intermediate14 min read

How to Read Clinical Trial Data: A Plain-English Guide for Investors

A biotech company announces Phase 3 results. The press release is full of numbers: p-values of 0.0001, hazard ratios of 0.58, median progression-free survival of 14.2 months. The stock is moving fast. Do you buy? Sell? Hold? Most retail investors either panic, follow the crowd, or freeze. The edge goes to investors who can actually read the data — who know within minutes whether the results are genuinely strong, borderline, or disappointing regardless of how the company spins them. This guide will teach you how to read clinical trial results like a biotech analyst. Not at a PhD level — you don't need that. At an investor level: enough to make informed decisions about what the data means for the stock. We'll cover endpoints, statistical significance, effect size, safety signals, and the specific red flags that experienced investors watch for.

Understanding endpoints: what the trial is actually measuring

Every clinical trial has a primary endpoint — the specific measurement used to determine whether the drug works. If the drug hits its primary endpoint, the trial is considered a success. If it misses, it's a failure. Everything else is secondary. The most common primary endpoints in biotech trials include: Overall Survival (OS): How long patients live. The gold standard for cancer trials. If a drug improves overall survival, the FDA almost always approves it. Downside: OS trials take years because you have to wait for enough patients to die. Progression-Free Survival (PFS): How long until the disease gets worse. Very common in oncology. Faster to measure than OS, but the FDA sometimes questions whether PFS improvements translate to patients actually living longer. Objective Response Rate (ORR): The percentage of patients whose tumors shrank by at least 30%. Used in early and accelerated approval settings. A high ORR is encouraging but doesn't guarantee the drug extends life. HbA1c reduction: For diabetes drugs, the change in average blood sugar levels over 3 months. ACR response rates: For rheumatoid arthritis drugs, measuring improvement in symptoms. The critical thing for investors: read the press release to find what the primary endpoint was and whether the trial met it. Companies will trumpet positive secondary endpoints even when the primary endpoint fails. Don't be fooled. A trial that misses its primary endpoint is a failed trial in the eyes of the FDA, regardless of how many secondary endpoints look promising.

Key takeaway

Always identify the primary endpoint first. A trial that meets its primary endpoint is a success; one that misses it is a failure — no matter how the company spins the secondary data.

Example

When Bristol-Myers Squibb reported results for Opdivo in first-line lung cancer, the primary endpoint was overall survival. The trial met this endpoint with a hazard ratio of 0.66 (meaning 34% reduction in risk of death), which was a clear win. If the primary endpoint had been response rate instead and OS was just a secondary endpoint, the regulatory and investment implications would have been very different.

P-values: what 'statistically significant' actually means for your portfolio

The p-value is the number that determines whether a clinical trial result is real or might be due to chance. In biotech investing, p-values are the gatekeeper between a stock doubling and a stock crashing. Here's the simple version: a p-value measures the probability that the observed result would occur by random chance if the drug didn't actually work. A p-value of 0.05 means there's a 5% chance the result is a fluke. A p-value of 0.001 means there's a 0.1% chance. The standard threshold for statistical significance in clinical trials is p < 0.05. If the p-value is below 0.05, the result is considered statistically significant. If it's above 0.05, the trial technically failed to demonstrate the drug works. What investors need to know about p-values: P < 0.001 is very strong. This is rock-solid data. The FDA will have a hard time arguing the drug doesn't work. P < 0.01 is strong. Comfortably significant. P < 0.05 is significant but just barely. The drug works, but not with overwhelming evidence. The FDA may have questions. P = 0.05 to 0.10 is the danger zone. Technically not significant, but close enough that the company will argue it's clinically meaningful. Stocks in this range are volatile because the FDA's decision becomes unpredictable. P > 0.10 means the trial failed. The drug did not demonstrate a statistically significant effect. Critical nuance: a low p-value tells you the result is real, but it doesn't tell you the result is large enough to matter clinically. A drug could have a p-value of 0.0001 but only extend survival by 2 weeks. That's statistically significant but clinically questionable — and the stock may not react the way you expect.

Key takeaway

P-values below 0.05 mean the result is statistically significant. But a low p-value alone isn't enough — you also need to check whether the effect size is clinically meaningful. The best biotech data has both a very low p-value AND a large treatment effect.

Example

When MindMed reported Phase 2b results for its anxiety drug, investors immediately looked at the p-value on the primary endpoint. A p-value of 0.003 would signal strong efficacy. A p-value of 0.06 would mean the trial technically missed — even if the drug showed some improvement. That decimal point difference can mean billions in market cap.

