When running clinical trials, pharmaceutical and biotech companies must focus on more than just the efficacy of the treatments being tested. Understanding and accurately communicating treatment effects on survival is one of the most critical aspects of evaluating new therapies. While survival analysis and hazard ratios are widely used, they are not always well understood, and there are more advanced methods available that can provide a more patient-oriented view of treatment benefits.
In this blog, we’ll break down 8 essential tips to help pharmaceutical and biotech companies understand and communicate treatment effects more effectively in clinical trials. These tips are grounded in statistical approaches on the measures of treatment effect discussed in the review paper “Understanding and Communicating Measures of Treatment Effect on Survival: Can We Do Better?”, by:
- Everardo D. Saad, Medical Director, IDDI
- John R. Zalcberg, at the time of publication, working at School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia
- Julien Péron, at the time of publication, working at the Department of Medical Oncology, Hospices Civils de Lyon, Pierre-Benite, France
- Elisabeth Coart, Director of Consultancy Services, IDDI
- Tomasz Burzykowski, Vice-President of Research, IDDI
- Marc Buyse, Founder and Chief Scientific Officer, IDDI
By the way, if survival analysis seems complicated, you may find it helpful to read our blog where we explain the fundamental aspects of survival analysis in simple terms. It might be useful to review that blog before continuing with the one below.
1. Go Beyond the Hazard Ratio: Use Complementary Measures
The hazard ratio (HR) is often the default measure for treatment effects in time-to-event trials, but it is not always the most comprehensive. While the HR should be interpreted as capturing the relative reduction in the hazard, its use as a measure to communicate treatment benefit in clinical practice is limited, as it doesn’t measure absolute benefit or hazard in terms of survival probabilities. Therefore, it cannot be directly related to the expected survival of an individual patient.
Tip: Combine the HR with absolute measures such as difference in survival probabilities at key time points or restricted mean survival time (RMST). This provides a clearer picture of how the treatment affects individual patients over time.
2. Use Net Benefit for a More Patient-Oriented Measure
The concept of net benefit provides a more personalized view of treatment effect. It calculates the probability that a random patient in one of the treatment groups will live longer (or longer at least by a clinically relevant amount of time) than a patient chosen at random in the other group minus the probability of the opposite situation. This is a more intuitive metric for patients than a hazard ratio.
Tip: Introduce the net benefit as part of your trial analysis to give patients and physicians a better understanding of the real-world implications of the treatment.
3. Consider the Difference Between Survival Probabilities
Rather than relying solely on the hazard ratio, the difference in survival probabilities at specific time points (such as one year or two years) offers a direct, absolute comparison between the groups. This can be easier to interpret for clinicians and patients.
Tip: Present your trial results with survival probability differences at critical time points. But remember that there is loss of information when presenting only survival probabilities at specific time points, as they are specific points in time, and you may not be able to cover the entire range of differences in survival probabilities.
4. Restricted Mean Survival Time: A Valuable Alternative
The restricted mean survival time (RMST) is gaining recognition as a useful alternative to the hazard ratio. RMST measures the average time survived by patients over a period of interest and doesn’t require the assumption of proportional hazards.
Once you have the RTSM for each group, you can calculate the difference by subtracting them. As a result, the difference of RMST measures the mean gain in life expectancy through time t associated with the superior treatment.
Tip: RTSM is valuable for patients and clinicians because it allows them to know how much longer a patient receiving the superior treatment would be expected to live (on average, of course) through a period of time, when compared to a patient treated with an inferior treatment.
5. Tailor Your Communication for Patients and Clinicians
One of the key challenges in clinical trials (and statistics…) is communicating the results effectively. Many clinicians and patients struggle to understand what a hazard ratio or relative hazard reduction means for them.
Tip: Use simpler, patient-oriented metrics to communicate the real-world implications of your treatment results. These metrics are more relatable and easier to grasp for non-statisticians.
6. Focus on Individual Patient Benefit
Clinical trials often focus on average treatment effects, but patients are more interested in how the treatment will benefit them specifically. Using measures like net benefit can help assess the probability of individual benefit in a way that’s meaningful for patients.
Tip: Incorporate individualized treatment measures like net benefit into your communication with patients and healthcare providers to better convey the potential personal impact of your treatment.
7. Use Median and Restricted Means Appropriately
While the median survival time is commonly reported, it doesn’t always provide the full picture, especially when survival curves don’t reach 50%. Additionally, median differences can be misleading if the sample is heavily censored with short follow-up.
Tip: Report both the median survival and restricted mean survival times (RMST) in your results to ensure you capture both short-term and longer-term benefits.
8. Incorporate Multiple Measures to Paint a Complete Picture
No single measure can adequately describe the full effect of a treatment in clinical trials. The paper emphasizes the need to use multiple statistical measures, including both relative and absolute metrics, to fully capture the treatment’s impact.
Tip: Don’t rely on one metric alone. Use a combination of hazard ratio, survival probabilities, and restricted mean survival times, for example, to offer a complete analysis of the treatment effect.
Conclusion: How IDDI Can Help
Effectively communicating the results of clinical trials is crucial, not just for regulatory purposes, but also for ensuring that physicians and patients can make informed decisions. IDDI has extensive experience in providing advanced statistical services to pharmaceutical and biotech companies, including the application of innovative survival analysis techniques.
By partnering with IDDI, you gain access to expert statisticians who can help you apply these cutting-edge methods, ensuring that your clinical trial data is interpreted correctly and communicated in a way that is both scientifically sound and easy to understand.
Contact IDDI today to see how we can support your clinical trials with our expertise in survival analysis reporting.