American medicine has a bias toward action. The default when a number is off is to treat it; the default when a scan finds something is to remove it. That bias has saved lives in many places. In several common conditions, it has also produced a steady flow of harm — interventions whose risks outweigh their benefits in the average patient.
This isn’t a conspiracy theory. It’s a quiet conclusion in JAMA, BMJ, and the New England Journal, repeated often enough that the term “overdiagnosis” now has its own annual conference.
Mild hypertension and the threshold creep
Blood pressure thresholds keep moving down. The 2017 ACC/AHA guidelines redefined hypertension at 130/80, instantly relabeling roughly 30 million Americans. For people with established cardiovascular disease, tight control matters. For low-risk adults with stage 1 hypertension and no other risk factors, randomized data show drug therapy produces minimal absolute benefit and meaningful rates of dizziness, falls, and kidney injury. Lifestyle changes — sodium reduction, weight loss, exercise — match or exceed the effect of a first-line drug in this group, without the side effects.
The conversation worth having with your doctor is about absolute risk over ten years, not the label on your chart.
Early-stage prostate cancer
PSA screening picks up cancers that would never have killed the patient. The PIVOT and ProtecT trials both showed that for low-risk localized prostate cancer, immediate surgery or radiation produced no survival advantage over active surveillance over more than a decade. The interventions did produce incontinence and erectile dysfunction at high rates.
Active surveillance — regular PSA tests, periodic MRIs, biopsy only on progression — is now the guideline-recommended first option for low-risk disease. It is still offered far less often than the evidence supports, partly because surgical pathways are faster, more lucrative, and emotionally satisfying for everyone except the patient living with the side effects.
Knee arthroscopy for osteoarthritis
Arthroscopic surgery for degenerative knee disease was a billion-dollar procedure for years. Then sham-surgery trials — yes, real anesthesia, real incisions, no actual repair — showed patients improved equally whether they got the procedure or not. Pain relief came from time, attention, and physical therapy, not the scope. Major guidelines now recommend against arthroscopy for osteoarthritis, but it is still performed at scale, often because patients ask for it and surgeons accommodate.
Other crowded examples include spinal fusion for nonspecific low back pain, stenting for stable angina (the ORBITA trial), and antibiotics for uncomplicated sinusitis. Each has its niche; each is used far past it.
The takeaway
Overtreatment is not a reason to distrust medicine; it’s a reason to ask better questions. What is my ten-year absolute risk if we do nothing? What’s the number needed to treat for benefit, and the number needed to harm? Is active surveillance an option? Doctors who answer those questions clearly are the ones worth keeping. The honest “we can watch this” is often the best medicine on offer.
Leave a Reply