Ready to start? Open a Webull account and trade biotech stocks commission-freevia Webull · Affiliate

Effect size: does the drug work well enough to matter?

After checking the p-value, the next question is: how big is the treatment effect? A drug can be statistically significant but not clinically meaningful. Investors who understand effect size have a major edge. Hazard Ratio (HR) is the most important metric in oncology trials. A hazard ratio compares the rate of events (death or disease progression) between the drug group and the control group. HR = 1.0 means no difference. HR < 1.0 means the drug is better. HR > 1.0 means the drug is worse. HR of 0.50 means the drug cut the risk of death or progression by 50%. This is outstanding. HR of 0.70 means a 30% reduction. This is good and likely approvable. HR of 0.85 means a 15% reduction. This is modest and may face FDA scrutiny. HR of 0.95 means only a 5% reduction. Even if statistically significant with a huge trial, this is clinically marginal. Median improvement is the other key metric. If the control group had median overall survival of 12 months and the drug group had 18 months, that's a 6-month improvement. Whether that's good depends on the disease: 6 months is huge in pancreatic cancer (where survival is short) but modest in breast cancer (where patients may live years). Absolute vs. relative risk reduction matters enormously. A press release might say the drug reduced the risk of death by 30% (relative risk reduction). But if the baseline risk of death was only 10%, the absolute risk reduction is only 3 percentage points (from 10% to 7%). Always look at both numbers. For non-oncology trials, look for the actual difference between drug and placebo groups. In a depression trial, did the drug improve scores by 3 points on a 60-point scale (modest) or 8 points (substantial)? Context from previous drugs in the same disease area helps you calibrate what's meaningful.

Key takeaway

A hazard ratio below 0.70 (30%+ risk reduction) is typically strong data in oncology. Always look at both the relative improvement AND the absolute improvement — and compare to what existing treatments achieve in the same disease.

Example

CRISPR Therapeutics' Casgevy showed that 29 out of 31 sickle cell disease patients achieved freedom from severe vaso-occlusive crises. That's a 93.5% response rate in a disease where patients previously had no curative option. The effect size was so dramatic that the data essentially spoke for itself — no complex statistical interpretation needed.

Safety data: the red flags that kill biotech stocks

Efficacy gets the headlines. Safety kills the deals. Many drugs that work beautifully get rejected or restricted because of safety concerns. As an investor, you need to scan safety data as carefully as efficacy data. Serious Adverse Events (SAEs) are side effects that result in hospitalization, disability, or death. Every clinical trial reports SAE rates for both the drug group and the placebo group. If the drug group has significantly more SAEs, that's a problem. Compare SAE rates: drug group at 15% vs. placebo at 12% is manageable. Drug group at 25% vs. placebo at 10% is a red flag. Deaths on treatment are the most serious safety signal. Any death that might be drug-related gets intense FDA scrutiny. Even one unexplained death in a small trial can derail an approval. Liver toxicity (elevated ALT/AST) is a classic drug killer. If a trial reports liver enzyme elevations, look at the severity. Grade 1-2 elevations are common and usually manageable. Grade 3-4 elevations (more than 5x normal) are serious and often require a black box warning or dose restrictions. Cardiac events including QT prolongation, heart attacks, or heart failure signals get heavy FDA attention. Drugs that cause cardiac issues face a very high bar for approval. Discontinuation rates tell you how tolerable the drug is in practice. If 20% of patients stopped taking the drug due to side effects, that limits real-world commercial potential even if the FDA approves it. The investor takeaway: a drug with moderate efficacy and clean safety often has a better regulatory path than a drug with strong efficacy but concerning safety. When reading trial results, look for the safety section (usually near the bottom of the press release) and specifically check SAE rates, deaths, liver signals, and discontinuation rates.

Key takeaway

Safety data can make or break an FDA approval. Always check SAE rates, treatment-related deaths, liver toxicity signals, cardiac events, and discontinuation rates. A clean safety profile can be as important for the stock as strong efficacy data.

Example

Lykos Therapeutics' MDMA therapy for PTSD showed promising efficacy in Phase 3, but the FDA rejected it in August 2024. Concerns about the trial design and safety monitoring contributed to the Complete Response Letter. Strong efficacy alone wasn't enough to overcome regulatory concerns about the safety framework.

