What preventive care actually prevents

“Preventive care” gets used to mean two different things, and conflating them is why the annual physical has such a confusing reputation. One is prevention proper: keeping a disease from developing. The other is screening: finding a disease that has already started, early enough that treating it goes better.

Both are worth doing. They work differently, and the evidence behind them looks different too.

Prevention is mostly unglamorous

The interventions with the strongest evidence for preventing disease outright are the ones nobody wants to hear about again: not smoking, staying physically active, sleeping enough, keeping blood pressure and metabolic markers in range, and vaccination.

Blood pressure is the clearest example. Hypertension causes no symptoms for years while it damages arteries, kidneys, and the heart. Treating it is undramatic — often a cheap generic medication and some changes that are easier to describe than to sustain — and it prevents strokes and heart failure that would otherwise arrive two decades later. Few interventions in medicine provide more health benefit for so little cost — vaccination being the other obvious contender.

Vaccination works the same way. So does moving your body most days. The reason these get less attention than they deserve is that a prevented event is invisible: nobody experiences the stroke they did not have.

Screening is a different bet

Screening looks for disease in people who feel fine. That makes it a probabilistic exercise, and it is why recommendations are specific about who should be screened and when, rather than testing everyone for everything.

A screening test earns its place when four things are true:

• The people being screened are at sufficient risk to make looking worthwhile
• The disease has a detectable phase before symptoms appear
• Effective treatment exists, and treating in that phase produces better outcomes than treating later
• The test itself does not cause more harm than it prevents

Colorectal cancer screening meets that bar clearly, which is why the recommended starting age for people at average risk was lowered to 45 — by the American Cancer Society in 2018, and by the U.S. Preventive Services Task Force in 2021. Cervical and breast cancer screening meet it, within defined age ranges and intervals. Many other tests do not, which is why they are not recommended even though they exist and can be ordered.

Why more testing is not better testing

Every test has a false positive rate. Run enough tests on a healthy person and something will come back abnormal — not because anything is wrong, but because that is how reference ranges work. A “normal range” is typically defined so that 95% of healthy people fall inside it, which means roughly one in twenty healthy people falls outside it on any given test.

That abnormal result then has to be chased. Sometimes the chase is a repeat blood draw. Sometimes it is a CT scan, a biopsy, months of worry, and a finding that would never have caused harm if it had gone undiscovered.

That last category is called overdiagnosis, and it is worth understanding because it is counterintuitive. Some detected abnormalities would never have caused symptoms or shortened life. Finding and treating those conditions can expose people to procedures and anxiety without improving health — the diagnosis is technically correct and the treatment still leaves the person worse off. It is the main reason serious screening guidelines are narrower than what a comprehensive panel would offer you.

The United States Preventive Services Task Force publishes graded recommendations for exactly this reason. A grade of A or B means the evidence supports the benefit. A grade of D means the evidence suggests the harms outweigh it. A grade of I means the evidence is insufficient to say — a category worth knowing about, because a lot of marketed testing lives there.

What a useful annual visit actually does

Worth being straight about the evidence here, because it is more mixed than the convention suggests. Randomized trials and Cochrane reviews of the periodic health evaluation have not consistently shown a reduction in all-cause or cardiovascular mortality. The annual physical as a standalone ritual does not have the evidence base people assume it has.

What these visits do reliably improve is narrower and still worth having: the delivery of recommended preventive services, detection of high blood pressure, vaccination rates, cancer screening uptake, risk-factor management, and how satisfied people are with their care.

So the honest framing is that much of the value, when present, comes from preventive care, risk-factor assessment, and shared decision making rather than from the physical examination itself:

• Reviewing what has changed since last year, including things you have stopped noticing
• Checking blood pressure, which is the single highest-yield measurement in primary care
• Deciding which screening is appropriate for your age, history, and family history — including which to skip
• Reconciling medications, which quietly accumulate
• Updating vaccines
• Talking about sleep, alcohol, mood, and stress, none of which show up on a standard panel

The visit is also where a baseline gets established. A single lab value is a data point; the same value tracked over five years is a trend, and trends are what tell you something is drifting before it becomes a diagnosis.

The honest summary

Prevention begins outside the clinic, but primary care reinforces it through vaccination, risk-factor management, and evidence-based screening. Screening works within limits that are narrower than the marketing suggests. Much of the job of a primary care visit is sorting one from the other for your particular situation — including saying no to tests that would generate more anxiety than information.

The same framing applies to treatment. Whether a medication is worth starting depends on the same question as whether a test is worth running — as it does with the GLP-1 medications, where the decision turns less on what the drug can do than on whether it is likely to make a particular person healthier.