Positive outcomes influence the rate and time to publication, but not the impact factor of publications of clinical trial results.
Suñé P, Suñé JM, Montoro JB. · 2013
Read Positive outcomes influence the rate and time to publication, but not the impact factor of publications of clinical trial results. by Suñé P, Suñé JM, Montoro JB. (2013) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc3559840.txt.
Introduction
Pilar Suñé, Josep Maria Suñé, and Bruno Montoro studied publication bias in PLoS ONE in 2013, asking whether a drug trial's outcome shapes whether its results ever appear in print. Their source material was every drug trial approved by one general hospital ethics committee between 1997 and 2004, followed forward to see which ones reached publication and how fast.
Fewer than half of the 785 trials were published. The outcome clearly mattered. About 85 percent of trials with positive results were published, against about 69 percent of trials with negative results. Positive trials also appeared sooner, in roughly two years at the median. Journal prestige, measured by impact factor, did not differ by outcome. The findings warn readers that the printed record favors good news about new drugs.
Catalog record
| CREATOR | Suñé P, Suñé JM, Montoro JB. |
|---|---|
| YEAR | 2013 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 33 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Positive Outcomes Influence the Rate and Time to Publication, but Not the Impact Factor of Publications of Clinical Trial Results By Pilar Suñé, Josep Maria Suñé, J. Bruno Montoro, Roberta W. Scherer Published in: PLoS ONE DOI: 10.1371/journal.pone.0054583 ABSTRACT Objectives Publication bias may affect the validity of evidence based medical decisions. The aim of this study is to assess whether research outcomes affect the dissemination of clinical trial findings, in terms of rate, time to publication, and impact factor of journal publications. Methods and Findings All drug-evaluating clinical trials submitted to and approved by a general hospital ethics committee between 1997 and 2004 were prospectively followed to analyze their fate and publication. Published articles were identified by searching Pubmed and other electronic databases. Clinical study final reports submitted to the ethics committee, final reports synopses available online and meeting abstracts were also considered as sources of study results. Study outcomes were classified as positive (when statistical significance favoring experimental drug was achieved), negative (when no statistical significance was achieved or it favored control drug) and descriptive (for non-controlled studies). Time to publication was defined as time from study closure to publication. A survival analysis was performed using a Cox regression model to analyze time to publication. Journal impact factors of identified publications were recorded. Publication rate was 48·4% (380/785). Study results were identified for 68·9% of all completed clinical trials (541/785). Publication rate was 84·9% (180/212) for studies with results classified as positive and 68·9% (128/186) for studies with results classified as negative (p<0·001). Median time to publication was 2·09 years (IC95 1·61–2·56) for studies with results classified as positive and 3·21 years (IC95 2·69–3·70) for studies with results classified as negative (hazard ratio 1·99 (IC95 1·55–2·55). No differences were found in publication impact factor between positive (median 6·308, interquartile range: 3·141–28·409) and negative result studies (median 8·266, interquartile range: 4·135–17·157). Conclusions Clinical trials with positive outcomes have significantly higher rates and shorter times to publication than those with negative results. However, no differences have been found in terms of impact factor. BACKGROUND Evidence-based medical decisions rely on the absence of bias in the available published results. When the publication or not of a particular study is determined by its outcomes or statistical significance, the derived bias is unacceptable, since it results in an exaggeration of the positive effects of a given treatment [1], [2]. Relevance of publication bias, in any of its manifestations, is undoubtedly more serious when it relates to clinical trials involving treatments in more advance phases of development, newly approved drugs, or new indications for registered drugs, since it may directly result in the use of less cost-effective, ineffective, or even harmful interventions in clinical practice [1]. Efforts have been made to prevent and reduce publication bias, including prospective registration of clinical trials [3], [4], publication of results in public databases [5] or in sponsor websites [6], and journal editors’ initiatives to promote trial registration [7]. However, a single universal tool to provide access to all clinical trial results, regardless of their outcomes, is not currently available. Previous research has documented the existence of lower publication rates [8]–[12] and delay of publication [9], [13] in studies with negative outcomes in comparison with studies with positive outcomes. However, to our knowledge, no previous studies have assessed publication bias in a large cohort of drug-related clinical trials with known results, both positive and negative. Direction and strength of outcomes might also affect the availability and extent of dissemination of study results as measured by article impact factor. Few studies have assessed the effect of study outcomes on impact factors, showing a tendency to publish studies with positive results in high impact factor journals [14], [15]. The aim of this study was to assess whether study outcomes affect the dissemination of clinical trials findings, in terms of rate, time to publication, and journal impact factor, in a cohort of clinical trials with known results, publically available or not. Furthermore, to evaluate the extent of publication bias in trials with a greater relevance on clinical practice, two subpopulation analyses were performed, with phase 3 and 4 clinical trials, and with trials in which conditions of use of experimental drugs had been approved, before or after trial completion. METHODS Source Data All drug-evaluating clinical trials submitted to and approved by the Ethics Committee (EC) of Hospital Vall d’Hebron in Barcelona, Spain, between 1997 and 2004 were prospectively followed to analyze their fate, results if available, and subsequent publication. The following information was recorded: experimental drug name, medical condition, name and specialty of principal investigator, main outcome, sample size, type of sponsor (industry or no industry), phase of study, study design and dates of EC approval, trial initiation and closure. Conditions of use of experimental drugs were classified as approved or off label, according to label information at the time of trial submission. Current status of the trial was recorded at the time of database closure (February 2010), categorized as not initiated, completed, prematurely terminated, or ongoing. Final reports submitted to the EC by the sponsor were also prospectively collected and filed for ulterior analysis of results. Study Results Search Strategy Completed and prematurely terminated studies were classified as published when at least one article was identified in peer reviewed literature. Other sources of results, if found, were also recorded: meeting abstracts, online final report synopses, and final reports submitted by the sponsor to the EC. We searched clinicaltrials.gov site to identify registered trials and study publications, if mentioned, using drug name, medical condition, sponsor identification and protocol number as keywords. Additionally, literature searches were performed to identify trial publications in the following databases of biomedical journals: Pubmed, ISI web of knowledge and Cochrane Library Plus. Keywords used in the search were drug name, medical condition, sponsor name, principal investigator, and clinicaltrials.gov identification number, when available. If no hits were found in the first strategy, a second search was performed, using only drug name and medical condition as keywords. We also reviewed Spanish language databases, Indice Médico Español and the Spanish Ministry of Education doctoral thesis database (TESEO) for additional publications. Criteria used to match identified publications to trials were classified as fixed, which should be included in the publication according to CONSORT guidelines [16], and optional, when their inclusion in the article was per author decision. Fixed criteria included identification of tested drugs (dosage and schedules), study population (inclusion criteria and medical condition), sample size, main outcome, statistical analysis and sponsor or funding source. Optional criteria were protocol number or clinicaltrials.gov identification, and presence of the hospital and principal investigator name in the list of participants. We also checked sponsor websites and the online database www.clinicalstudyresults.org to locate study final report synopses, using drug name, medical condition, protocol number, and clinicaltrials.gov identification number, when available, as keywords. Last search was performed during March 2010. The following information was collected for identified publications: date of publication, journal name, and journal impact factor, according to the 2008 list of journal impact factors [17]. Classification of Study Outcomes For each trial, identified result sources were recorded and quantified. Results were classified as: 1) complete, 2) partial, when analysis of results was preliminary or incomplete,