Evidence for the Selective Reporting of Analyses and Discrepancies in Clinical Trials: A Systematic Review of Cohort Studies of Clinical Trials

Abstract

In a systematic review of cohort studies, Kerry Dwan and colleagues examine the evidence for selective reporting and discrepancies in analyses between journal publications and other documents for clinical trials. Please see later in the article for the Editors' Summary

Why it can be ethical to use placebos in clinical practice

“…For people considering a statin for primary prevention, reasons for not wishing to take it can range from a feeling that one is already consuming too many tablets; that friends or neighbours have had problems; that one would prefer to avoid side effects, visits to the doctor or pharmacist, or, as one person as put it to me 'I'd rather die of a heart attack than cancer' . Or, in choosing between trials are published, meaning that even the most thorough systematic reviews are capable of reaching misleading conclusions [2,3]. This is not the only bias, but may be ethically challenging.…”

Fostering EMA's transparency policy

“…First, medical research is a public good as it involves the efforts of participants, researchers, medical staff, not to mention tax-payers who fund most of the researchers concerned and the infrastructures and facilities supporting clinical research. Second, data-sharing reduces the potential for incomplete reporting of study outcomes, encourages the exploration of new research questions, allows more accurate data scrutiny and pooling, thus improving the medical evidence base and, in the end, clinical decision-making [4][5][6][7][8].…”

“…First, medical research is a public good as it involves the efforts of participants, researchers, medical staff, not to mention tax-payers who fund most of the researchers concerned and the infrastructures and facilities supporting clinical research. Second, data-sharing reduces the potential for incomplete reporting of study outcomes, encourages the exploration of new research questions, allows more accurate data scrutiny and pooling, thus improving the medical evidence base and, in the end, clinical decision-making [4][5][6][7][8].Third, secondary analyses of study data have to rely on complete records of the planning, execution, and results of clinical trials, including individual participant-level data. The aggregate, or summary data used to report clinical trial results in scientific journals are not always a valid reflection of the underlying data or an adequate source for secondaryanalyses, including meta-analyses [9].…”

“…allows more accurate data scrutiny and pooling, thus improving the medical evidence base and, in the end, clinical decision-making [4][5][6][7][8].…”

Subgroup analyses in randomized trials—more rigour needed

“…Secondly, results of subgroup analyses are reported selec tively and such analyses are often not pre specified, resulting in frequent discrepan cies between trial protocols and subsequent published trial reports. 3,4 How subgroup analyses are performed is another key issue. The important question is whether the observed difference between the treat ments (the 'treatment effect') varies impor tantly between the subgroups.…”