Risk of bias in industry-funded oseltamivir trials: Comparison of core reports versus full clinical study reports

Tom Jefferson, Mark A. Jones, Peter Doshi, Chris B. Del Mar, Rokuro Hama, Matthew J. Thompson, Igho Onakpoya, Carl J. Heneghan

Research output: Contribution to journalArticleResearchpeer-review

12 Citations (Scopus)

Abstract

Background: The Cochrane risk of bias tool is a prominent instrument used to evaluate potential biases in clinical trials. In three updates of our Cochrane review on neuraminidase inhibitors, we assessed risk of bias on the same trials using different levels of detail: the trials in journal publications, in core reports, and in full clinical study reports. Here we analyse whether progressively greater amounts of information and detail in full clinical study reports (including trial protocols, statistical analysis plans, certificates of analyses, individual participant data listings and randomisation lists) affected our risk of bias assessments.

Methods and findings: We used the Cochrane risk of bias tool to assess and compare risk of bias in 14 oseltamivir trials (reported in 10 clinical study reports) obtained from the European Medicines Agency (EMA) and the manufacturer, Roche. With more detailed information, reported in clinical study reports, no previous assessment of 'high' risk of bias was reclassified as 'low' or 'unclear' in the main analysis, and over half (55%, 34/62) of the previous assessments of 'low' risk of bias were reclassified as 'high'. Most assessments of 'unclear' risk of bias (67%, or 28/42) were reclassified as 'high' risk of bias when our judgements were based on full clinical study reports. The limits of our study were our relative inexperience in dealing with large information sets, sometimes subjective bias judgements and focus on industry trials. Comparison with journal publications was not possible because of the low number of trials published.

Conclusions: We found that as information increased in the document, this increased our assessment of bias. This may mean that risk of bias has been insufficiently assessed in Cochrane reviews based on journal publications.

Original languageEnglish
Article numbere005253
JournalBMJ Open
Volume4
Issue number9
DOIs
Publication statusPublished - 2014

Fingerprint

Oseltamivir
Industry
Publications
Neuraminidase
Clinical Protocols
Random Allocation
Clinical Studies
Clinical Trials

Cite this

Jefferson, Tom ; Jones, Mark A. ; Doshi, Peter ; Del Mar, Chris B. ; Hama, Rokuro ; Thompson, Matthew J. ; Onakpoya, Igho ; Heneghan, Carl J. / Risk of bias in industry-funded oseltamivir trials : Comparison of core reports versus full clinical study reports. In: BMJ Open. 2014 ; Vol. 4, No. 9.
@article{ad987e7f7b264bbdb9641ae85c00ff44,
title = "Risk of bias in industry-funded oseltamivir trials: Comparison of core reports versus full clinical study reports",
abstract = "Background: The Cochrane risk of bias tool is a prominent instrument used to evaluate potential biases in clinical trials. In three updates of our Cochrane review on neuraminidase inhibitors, we assessed risk of bias on the same trials using different levels of detail: the trials in journal publications, in core reports, and in full clinical study reports. Here we analyse whether progressively greater amounts of information and detail in full clinical study reports (including trial protocols, statistical analysis plans, certificates of analyses, individual participant data listings and randomisation lists) affected our risk of bias assessments.Methods and findings: We used the Cochrane risk of bias tool to assess and compare risk of bias in 14 oseltamivir trials (reported in 10 clinical study reports) obtained from the European Medicines Agency (EMA) and the manufacturer, Roche. With more detailed information, reported in clinical study reports, no previous assessment of 'high' risk of bias was reclassified as 'low' or 'unclear' in the main analysis, and over half (55{\%}, 34/62) of the previous assessments of 'low' risk of bias were reclassified as 'high'. Most assessments of 'unclear' risk of bias (67{\%}, or 28/42) were reclassified as 'high' risk of bias when our judgements were based on full clinical study reports. The limits of our study were our relative inexperience in dealing with large information sets, sometimes subjective bias judgements and focus on industry trials. Comparison with journal publications was not possible because of the low number of trials published.Conclusions: We found that as information increased in the document, this increased our assessment of bias. This may mean that risk of bias has been insufficiently assessed in Cochrane reviews based on journal publications.",
author = "Tom Jefferson and Jones, {Mark A.} and Peter Doshi and {Del Mar}, {Chris B.} and Rokuro Hama and Thompson, {Matthew J.} and Igho Onakpoya and Heneghan, {Carl J.}",
year = "2014",
doi = "10.1136/bmjopen-2014-005253",
language = "English",
volume = "4",
journal = "BMJ Open",
issn = "2044-6055",
publisher = "BMJ Publishing Group",
number = "9",

}

Risk of bias in industry-funded oseltamivir trials : Comparison of core reports versus full clinical study reports. / Jefferson, Tom; Jones, Mark A.; Doshi, Peter; Del Mar, Chris B.; Hama, Rokuro; Thompson, Matthew J.; Onakpoya, Igho; Heneghan, Carl J.

