
December 24, 2021
AMA #30: How To Read And Understand Scientific Studies| The Drive with Peter Attia #188
Check out The Drive episode page
Key Takeaways
- It’s a valuable skill to be able to sift through the noise and find the signal in scientific studies
- Though meta-analyses summarize findings from previous studies, it’s still prudent to look at each individual study to make sure they’re sound – garbage in, garbage out (and vice versa)
- Observational studies are non-intervention studies looking at the outcomes of people and data to find patterns
- Experimental studies are intervention studies to determine safety and efficacy between groups of people assigned to intervention or control (non-intervention)
- The gold standard of experimental studies is a randomized controlled trial, ideally, double-blind so neither participant nor researcher knows who is assigned to intervention or control
- While clinical trials are regulated by FDA, most of the world follows the same guidelines for Phases I – IV
Introduction
In this special episode of The Drive, Peter Attia sits down with Head of Research at Attia Medical, PC Bob Kaplan (@bobkaplan) to take a deep dive into reading & understanding studies. Attia highlights how to find the important points, defines various types of studies, the evolution of studies from idea to execution, and much more. Note, this is only a preview. The full episode is only available to subscribers.
Host: Peter Attia (@PeterAttiaMD)
How Does A Study Go From Idea To Execution
- Good science is hypothesis-driven
- Studies are often guided by the null hypothesis versus the alternative hypothesis
- Null hypothesis: there is no relationship between two phenomena
- Alternative hypothesis: there is a relationship between the phenomena
- The gold standard to test the hypothesis is an experimental design, randomized controlled trial – ideally, be double-blinded
Different Types Of Studies
- Study design can be broken down into a few categories: (1) observational studies; (2) experimental studies; (3) meta-analysis or review
- Observational studies – no intervention, just looking at data and people to find patterns
- Individual case report: dive into one specific patient case that is driven by an observation that has not been studied before and can inform future treatment
- Case studies: similar to individual case reports but with a larger set of patients
- Retrospective cohort study: observational (not intervention) look back at a set of people who took part in ‘x’ and see how they are doing today versus people who did not take part in ‘x’
- The goal of cohort studies is to establish a pattern and potentially determine casualty
- Prospective cohort study: follow people who use ‘x’ over the next 5-10 years and compare them to people who do not use ‘x’ and observe outcomes
- Experimental studies – intervention
- Non-randomized trial (AKA open-label trial): two groups of people either given placebo or intervention but not randomized; grouping people based on certain characteristics
- Comes with a lot of bias because there is a reason you put certain people in a group
- Patients can also self-select into this type of study
- Randomized controlled trial: the gold standard of experimental design; study the question but take bias out by randomly assigning people to treatment or control group
- Non-randomized trial (AKA open-label trial): two groups of people either given placebo or intervention but not randomized; grouping people based on certain characteristics
- Meta-analyses: combine data from multiple studies that are attempting to look at the same question
- This seems great on the surface but is highly dependent on the quality of the original studies: garbage in, garbage out – and vice versa
- You can gain great insight from meta-analyses but take a closer look and examine each of the studies within the analyses
Phases Of Clinical Trials
- FDA clinical trials: drug or therapeutic you think will have some benefit in humans after showing success in animal models
- The first step is to file an investigational new drug (IND) application showing pre-clinical work and setting an intention to study in humans
- Phase I: small studies to determine toxicity (not whether the drug works) in a range of doses with efficacy
- Phase I studies look to see whether drug hurts, not how well it works
- Phase II: evaluate the safety and start exploring efficacy through open-label, usually non-randomized design
- Phase II studies are small and will compare to the natural history of patient outcomes to start examining the efficacy
- The goal is to randomize Phase II but patients will often self-select or doctors will recommend
- If drug or therapeutic has been deemed safe enough for use and has some efficacy, it will move on to Phase III
- Phase III: huge step up from Phase II studies; this is a randomized controlled trial and ideally blinded with the goal of measuring efficacy and determining broad use
- Phase IV: post-marketing study after the drug has been approved and uses information from a broader audience and exploring potential off-label uses or additional side effects
To learn more: subscribe and/or read Peter Attia’s series, Studying Studies.