Pharmaceutical companies which do not share Individual participant data confound science
Executive summary
Pharmaceutical companies’ refusal or failure to share individual participant data (IPD) does materially impede scientific verification, secondary analyses, and cumulative knowledge-building, but it does not exist in a vacuum: legitimate privacy, proprietary, and security concerns complicate a simple “share or science suffers” narrative [1] [2] [3]. Empirical audits show uneven and often poor IPD availability from industry-sponsored trials, and the literature acknowledges both clear benefits to openness and a persistent evidence gap about the net impact and best safeguards for IPD sharing [4] [1] [5].
1. The transparency deficit: how often industry keeps IPD out of reach
Multiple audits and scoping reviews document that industry commitments to share IPD have not uniformly translated into accessible data: analyses of high-impact journals and pharmaceutical-sponsored trials find single-digit to low-double-digit rates of IPD availability after publication, and company policies vary widely in scope and enforcement [4] [1] [2]. Recent quality-improvement work in oncology reinforces heterogeneity in what companies actually provide when asked for IPD, undermining expectations created by industry pledges [6] [7].
2. Why researchers argue IPD sharing matters to science
IPD enables independent verification of published analyses, more accurate pooled meta-analyses, better-powered subgroup and safety investigations, and the chance to test new hypotheses on existing trials—functions that advance cumulative science and patient care in ways summary results cannot [1] [2] [8]. When raw trial-level data are unavailable, replication is limited, safety signals can be missed or masked, and the scientific record becomes harder to correct or refine [1] [9].
3. Legitimate constraints: privacy, proprietary claims, and security risks
Pharma resistance is not purely defensive: large-scale sharing of detailed participant-level data raises real re‑identification and privacy risks, particularly when datasets can be linked with other public or commercial sources, prompting companies to withhold or heavily redact IPD to protect participants [3] [10]. Regulators and funders also recognize licensing or co‑funding arrangements that legitimately limit data dissemination, and data sovereignty and IP protection remain core business and compliance concerns [11] [12]. Cybersecurity threats and the high cost of breaches add another practical deterrent to broader, uncontrolled sharing [13].
4. When non‑sharing confounds science: documented harms and historical lessons
There are concrete examples where withheld or selectively reported internal analyses changed interpretation of safety or efficacy—litigation reviews of past cases show internal company analyses that conflicted with published conclusions, illustrating how restricted access can mislead clinicians and patients [9]. Without access to IPD, independent researchers cannot reliably verify results, perform robust pooled safety analyses, or pursue research questions that reduce waste and improve trial design, leaving knowledge gaps that can harm patient care and policy decisions [1] [2].
5. Tradeoffs, partial solutions, and emerging norms
The field is converging on middle paths: controlled-access platforms, risk‑based de‑identification, independent review committees, and formal data request processes aim to balance scientific utility with privacy and commercial interests, and industry alliances and policy updates (PhRMA/EFPIA, TransCelerate) have tried to create frameworks for this [7] [2]. Yet implementation is uneven and evidence on which sharing models maximize scientific value while minimizing harms remains limited, so policy and technical standards are still evolving [5] [7].
6. What remains uncertain and why it matters for accountability
High-level consensus accepts that IPD sharing can accelerate discovery and safeguard patients, but systematic, high-quality evidence quantifying the incremental scientific benefit of different IPD sharing models—and the operational costs or risks to participants and firms—is still lacking, making assertions that non‑sharing always “confounds science” too absolutist without nuanced appraisal of context and safeguards [5] [4]. Accountability mechanisms—clearer company reporting, independent post‑publication audits, and funder/regulator enforcement—are the levers most likely to align industry practice with scientific needs while addressing privacy, IP, and security concerns [6] [11].