Why Conflicts Matter to Participants, Data and the Site

Part of Conflict of Interest in Clinical Research: What a Conflict of Interest Is and Why It Matters.

About 5 minutes. Free, in your language, no account needed.

Expert review pending. CRLN wrote this lesson against the primary sources listed under References. A subject-matter expert has not yet reviewed it. Where your institution's policy or your sponsor's instructions are stricter, follow them.

Why this matters

Conflicts rarely look dramatic. Nobody at a site sets out to falsify data because they own shares. The harm comes from small, honest-seeming decisions that all lean the same way. This lesson walks through where those decisions happen and what an undisclosed conflict costs the people who rely on the trial.

Where bias can enter at a site

Think about the points in a trial where a person's judgment shapes what happens:

Each of these can tilt without anyone noticing, including the person doing the tilting. That is why the rules focus on disclosure and structure rather than on trusting individuals to correct for their own interests.

What it costs participants

Participants trust the site to put their welfare ahead of the trial's success. ICH E6(R3) says that neither the investigator nor site staff should coerce or unduly influence a participant to take part or to continue (Annex 1 §2.8.3). In the United States, FDA's consent rule requires that consent be sought only in circumstances that minimize the possibility of coercion or undue influence (21 CFR 50.20).

A conflict can undermine both. A person who learns later that the doctor who urged them to stay in the study held a royalty on the drug may feel, with reason, that their choice was not fully free. That loss of trust reaches beyond one participant: it is one of the reasons people decline research altogether.

What it costs the data

Regulators judge whether trial data can be relied on. FDA's financial disclosure rule states its purpose plainly: FDA may consider a study's data inadequate if steps were not taken to minimize bias, and one potential source of bias is a financial interest of the clinical investigator in the outcome (21 CFR 54.1).

When a disclosable interest exists, FDA weighs its size and nature and the steps taken to reduce bias. Design features such as multiple investigators, blinding, objective endpoints, and measurement of endpoints by someone other than the investigator can help (21 CFR 54.5(a) and (b)). If FDA decides an interest raises a serious question about the integrity of the data, it can audit the investigator's data, ask for further analyses, ask for additional independent studies, or refuse to treat the study as a basis for its decision (21 CFR 54.5(c)). FDA may also refuse to file a marketing application that lacks the required financial information (21 CFR 54.4(c)).

In other words, an undisclosed or unmanaged conflict can make a site's work, and the participants' contribution, count for less.

What it costs the site

For federally funded research, the consequences are spelled out. Where the US Department of Health and Human Services finds that a federally funded clinical study of the safety or effectiveness of a drug, device or treatment was designed, conducted or reported by an investigator with an unmanaged, unreported financial conflict, the institution must require the investigator to disclose the conflict in each public presentation of the results and to request an addendum to previously published presentations (42 CFR 50.606(c)).

Beyond the rules, a site's reputation with sponsors, IRBs and its own community depends on being seen as a place where judgment is independent. A single undisclosed conflict found during an audit raises questions about every other decision.

In practice

A sub-investigator is rating a pain scale at a follow-up visit. She recently accepted consulting work from the sponsor and has disclosed it. Her institution's management plan says another qualified rater performs the primary efficacy ratings for her patients. She hands the rating to her colleague, documents the handover as the plan requires, and continues with the safety review she is still allowed to do.

What this looks like when it goes wrong

An investigator holds stock options in a small private company sponsoring a device trial. He does not disclose them. He enrolls several borderline participants and rates outcomes himself. An inspection finds the options. Every outcome he rated is now in question, the sponsor must explain the gap to FDA, and participants who gave their time may have contributed to data that cannot be used.

Key takeaways

References