Is this a good study? The statistics tell the story

Thursday, March 5, 2020

Ever read through a study and wondered how to apply the hazard ratio, or if you should change your practice because of a secondary endpoint finding? In this episode, Lauren M. Catalano, MD, of the University of Pennsylvania, Philadelphia, explains all the common terms and why they matter in the context of the KEYNOTE-024 trial.

In Clinical Correlation, Ilana Yurkiewicz, MD, of Stanford (Calif.) University, talks about how to prepare for an unexpected bad outcome.

Practice points:

  • Don’t skip over the statistical analysis portion of a paper.
  • Use Google to find simple definitions for unfamiliar biostatistics terms.
  • Understanding the statistical elements is essential to determining the quality of the research.

* * *

Understanding statistics in the context of KEYNOTE-024

Article discussed: Updated analysis of KEYNOTE-024: Pembrolizumab versus platinum-based chemotherapy for advanced non–small cell lung cancer with PD-L1 tumor proportion score of 50% or greater. J Clin Oncol. 2019 Mar 1;37(7):537-46.

Primary endpoint:

  • The outcome that is necessary to ensure the efficacy of the trial. What is the study’s objective?
  • The primary endpoint is defined prior to starting the study, which influences how many patients need to be enrolled to ensure statistical significance.
  • This paper’s primary endpoint: Time since random assignment to disease progression or death.

Secondary endpoint:

  • These are interesting trends or observations that the investigators were able to determine, but for which the original study may not have been powered, meaning that they may not have enough data to determine the statistical significance.
  • This paper’s secondary endpoints: Objective response rate (confirmed complete and partial responses) and safety.

Hazard ratio:

  • “Ratio” suggests that this is a comparison between the intervention and control arm.
  • HR is a measure of an effect or intervention on the outcome of interest over a period of time (risk per unit of time). The outcome can be positive or negative.

Confidence interval:

  • Since it is not possible to survey the entire population, a confidence interval provides a range of values where the true value most likely falls.
  • If the confidence interval crosses “1” then there is no difference between the arms of the study.

Kaplan-Meier curve:

  • Often used to illustrate survival.
  • This is a graphical representation of hazard ratio, usually drawn as a step function.

P value:

  • The degree of error that we are willing to accept.
  • Often P = .05, which means we are willing to accept a 5% risk that the hypothesis is incorrect.

Crossover:

  • Patients assigned in one arm of the study (usually the control arm) can be reassigned to the other arm (usually the intervention group).

Intention to treat:

  • A technique used in randomized, controlled trials in which patient outcomes are compared within the group the patient was originally assigned to. This may not reflect the treatment that the patient actually received.
  • If the patient is in a group that is treated but then leaves that group, they are still counted in the original group.

Show notes by Ronak Mistry, DO, resident in the department of internal medicine, University of Pennsylvania, Philadelphia.

* * *

For more MDedge Podcasts, go to mdedge.com/podcasts

Email the show: [email protected]

Interact with us on Twitter: @MDedgehemonc

David Henry on Twitter: @davidhenrymd

Ilana Yurkiewicz on Twitter: @ilanayurkiewicz

Podcast Participants

David Henry, MD
David Henry, MD, FACP, is a clinical professor of medicine at the University of Pennsylvania and vice chairman of the department of medicine at Pennsylvania Hospital in Philadelphia. He received his bachelor’s degree from Princeton University and his MD from the University of Pennsylvania, then completed his internship, residency, and fellowship at the Hospital of the University of Pennsylvania. After 2 years as an attending in the U.S. Air Force, he was drawn to practicing as a hem-onc because of the close patient contact and interaction, and his belief that, win or lose with each patient, one can always make a difference in their care and lives. Follow Dr. Henry on Twitter: @davidhenrymd.
Ilana Yurkiewicz, MD
Ilana Yurkiewicz, MD, is a fellow in hematology and oncology at Stanford University, where she also completed her internal medicine residency. Dr. Yurkiewicz holds an MD from Harvard Medical School and a BS from Yale University. She went into hematology and oncology because of the high-stakes decision-making, meaningful relationships with patients, and opportunity to help people through some of the toughest challenges of their lives. Dr. Yurkiewicz is also a medical journalist. She is a former AAAS Mass Media Fellow and Scientific American blog columnist, and her writing has appeared in numerous media outlets including Hematology News, where she writes the monthly column Hard Questions. Dr. Yurkiewicz is on Twitter: @ilanayurkiewicz.