How is sensitive data protected?

All documents are encrypted with AES-256-GCM at rest and TLS 1.3 in transit. A GDPR anonymization layer removes PII before any AI processing. Your data never leaves our SOC 2 certified infrastructure.

How long does integration take?

Less than 30 minutes. Upload documents directly, connect via OAuth for Slack, or use our REST API for bulk processing.

How does outcome tracking work?

Outcomes are detected automatically from follow-up documents, Slack messages, and web intelligence. You confirm with one click. Each reported outcome makes your future analyses more accurate.

How is Decision Intel different from ChatGPT?

ChatGPT gives one opinion from one model. Decision Intel uses 3 independent judges for noise measurement, a 20×20 bias interaction matrix for compound scoring, 31 domain-specific biases, and an outcome flywheel that gets smarter with every decision.

All case studies

Catastrophic failureHealthcare2004

Merck

Vioxx Withdrawal

Estimated impact: $4.85B

F17

Decision Quality Index

Critical — 5 biases detected, 1 toxic combinations

Merck voluntarily withdrew its blockbuster painkiller Vioxx after a study confirmed it doubled the risk of heart attacks and strokes. Internal documents later revealed Merck scientists had identified cardiovascular risks years earlier but selectively presented data to regulators and physicians to protect $2.5 billion in annual revenue.

Decision context

Whether to investigate and disclose emerging cardiovascular risk signals from clinical trials or continue marketing Vioxx while designing studies to minimize unfavorable findings.

Deep hindsight-stripped analysis coming soon. This case is part of our ongoing backfill to reconstruct what the platform would have flagged at decision time. The summary and bias profile below are drawn from primary sources.

Biases present in the decision

★ Confirmation BiasSelective PerceptionGroupthinkLoss AversionAuthority Bias

3D Decision Knowledge Graph

Merck — 5 biases, 1 toxic combination

Drag to rotate · Scroll to zoom · Click to explore

Decision

Critical

High

Medium

Toxic

Bias Intensity Profile

Severity × confidence across 5 detected biases

Toxic combinations

Echo Chamber

Reference class base rates

Across all 146 curated case studies in our library:

90%

base failure rate

10%

base success rate

Lessons learned

Selective perception of clinical trial data to support a desired safety narrative can cause widespread patient harm.
Confirmation bias in pharmaceutical research leads to study designs that are more likely to produce favorable results rather than accurate ones.
Post-market surveillance systems must be independent of commercial interests to detect safety signals early.

Source: FDA NDA 21-042/S-007; Merck SEC filings; Bombardier et al., NEJM (2000) (FDA Action)

Pilot slots available

See what we'd flag in your next strategic memo.

Upload a strategic memo or board deck. Get the same bias audit you just saw for Merck, on your own high-stakes call, in under 60 seconds.

Try the Demo — Free Request a Pilot

Or leave your email, we'll run a strategic memo of your choosing and send the readout within a business day.

Ready to audit your own memo right now? Create a free account →

Related cases in Healthcare

Catastrophic failure1996

Purdue Pharma

Purdue Pharma OxyContin Marketing

Failure1970

Johnson & Johnson

Johnson & Johnson Talc Powder Litigation

Catastrophic failure2014

Theranos

Theranos Blood Testing Patient Safety Failures

Lessons learned

Selective perception of clinical trial data to support a desired safety narrative can cause widespread patient harm.

Confirmation bias in pharmaceutical research leads to study designs that are more likely to produce favorable results rather than accurate ones.

Post-market surveillance systems must be independent of commercial interests to detect safety signals early.