Introduction
Medical screening is often framed as an uncomplicated good: if disease is found earlier, treatment can begin earlier and outcomes should improve.
This episode examines why that reasoning is incomplete. Screening can produce benefit, but it can also generate false positives, incidental findings, overdiagnosis, overtreatment and diagnostic cascades.
Dr. Austin Baraki discusses the distinction between screening people without relevant symptoms and investigating a clinical problem, the probabilistic interpretation of test results, biases that can make screening appear more effective than it is, and the importance of shared decision-making before a test is ordered.
Timestamps
- [03:35] What screening means
- [06:36] Choosing what to screen
- [11:59] Sensitivity and specificity
- [16:54] Predictive value and prevalence
- [22:16] Harms of screening and testing
- [31:02] Overdiagnosis and biases
- [41:40] The $50,000 physical
- [44:49] Overdiagnosis vs false positives
- [53:03] Overtreatment and surveillance
- [57:22] Clinician action bias
- [01:02:30] What to screen for
On this episode:
Episode Resources
- Join the Sigma newsletter for free
- Subscribe to Sigma Nutrition Premium
- Enroll in the next cohort of our Applied Nutrition Literacy course
- Barbell Medicine
- Instagram: @austin_barbellmedicine
- Twitter: @AustinBaraki
Further Reading:
- Prevention TaskForce – recommendations from the USPSTF Preventive Services Database
- ePrognosis (from UCSF)
- Early Detection of Cancer – Harding Center for Risk Literacy
- Maxim et al., 2014 – Screening tests: a review with examples
- Hoffmann et al., 2017 – Clinicians’ Expectations of the Benefits and Harms of Treatments, Screening, and Tests
- Welch & Albersen, 2020 – Reconsidering Prostate Cancer Mortality — The Future of PSA Screening
- Andermann et al., 2008 – Revisiting Wilson and Jungner in the genomic age: a review of screening criteria over the past 40 years
- Rothberg, 2014 – A Piece of My Mind. The $50,000 Physical
- Broderson et al., 2018 – Overdiagnosis: what it is and what it isn’t
- Armstrong & Eborall, 2012 – The sociology of medical screening: past, present and future
- Gillespie, 2011 – The experience of risk as ‘measured vulnerability’: health screening and lay uses of numerical risk
- Hofmann et al., 2019 – Expanding Disease and Undermining the Ethos of Medicine
- Overdiagnosed: Making People Sick in the Pursuit of Health
Guest Information
Austin Baraki, MD is a practicing Internal Medicine Physician, Clinical Assistant Professor of Medicine, educator, and coach. He completed his doctorate in medicine at Eastern Virginia Medical School, and Internal Medicine Residency at the University of Texas Health Science Center in San Antonio.
Study Notes
Useful Terminology
- Screening: Testing people without relevant symptoms to identify an unrecognised disease, precursor state or risk factor.
- Diagnostic evaluation: Investigation prompted by symptoms, signs or other clinical information, with the aim of explaining the presentation and guiding management.
- Pre-test probability: The estimated probability that a person has a condition before the result of a particular test is known.
- Post-test probability: The updated probability of the condition after the test result and other relevant information are considered.
- Sensitivity: The proportion of people with the condition who test positive. A highly sensitive test produces relatively few false negatives.
- Specificity: The proportion of people without the condition who test negative. A highly specific test produces relatively few false positives.
- Positive predictive value: Among people who test positive, the proportion who truly have the condition. It depends on both test performance and the prevalence of the condition in the tested population.
- Negative predictive value: Among people who test negative, the proportion who truly do not have the condition. This also varies with prevalence.
- False positive: A positive test result in a person who does not have the condition.
- False negative: A negative test result in a person who does have the condition.
- Overdiagnosis: Correctly identifying a condition that would never have caused symptoms or death during the person’s expected lifetime.
- Misdiagnosis: Assigning a diagnosis that the patient does not have.
- Overtreatment: Treatment that is unnecessary or provides no net benefit. It may follow overdiagnosis, although the two concepts are not identical.
- Lead-time bias: The appearance of longer survival because a condition is detected earlier, even when the time of death is unchanged.
- Length-time bias: The tendency of periodic screening to detect more slow-growing, long-duration disease than rapidly progressive disease.
- Disease mongering: Expanding or promoting disease definitions in ways that turn ordinary experiences, low-risk states or weakly consequential findings into medical problems.
- Commission bias: A preference for action over inaction, even when observation or deliberate non-intervention may be the better choice.
- Diagnostic cascade: A sequence of further tests, consultations or treatments triggered by an initial finding, including an incidental or false-positive result.
Screening and Diagnosis Answer Different Questions
Screening begins with an asymptomatic population. Diagnostic evaluation begins with a symptom, sign or other clinical concern that changes the probability of different explanations.
- Screening searches for unrecognised disease in people who appear well.
- It may be offered to an entire population, a demographic group or a more selective high-risk group.
- Its intended purpose is not simply to find an abnormality. It should reduce later morbidity, mortality or suffering by creating a useful opportunity to intervene.
- Because the starting prevalence is often low, even a technically good test can generate a substantial number of false positives.
- Diagnostic testing begins with a reason for suspicion.
- Symptoms, examination findings, medical history and prior results alter the pre-test probability of