Reading the press release like a biotech analyst

When a biotech company reports clinical trial results, the press release follows a predictable structure. Here's how to read it efficiently and spot the key signals: The headline tells you the company's spin, not necessarily the truth. 'Positive topline results' usually means the primary endpoint was met. 'Demonstrated clinically meaningful improvements' without mentioning statistical significance is often a red flag — it may mean the trial missed its primary endpoint. Skip to the primary endpoint result first. Find the sentence that states whether the primary endpoint was met, the p-value, and the effect size. This tells you 80% of what you need to know. Check the control arm. How did the placebo or standard-of-care group do? Sometimes a trial 'fails' not because the drug didn't work, but because the control group did unexpectedly well. This matters for interpreting the data. Look at subgroup analyses carefully. Companies love to highlight subgroups where the drug worked even if the overall trial was lukewarm. Subgroup analyses are exploratory, not definitive. The FDA usually wants to see the overall population results, not cherry-picked subgroups. Read the safety paragraph. It's usually the last substantive section. Look for the specific numbers: SAE rates, discontinuations, and any named serious side effects. Check the company's next steps. Are they filing an NDA immediately? That suggests confidence. Are they planning 'additional analyses' or 'discussions with FDA'? That suggests the data may not be clean enough for a straightforward filing. Finally, compare to analyst consensus expectations. The stock moves based on whether results beat, meet, or miss expectations — not on absolute numbers. A hazard ratio of 0.72 could send a stock up if the Street expected 0.80, or down if they expected 0.65.

Key takeaway

Read the primary endpoint result and p-value first. Be skeptical of press releases that emphasize secondary endpoints or subgroup analyses without clearly stating the primary endpoint was met. Check the safety section and the company's stated next steps.

Example

When a biotech reports 'the trial demonstrated a strong trend toward significance (p=0.06) on the primary endpoint with highly significant results on multiple secondary endpoints,' experienced investors know this means the trial failed its primary endpoint. The company is spinning the data to soften the blow.

Key terms

Primary Endpoint

The main outcome measured in a clinical trial to determine if the drug works. The trial is designed around this single measurement, and the FDA bases its approval decision primarily on whether this endpoint was met.

P-value

A statistical measure of the probability that an observed result occurred by chance. A p-value below 0.05 is considered statistically significant, meaning the result is unlikely to be a fluke.

Hazard Ratio

A measure used in survival analyses comparing the rate of events between two groups. HR less than 1.0 means the treatment group had fewer events. HR of 0.70 means a 30% reduction in the risk of an event.

Overall Survival (OS)

How long patients live from the start of the trial. The gold standard endpoint in oncology because it directly measures whether a drug helps patients live longer.

Progression-Free Survival (PFS)

How long until a patient's disease gets worse (progresses) or the patient dies. A common oncology endpoint that's faster to measure than overall survival.

Objective Response Rate (ORR)

The percentage of patients whose tumors shrank by at least 30% on imaging. Used to measure anti-tumor activity. Does not measure duration of response or survival.

Serious Adverse Event (SAE)

A side effect that results in death, hospitalization, disability, or other medically significant outcome. High SAE rates relative to placebo are a safety red flag.

Confidence Interval (CI)

A range of values within which the true treatment effect likely falls. A 95% CI of 0.55-0.85 for a hazard ratio means we're 95% confident the true HR is between those values. Narrower is better.

Intent-to-Treat (ITT)

An analysis that includes all patients who were enrolled in the trial, regardless of whether they completed treatment. The FDA prefers ITT analyses because they reflect real-world treatment conditions.

Subgroup Analysis

Breaking trial results down by patient characteristics (age, gender, disease severity). Useful for generating hypotheses but not definitive — the FDA generally requires the overall population result to be positive.

Next steps

Find the most recent Phase 3 press release from a biotech stock you follow and practice identifying the primary endpoint, p-value, and effect size

Use ClinicalInvestor's Trial Translator to get a plain-English analysis of any clinical trial by pasting its ClinicalTrials.gov ID

Compare the clinical trial data to analyst consensus expectations — check whether results beat or missed what the Street expected

Look at 2-3 historical examples of positive and negative trial readouts to calibrate your understanding of what 'good' and 'bad' data looks like

Bookmark ClinicalTrials.gov and practice looking up the trial design for companies you're researching — understanding the endpoints in advance makes reading results much easier

Keep learning

What Is a PDUFA Date? A Complete Guide → →How to Invest in Biotech Stocks → →Biotech Stock Catalyst Trading Strategies → →