NCT00681564

1. Castellani: "We are reaching out to groups such as the Institute of Medicine, the Harvard Multi-Regional Clinical Trial Center,” Response: - From Harvard: http://blogs.sph.harvard.edu/ashish-jha/<br>"an ounce of data is worth a thousand pounds of opinion" 2. "The biopharmaceutical industry is firmly committed to enhancing public health through responsible reporting and publication of clinical research and safety information. In the process of drug development, companies routinely publish their research, collaborate with academic researchers, and disclose clinical trial information at the time of patient registration, drug approval, and for medicines whose research programs have been discontinued." "Information on clinical trials for potential new medicines is already required by US law to be posted on ClinicalTrials.gov, the publicly accessible clearing house maintained by the National Institutes of Health (NIH). As of May 2013, ClinicalTrials.gov has information on 146 213 studies in all 50 states and in 185 countries. NIH reported last year that ClinicalTrials.gov “receives more than 95 million page views per month and 60 000 unique visitors daily.”" Response: - About the disclosure of clinical trial information - From the same data presented above, the following Figure and Table shows evidence about: "The biopharmaceutical industry is firmly committed to enhancing public health through responsible reporting and publication of clinical research and safety information." - Other clinical trial registries: EudraCT (Europe): No results = 23265 / Has results = 25 JAPIC (Japan): No results = 1498 / Has results = 291 Raw data of US phase 3 trials is publicly available and shown below according to what I think "data sufficiently shared" actually means. 3. " In addition, PhRMA has set out voluntary principles to fortify biopharmaceutical companies’ commitment to the highest standards for ethics and transparency in the conduct of clinical trials. PhRMA’s Principles on Conduct of Clinical Trials and Communication of Clinical Trial Results are designed to help ensure that clinical research conducted by biopharmaceutical research companies continues to protect patients and provide meaningful medical research results to healthcare professionals and patients." Response: 4. "Processes for data sharing or disclosure must take account of patients’ informed consent and the reality that re-identification of patients based on anonymized information is possible. Threats to patient privacy will jeopardize patient willingness to participate in clinical trials, which would delay the availability of new therapies." Response: - ClinicalTrials.gov search query: “genome” (July 30, 2014) = 0.6% (1083/172006) - Studies could choose to disclose other variables not related with genomic data: anonymized patient level raw data is possible for those studies (once again: 0.6%). - Patient willingness to participate on trials could be jeopardized if results of human studies are hidden from prescribers and healthcare systems around the world. - Red flags are frequently not raised when open data is available because: (1) Researchers are willing to correct themselves in cases were they detect any errors; (2) Researchers appreciate to be notified about any possible inconsistencies in their data, this is because: (3) open data does not raise red flags; (4) One example: I can see 1507 red flags, grouped by different color intensities and brightness, but 891 have a very particular red tonality. It'll be also possible to detect red flags hiding among 1450 white flags, as it was shown on the top 10 largest US phase 3 trials: it's not only about publishing a clinical trial, it's also about misreporting. - Please allow me to prove you that it is possible to share anonymized raw level patient data: NCT00681564.(1) <strong>5. Pending Table: analysis of unpublished studies US FDA novel drugs approved in 2013</strong> Raw data: 6. "Dumping millions of pages of clinical trial information into the public domain without providing appropriate scientific and clinical context or guidelines for meta-analysis could lead to second guessing of the expert decisions of national regulators worldwide, undermining patient trust and confidence in the safety and effectiveness of approved medicines." Response: - "Dumping" millions of numbers (data) could replace text and opinions I do not trust "expert opinions" due to several cases crossing my mind right now: rofecoxib, ticagrelor, statins, agomelatine, duloxetine, sertraline, paroxetine, liraglutide, exenatide, venlafaxine, gabapentin, pregabaline, dabigatran, rosiglitazone... I could go on... - I agree with your disagreement about clinical practice guidelines. - I also do not believe meta-analysis are at the top of the hierarchical pyramid of evidence because over half percent of clinical trials are unpublished. - I also do not think that clinical trials are superior to observational studies in their strength of evidence. I think we are forgetting the importance of observational studies and relying on clinical trials poorly registered (over 50% of US phase 3 trials were retrospectively registered), with insufficient disclosure of data, published many times (10 - 30 times) under the pretext of "subgroup analysis" while keeping hidden data conveniently on the shadows, out of the sight and attention by anybody that a "second guessing" could actually be accurate. - What is the percentage of unregistered clinical trials and how many will be retrospectively registered?A number could give us a clue = 1899 out of 2957 US phase 3 clinical trials were retrospectively registered (64.2%). Availability of results and prospective registration at ClinicalTrials.gov are related to publication of a clinical trial. References 1. Ramirez, Jorge H; R, Arce; Parra, Beatriz; Bravo, Luis E.; Gutierrez, Sonia Jackeline; Jaramillo, Adriana; Vivas, Juliana; Gonzalez, Viviana; JE, Soto; Otalvaro A; J, Consuegra; A, Contreras (2014): Periodontal therapy does not improve cardiovascular disease biomarkers: results of a randomized clinical trial. figshare. http://dx.doi.org/10.6084/m9.figshare.1098548 - See more at: http://figshare.com/preview/_preview/1121828#sthash.pmx78SsI.dpuf