How do I evaluate whether an oncology biotech news site is credible and unbiased about recent trial data?

To judge whether an oncology biotech news site is credible and reasonably unbiased about recent trial data, check it on three levels: source quality, reporting quality, and omission/bias signals.

1) Check the underlying source

A credible site should clearly tell you:

• Who wrote the article and their credentials
• What the original source was: journal paper, ASCO/ESMO abstract, company press release, FDA briefing, conference presentation, etc.
• Whether the data are primary or secondary
  Primary sources are better than press releases or reposted summaries.

Red flags:

• No author byline
• No link to the original abstract/paper
• Heavy reliance on company press releases with little independent verification

2) Assess how they report the trial

Good coverage should include the basics:

• Trial phase (1/2/3), design, randomization, blinding
• Patient population and line of therapy
• Primary endpoint(s) and whether they were met
• Absolute numbers and not just percentages or headline claims
• Statistical context: p-values, hazard ratios, confidence intervals, follow-up length
• Safety/toxicity data, not just efficacy
• Limitations: small sample size, immature OS, crossover, subgroup findings, surrogate endpoints

Red flags:

• Headlines like “game-changing,” “breakthrough,” “cures,” or “smashes expectations”
• Emphasis on only one positive subgroup or one cherry-picked endpoint
• No mention of adverse events, comparator arm, or maturity of data
• Confusing response rate with overall survival

3) Look for signs of bias

Bias often shows up in what they emphasize or leave out.

Possible pro-company bias:

• Uses promotional language
• Repeats stock-sounding phrases from company PR
• Quotes company executives more than independent experts
• Downplays missed endpoints or toxicity

Possible anti-company bias:

• Overstates weak signals as failures
• Focuses only on adverse events without context
• Ignores positive efficacy when data are genuinely strong

A balanced site should:

• Present both strengths and weaknesses
• Distinguish hypothesis-generating from practice-changing
• Avoid overclaiming before peer review or mature survival data

4) Compare with primary and independent sources

For any major trial, compare the news story against:

• The conference abstract/poster
• The full paper if published
• FDA/EMA documents if regulatory
• Other reputable outlets with different editorial angles
• Independent clinician or statistician commentary

If the site’s framing differs wildly from the data, be cautious.

5) Evaluate their track record

Over time, credible outlets tend to:

• Correct errors transparently
• Update stories when later data change interpretation
• Avoid sensational language
• Get trials right more often than not

You can test this by reviewing:

• Older stories on therapies later shown to fail
• Whether they updated early “positive” coverage after OS or confirmatory data matured

6) Check for conflicts of interest

Look for:

• Ad sponsorship from biotech/pharma
• Sponsored content labels
• Access journalism patterns
• Contributor relationships with companies or investor audiences

Funding doesn’t automatically invalidate reporting, but it raises the need for skepticism.

7) A quick credibility checklist

A site is more credible if it:

• Names authors and sources
• Links to primary data
• Reports endpoints, statistics, and limitations
• Includes toxicity and context
• Uses restrained language
• Corrects itself publicly
• Separates news from opinion or sponsorship

Practical rule of thumb

If a story makes a trial sound definitive, ask:

1. What was the actual endpoint?
2. How big was the effect, and in whom?
3. Was the result statistically and clinically meaningful?
4. What were the harms?
5. What do independent experts say?
6. Is this from a press release, abstract, or peer-reviewed paper?

If you want, I can also give you a one-page scoring rubric you can use to rate any oncology biotech news article for credibility and bias.