Authors' reports about research integrity problems in clinical trials

https://doi.org/10.1016/j.cct.2004.11.013Get rights and content

Abstract

Background

There is little information about the prevalence of research integrity problems in the scientific literature. We sought to determine how frequently authors of published pharmaceutical clinical trials reported fabrication of data or misrepresentation of research.

Methods

We conducted a mail survey of 549 authors who had published reports of pharmaceutical clinical trials from 1998 to 2001 that appeared in the Cochrane Database of Systematic Reviews. We asked authors about fabricated data or misrepresentations of research in three contexts: the target study (the report from which their name was obtained), another study they had participated in, or a study that they personally knew about.

Results

We received replies from 64% of authors with valid addresses. Two authors (1%) reported that the target article misrepresented the research. Almost 5% reported fabrication or misrepresentation in a study they had participated in the last 10 years, and 17% of authors personally know about a case of fabrication or misrepresentation in the last 10 years from a source other than published accounts of research misconduct.

Conclusions

Fraud and misrepresentation in clinical trials appear to be rare on a per-published report basis. However, they occur sufficiently frequently that scientists have a significant chance of participating in a project affected by fraud or misrepresentation during their research careers. These rates of exposure justify vigorous efforts to prevent research misconduct.

Section snippets

Sample

We gathered a sample of authors from The Cochrane Database of Systematic Reviews, an online collection of meta-analytic reviews published by the Cochrane Collaboration [16], a coalition of clinical researchers who conduct research reviews according to a common set of methods. The reviews summarize all known reports on the effectiveness of a given drug or treatment. We had two reasons for gathering articles from Cochrane. First, the Cochrane reviewers saved us time because they had already

Results

Of the 549 authors mailed surveys, 45 surveys were returned with bad addresses, so we counted 504 surveys as having reached authors. We received 322 completed surveys, so our return rate was 59% of 549 authors, or 64% of 504 authors net of those not reached because of bad addresses.

Table 1 describes the authors who returned surveys. Most were MDs and identified themselves as “organizing PI” or “site PI.” The large proportion of authors who described themselves as “organizing PI” suggests that

Discussion

This study presents authors' reports of research integrity problems in three contexts. First, we found that 0.6% of authors reported that the article misrepresented the research it published. Although the confidence interval around this rate includes the rate reported by Weiss et al. [10], the results are difficult to compare, for two reasons. First, Weiss's methods detected only falsification, whereas our survey asked about falsification and misrepresentation. Second, Weiss's methods would

References (17)

There are more references available in the full text version of this article.

Cited by (51)

  • Open research practices in psychology

    2021, Psychologie Francaise
  • Ethical issues in research

    2017, Best Practice and Research: Clinical Obstetrics and Gynaecology
    Citation Excerpt :

    In one systematic review and meta-analysis, it was determined that 1.97% of scientists admitted to have fabricated, falsified, or modified data or results at least once; 33.7% admitted questionable research practices; and up to 72% admitted following other questionable research practices [30]. Another survey determined that 29% of the respondents were involved in misconduct that they were not aware of [31]. The basic requirements for ethical research should address the following [32]:

View all citing articles on Scopus

Grant support: the research was supported by the Office of Research Integrity Research Program on Research Integrity (R01 NS42449, PI: Gardner).

View full text