H. Gilbert Welch, Lisa M. Schwartz, and Steven Woloshin · 2011

Overdiagnosed

Making People Sick in the Pursuit of Health

Rough AI truth score

83/100

The book's core warning is well supported: more sensitive tests and broader disease definitions can find abnormalities that would never cause symptoms, exposing people to labels and treatment without benefit. Magnitude varies sharply by condition and screening program, so the lesson is informed comparison of absolute benefits and harms, not rejection of screening.

Based on three central claims · high confidence

The three claims

01Supported

Screening can detect real abnormalities that would never progress to symptoms or death during a person's lifetime, creating overdiagnosis.

Randomized and incidence-trend evidence establishes overdiagnosis in multiple cancers and other conditions. Exact rates depend on disease, test, follow-up, age, competing mortality, and method, but the phenomenon follows directly from heterogeneous progression plus sensitive detection.

BMJ review of overdiagnosis in mammography programs BMJ review of methods for quantifying overdiagnosis

02Mostly supported

Expanding disease thresholds and using increasingly sensitive tests can turn risk factors or incidental findings into diagnoses without proportional health benefit.

Threshold changes, imaging, biomarkers, and surveillance increase detection and treatment, while benefit depends on baseline risk and effective intervention. Some earlier diagnoses meaningfully prevent morbidity; the problem is failure to distinguish beneficial detection from labeling with low expected value.

BMJ quantification of harms in cancer screening trials National Cancer Institute: cancer screening overview

03Mostly supported

Patients cannot judge screening value from detection or survival statistics alone and need absolute mortality benefits, false positives, overdiagnosis, and treatment harms.

Lead-time, length, and overdiagnosis biases can improve five-year survival without postponing death. Randomized mortality outcomes and absolute risk better address benefit, while quality of life, false reassurance, downstream procedures, and heterogeneous preferences remain essential.

BMJ quantification of harms in cancer screening trials UK National Screening Committee on surrogate outcomes

Other claims worth checking

Disease labels can produce psychological and financial harms independent of treatment.
Screening decisions should differ by baseline risk and patient preference.

What this number means. It is an AI-generated first-pass judgment of three central factual or causal claims—not a rating, exhaustive fact-check, or human peer review. Claim credits are 100% for supported, 75% for mostly supported, 50% for mixed, and 25% for weak, then averaged and rounded. Lower confidence means the score should move more as better evidence arrives.

Method three-central-claims/0.1.0 · checked 2026-09-01 · 3/3 selected claims assessed · method and source audit