In: BMJ Open, Vol. 4, No. 9, e005253, 2014.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Risk of bias in industry-funded oseltamivir trials

T2 - Comparison of core reports versus full clinical study reports

AU - Jefferson, Tom

AU - Jones, Mark A.

AU - Doshi, Peter

AU - Del Mar, Chris B.

AU - Hama, Rokuro

AU - Thompson, Matthew J.

AU - Onakpoya, Igho

AU - Heneghan, Carl J.

PY - 2014

Y1 - 2014

N2 - Background: The Cochrane risk of bias tool is a prominent instrument used to evaluate potential biases in clinical trials. In three updates of our Cochrane review on neuraminidase inhibitors, we assessed risk of bias on the same trials using different levels of detail: the trials in journal publications, in core reports, and in full clinical study reports. Here we analyse whether progressively greater amounts of information and detail in full clinical study reports (including trial protocols, statistical analysis plans, certificates of analyses, individual participant data listings and randomisation lists) affected our risk of bias assessments.Methods and findings: We used the Cochrane risk of bias tool to assess and compare risk of bias in 14 oseltamivir trials (reported in 10 clinical study reports) obtained from the European Medicines Agency (EMA) and the manufacturer, Roche. With more detailed information, reported in clinical study reports, no previous assessment of 'high' risk of bias was reclassified as 'low' or 'unclear' in the main analysis, and over half (55%, 34/62) of the previous assessments of 'low' risk of bias were reclassified as 'high'. Most assessments of 'unclear' risk of bias (67%, or 28/42) were reclassified as 'high' risk of bias when our judgements were based on full clinical study reports. The limits of our study were our relative inexperience in dealing with large information sets, sometimes subjective bias judgements and focus on industry trials. Comparison with journal publications was not possible because of the low number of trials published.Conclusions: We found that as information increased in the document, this increased our assessment of bias. This may mean that risk of bias has been insufficiently assessed in Cochrane reviews based on journal publications.

AB - Background: The Cochrane risk of bias tool is a prominent instrument used to evaluate potential biases in clinical trials. In three updates of our Cochrane review on neuraminidase inhibitors, we assessed risk of bias on the same trials using different levels of detail: the trials in journal publications, in core reports, and in full clinical study reports. Here we analyse whether progressively greater amounts of information and detail in full clinical study reports (including trial protocols, statistical analysis plans, certificates of analyses, individual participant data listings and randomisation lists) affected our risk of bias assessments.Methods and findings: We used the Cochrane risk of bias tool to assess and compare risk of bias in 14 oseltamivir trials (reported in 10 clinical study reports) obtained from the European Medicines Agency (EMA) and the manufacturer, Roche. With more detailed information, reported in clinical study reports, no previous assessment of 'high' risk of bias was reclassified as 'low' or 'unclear' in the main analysis, and over half (55%, 34/62) of the previous assessments of 'low' risk of bias were reclassified as 'high'. Most assessments of 'unclear' risk of bias (67%, or 28/42) were reclassified as 'high' risk of bias when our judgements were based on full clinical study reports. The limits of our study were our relative inexperience in dealing with large information sets, sometimes subjective bias judgements and focus on industry trials. Comparison with journal publications was not possible because of the low number of trials published.Conclusions: We found that as information increased in the document, this increased our assessment of bias. This may mean that risk of bias has been insufficiently assessed in Cochrane reviews based on journal publications.

UR - http://www.scopus.com/inward/record.url?scp=84907560908&partnerID=8YFLogxK

U2 - 10.1136/bmjopen-2014-005253

DO - 10.1136/bmjopen-2014-005253

M3 - Article

VL - 4

JO - BMJ Open

JF - BMJ Open

SN - 2044-6055

IS - 9

M1 - e005253

ER -