Module 4.2 Are the results clinically important?

Are the benefits clinically meaningful as well as statistically significant, and do the benefits outweigh any harms of the intervention or diagnostic test?

Explore the following aspects of clinical importance:

Results and benefits

What are the results, and do you understand how they are presented?

• The differences between relative (RRR) and absolute risk reductions (ARR) are often poorly understood by health professionals, and patients have even more trouble. However, you need to understand these concepts to communicate information to your patients.
  • RRR tells you how much smaller the risk is compared to before the intervention
  • ARR tells you how many people out of 100 are now safe because of the treatment
  • Read more about absolute and relative risk reduction.
• Become familiar with concepts such as Kaplan Meier survival curves  for trials of therapy, Forest plots for meta-analysis, and Receiver Operating Curves (ROC) for diagnostic tests. Read more about these concepts in the Resources section.
• The number needed to treat (NNT) is calculated as 1/ARR. It is the number of people that you would have to treat with treatment B in order to save one additional life. In the examples above, treatment B may give a NNT that varies from 10 to 40 depending on the expected event rate. NNT will differ in individual patients as their baseline risk differs.

Assess whether the treatment effect is clinically relevant.

Determine what level of uncertainty surrounds the results.

Consider how results are put into context in the discussion.

Risks and Toxicities

Ask yourself if the risks or toxicities outweigh the benefits of treatment.

• What is the Number Needed to Harm (NNH)?
  • This includes calculating the Absolute Risk Increase (ARI). This includes determining the risk of harm in the treatment group and the risk of harm in the control group. The difference between these two risks equals the ARI.
  • NNH=1/ARI

Clinical v Statistical significance

• Almost any difference can be statistically significant if a large enough clinical trial is done. However, a statistically significant clinical trial result does not always imply a clinically important result.
• A clinically important result must show a difference or benefit that is large enough to be worth the costs, toxicities and inconveniences involved.

For example, in 2024, the PEACE-3 trial (1) evaluated the addition of radium-223 to enzalutamide in patients with metastatic castration-resistant prostrate cancer. The randomised control trial showed a statistically significant improvement in radiographic progression-free survival (rPFS) with a 31% reduction in the risk compared to enzalutamide alone (HR=0.69, p=0.0009). However, the treatment produced an increase in median survival from 16.4 months with enzalutamide alone compared to 19.4 months with the combination therapy. While the improvements in rPFS was statistically significant the absolute increase in median rPFS was 3 months. The modest extension raises questions about the clinical significance of the findings, especially when considering treatment cost, potential side effects and impact on quality of life.

1 Gillessen S, Choudhury A, Saad F, et al. LBA1 A randomized multicenter open label phase III trial comparing enzalutamide vs a combination of Radium-223 (Ra223) and enzalutamide in asymptomatic or mildly symptomatic patients with bone metastatic castration-resistant prostate cancer (mCRPC): First results of EORTC-GUCG 1333/PEACE-3. Annals of Oncology. 2024;35:S1254.