British Medical Journal (BMJ)- of 184 original articles there were 68 that required further study (37%) -Journal of the American Medical Association (JAMA)- of 177 original articles there were 55 that required further study (31%) -The Lancet- of 184 original articles there were 68 that required further study (37%) -New England Journal of Medicine (NEJM)- of 195 original articles there were 60 that required further study (31%) The data collected was assessed by another group from the collection team. The details above summarize our findings and clearly demonstrates the high numbers of articles that are still being published in each of these major medical journals that self-report incomplete findings. The responsibility for improving the position of research lies with all of us. Researchers must perform adequate power calculations before instigating research and be prepared to accept that data that cannot prove or disprove their null hypothesis is inadequate and should not be submitted without further correction1. Likewise peer-reviewers and publishers have the difficult job of monitoring this process and rejecting articles that fail to reach conclusions, regardless of how important the topic of interest2. We appreciate the difficulties and suggest larger research networks with collaborative working may be a solution and ergo help reduce costly incomplete research. 1. Altman D. Poor-Quality Medical Research What Can Journals Do? JAMA; 2002:2765-7. 2. Young S. Peer review of manuscripts: theory and practice. Journal of Psychiatry and Neuroscience 2003;28:327-30. Competing interests: None declared |
|||