Hazard Ratio Explained for Biotech Investors

Q: What is a hazard ratio in a clinical trial?

A hazard ratio (HR) compares the risk of an event, such as death or disease progression, between two groups over time. An HR below 1 favors the experimental treatment and an HR above 1 favors the control arm. For example, an HR of 0.70 corresponds to an estimated 30% lower risk of the event in the treatment arm versus control over the analysis period.

Q: What does an HR of 0.70 mean?

An HR of 0.70 means the treatment arm had an estimated 30% lower risk of the event than the control arm over the period analyzed. An HR of 0.50 implies a 50% risk reduction, an HR of 1.00 means no difference between the arms, and an HR above 1.00 suggests the treatment arm may be worse. HR describes relative risk over time, not a difference in median survival.

Q: Is a hazard ratio the same as a difference in median survival?

No. Median survival is a single point in time where half the patients have had the event. A hazard ratio summarizes the relative event risk across the whole survival curve. A trial can show a small median difference but a strong HR if the curves separate consistently, or a larger median difference with a weaker HR if the curves cross or separation is inconsistent.

Q: Why should a hazard ratio be read together with the p-value?

Because HR and the p-value answer different questions. The HR estimates the size and direction of the treatment effect, the p-value indicates whether the result is statistically significant, and the confidence interval shows how uncertain the estimate is. An HR of 0.70 with p<0.05 may be statistically positive, while the same HR with p=0.12 may not be.

Hazard ratio, or HR, is one of the most important statistics in clinical trials. It compares the risk of an event between two groups over time.

In oncology trials, the event is often death or disease progression. An HR below 1 usually favors the experimental treatment. An HR above 1 usually favors the control arm.

Simple example

If a cancer drug reports an OS hazard ratio of 0.70, that means the risk of death was estimated to be 30% lower in the treatment arm than in the control arm over the analysis period.

If HR is 0.50, the estimated risk reduction is 50%. If HR is 1.00, there is no difference in event risk between the arms. If HR is 1.20, the treatment arm may be worse.

HR is not the same as median survival

Investors often confuse hazard ratio with median overall survival. Median overall survival tells you the time point when half the patients have died. Hazard ratio summarizes the relative event risk across the survival curve.

A trial can show a modest median survival difference but a strong hazard ratio if the curves separate consistently over time. Another trial can show a median difference but a weaker hazard ratio if the curves cross or separation is inconsistent.

What does HR 0.636 mean?

An HR of 0.636 means the treatment arm has an estimated 36.4% lower risk of the event compared with control. In oncology, HR targets around 0.60–0.70 can be meaningful, but interpretation depends on disease, setting, sample size and control-arm performance.

Why control-arm performance matters

The stronger the control arm, the harder it may be for an experimental drug to show a large relative benefit. If the control arm was expected to live 8 months but actually lives 18 months, the experimental drug must produce a larger absolute survival gain to achieve the same relative hazard reduction.

HR and p-value

A hazard ratio should be interpreted together with the p-value.

HR tells you the estimated size and direction of the treatment effect.
P-value tells you whether the result is statistically significant.
Confidence interval tells you how uncertain the estimate is.

A trial with HR 0.70 and p<0.05 may be statistically positive. A trial with HR 0.70 and p=0.12 may not be.

HR and clinical significance

Statistical significance is not the same as clinical significance. Investors should ask whether the HR is statistically significant, whether the absolute benefit is meaningful, whether safety is acceptable and whether the endpoint matters to patients and physicians.

Common investor mistake: focusing only on HR

A strong HR is important, but investors should also look at Kaplan-Meier curves. Ask whether curves separate, cross, show late benefit, have balanced censoring and include enough patients at risk late in the curve.

Frequently asked questions

What is a hazard ratio in a clinical trial?

A hazard ratio (HR) compares the risk of an event, such as death or disease progression, between two groups over time. An HR below 1 favors the experimental treatment and an HR above 1 favors the control arm. For example, an HR of 0.70 corresponds to an estimated 30% lower risk of the event in the treatment arm versus control over the analysis period.

What does an HR of 0.70 mean?

An HR of 0.70 means the treatment arm had an estimated 30% lower risk of the event than the control arm over the period analyzed. An HR of 0.50 implies a 50% risk reduction, an HR of 1.00 means no difference between the arms, and an HR above 1.00 suggests the treatment arm may be worse. HR describes relative risk over time, not a difference in median survival.

Is a hazard ratio the same as a difference in median survival?

No. Median survival is a single point in time where half the patients have had the event. A hazard ratio summarizes the relative event risk across the whole survival curve. A trial can show a small median difference but a strong HR if the curves separate consistently, or a larger median difference with a weaker HR if the curves cross or separation is inconsistent.

Why should a hazard ratio be read together with the p-value?

Because HR and the p-value answer different questions. The HR estimates the size and direction of the treatment effect, the p-value indicates whether the result is statistically significant, and the confidence interval shows how uncertain the estimate is. An HR of 0.70 with p<0.05 may be statistically positive, while the same HR with p=0.12 may not be.

Bottom line

Hazard ratio is one of the most important numbers in biotech trial analysis, but it should never be read alone. The best interpretation combines HR, p-value, confidence interval, mOS, survival curves, control-arm performance, safety and trial design.