TY - JOUR
T1 - Commercial funding and estimated intervention effects in randomized clinical trials
T2 - Systematic review of meta-epidemiological studies
AU - Nejstgaard, Camilla Hansen
AU - Laursen, David Ruben Teindl
AU - Lundh, Andreas
AU - Hróbjartsson, Asbjørn
N1 - © 2022 The Authors. Research Synthesis Methods published by John Wiley & Sons Ltd.
PY - 2023/3
Y1 - 2023/3
N2 - We investigated to which degree commercial funding is associated with estimated intervention effects in randomized trials. We included meta-epidemiological studies with published data on the association between commercial funding and results or conclusions of randomized trials. We searched five databases and other sources. We selected one result per meta-epidemiological study, preferably unadjusted ratio of odds ratios (ROR), for example, odds ratio(commercial funding)/odds ratio(noncommercial funding). We pooled RORs in random-effects meta-analyses (ROR <1 indicated exaggerated intervention effects in commercially funded trials), subgrouped (preplanned) by study aim: commercial funding per se versus risk of commercial funder influence. We included eight meta-epidemiological studies (264 meta-analyses, 2725 trials). The summary ROR was 0.95 (95% confidence interval 0.85-1.06). Subgroup analysis revealed a difference (p = 0.02) between studies of commercial funding per se, ROR 1.06 (0.95-1.17) and studies of risk of commercial funder influence, ROR 0.88 (0.79-0.97). In conclusion, we found no statistically significant association between commercial funding and estimated intervention effects when combining studies of commercial funding per se and studies of risk of commercial funder influence. A preplanned subgroup analysis indicated that trials with high risk of commercial funder influence exaggerated intervention effects by 12% (21%-3%), on average. Our results differ from previous theoretical considerations and findings from methodological studies and therefore call for confirmation. We suggest it is prudent to interpret results from commercially funded trials with caution, especially when there is a risk that the funder had direct influence on trial design, conduct, analysis, or reporting.
AB - We investigated to which degree commercial funding is associated with estimated intervention effects in randomized trials. We included meta-epidemiological studies with published data on the association between commercial funding and results or conclusions of randomized trials. We searched five databases and other sources. We selected one result per meta-epidemiological study, preferably unadjusted ratio of odds ratios (ROR), for example, odds ratio(commercial funding)/odds ratio(noncommercial funding). We pooled RORs in random-effects meta-analyses (ROR <1 indicated exaggerated intervention effects in commercially funded trials), subgrouped (preplanned) by study aim: commercial funding per se versus risk of commercial funder influence. We included eight meta-epidemiological studies (264 meta-analyses, 2725 trials). The summary ROR was 0.95 (95% confidence interval 0.85-1.06). Subgroup analysis revealed a difference (p = 0.02) between studies of commercial funding per se, ROR 1.06 (0.95-1.17) and studies of risk of commercial funder influence, ROR 0.88 (0.79-0.97). In conclusion, we found no statistically significant association between commercial funding and estimated intervention effects when combining studies of commercial funding per se and studies of risk of commercial funder influence. A preplanned subgroup analysis indicated that trials with high risk of commercial funder influence exaggerated intervention effects by 12% (21%-3%), on average. Our results differ from previous theoretical considerations and findings from methodological studies and therefore call for confirmation. We suggest it is prudent to interpret results from commercially funded trials with caution, especially when there is a risk that the funder had direct influence on trial design, conduct, analysis, or reporting.
KW - Epidemiologic Studies
KW - Meta-Analysis as Topic
KW - Odds Ratio
KW - Randomized Controlled Trials as Topic
KW - meta-analyses
KW - bias
KW - meta-epidemiology
KW - randomized clinical trials
KW - commercial funding
UR - http://www.scopus.com/inward/record.url?scp=85143670745&partnerID=8YFLogxK
U2 - 10.1002/jrsm.1611
DO - 10.1002/jrsm.1611
M3 - Review
C2 - 36357935
SN - 1759-2879
VL - 14
SP - 144
EP - 155
JO - Research Synthesis Methods
JF - Research Synthesis Methods
IS - 2
ER -