Association of conflicts of interest with the results and conclusions of goal-directed hemodynamic therapy research: a systematic review with meta-analysis.

PURPOSE: The association between conflicts of interest (COI) and study results or article conclusions in goal-directed hemodynamic therapy (GDHT) research is unknown.

METHODS: Randomized controlled trials comparing GDHT with usual care were identified. COI were classified as industry sponsorship, author conflict, device loaner, none, or not reported. The association between COI and study results (complications and mortality) was assessed using both stratified meta-analysis and mixed effects meta-regression. The association between COI and an article's conclusion (graded as GDHT-favorable, neutral, or unfavorable) was investigated using logistic regression.

RESULTS: Of the 82 eligible articles, 43 (53%) had self-reported COI, and 50 (61%) favored GDHT. GDHT significantly reduced complications on the basis of the meta-analysis of studies with any type of COI, studies declaring no COI, industry-sponsored studies, and studies with author conflict but not on studies with a device loaner. However, no significant relationship between COI and the relative risk (GDHT vs. usual care) of developing complications was found on the basis of meta-regression (p = 0.25). No significant effect of GDHT was found on mortality. COI had a significant overall effect (p = 0.016) on the odds of having a GDHT-favorable vs. neutral conclusion based on 81 studies. Eighty-four percent of the industry-sponsored studies had a GDHT-favorable conclusion, while only 27% of the studies with a device loaner had the same conclusion grade.

CONCLUSIONS: The available evidence does not suggest a close relationship between COI and study results in GDHT research. However, a potential association may exist between COI and an article's conclusion in GDHT research.