When I was in medical school lo these many years ago, my professors believed that whether or not you saw a physician regularly made little difference in your health. Certainly enormous gains had been made in extending life expectancy, but my teachers thought these were due to a few specific advances: sanitation, obstetrics, and antibiotics. Whether you saw a physician regularly or waited until you became ill didn’t obviously make any difference in outcomes.

But this belief was just a gut feeling — there was no scientific data about whether doctors could help people stay healthy. Then as now, heart disease was the most common cause of illness and death in America. Back then, there was no prevention. Treating high blood pressure was just becoming routine. Otherwise you waited until something happened, and most people with heart disease didn’t do well. Often we physicians could only stand by and watch helplessly.

What a difference thirty years of progress has made! Though heart disease is now one of the most treatable conditions, prevention works even better than treatment. Similar advances have been made in many other illnesses. Solid scientific data is now irrefutable: you’ll be a lot healthier if you see a physician regularly, even when you feel well.

For instance, people in their early 60s who don’t have health insurance often forego needed care until they’re 65 years old and can get Medicare insurance. The data show clearly that they are sicker and have poorer health outcomes than those who had health insurance all along. The newly insured group stayed sicker than those who had been covered, even five years later.

Here’s another good example. As most people know, of major industrialized countries, only in America is health insurance a private, for-profit, hit-and-miss proposition. One-sixth of us have no insurance at all, and many others are underinsured. The most disturbing example of the need to get regular care is that America has the highest preventable death rate of the major industrialized nations (See US ranks last among other industrialized nations on preventable deaths and also here). Our shortened lifespans and poor health outcomes are due to poor health practices (Americans are among the world’s most sedentary and obese), poor access to care, and failure to use care when it is available.

A lot of us believe firmly “if it ain’t broke, don’t fix it.” Why should I see a doctor if I feel well? But we do not feel that way about our kids or prenatal care. Parents who have the means to do so bring their children to the pediatrician regularly. And virtually every woman knows that failure to see an obstetrician during pregnancy greatly increases the risks of bad outcomes for both mother and child.

What about adults who feel well and aren’t pregnant? Why should they ever see a physician?

At about this point in discussing prevention with my patients, I ask them if they ever change the oil in their car. “Sure!!” How often? “Every 3,000 miles!”

Then I ask why don’t you just wait until the engine starts acting up and then see the mechanic? At this point, everybody gets it: because that would ruin the engine. A car needs regular maintenance that must be done while it is still acting normally, or it will break down at 10-20% of its usual life expectancy.

To be sure, cars don’t heal themselves the way people do. But the science is clear. On average, people who get regular care live several years longer than those who don’t, and they have more years of normal function, free from disability.

So how do you tell if you need a particular preventive medicine intervention? Should you get every health screening procedure that comes along? I’d suggest the following four-fold test:

1. You should only test for conditions that are common, or for which you are personally at increased risk. For example, heart disease, strokes, and certain kinds of cancer kill 60% of us. On the other hand, hemochromatosis is a rare disorder in which a buildup of iron in the body can cause heart and liver disease and diabetes. Unless you have a relative with this condition or have certain other health problems, there’s no evidence that screening for hemochromatosis is helpful. It’s just too uncommon.

2. Screen for conditions that are catastrophic if you wait until they cause symptoms. Again, heart disease is a great example. You find out you have coronary heart disease in one of three ways, each equally likely. A third of the time you drop dead. A third of the time you have a heart attack, and a third of the time you develop chest pain or similar symptoms. These are all bad for you. On the other hand, if you develop the most common form of skin cancer (not melanoma), there’s no harm in waiting until it’s clear there’s something growing on your skin.

3. The screening test should be simple, accurate, and safe. It should not often lead to red herrings: false positive results that then have to further evaluated, often at some risk. I discuss how you evaluate heart disease risk here. Such screening has been shown to save lives. On the other hand, many commonly used screening tests produce lousy results. One example of a crummy test is whole-body CT scanning, which floods you with radiation and usually turns up questionable abnormalities that probably don’t mean anything, but no one can be sure. Do you do a biopsy or repeat the CT scan in six months just to be certain? What about the psychological distress caused by a false positive test?

4. Ultimately, we’d like scientific proof that doing the screening test saves lives. Doing the necessary scientific studies takes a long time and a lot of money, and so this confirmation is available only for certain conditions. Moreover, the quality of the scientific evidence varies, so the level of proof may be soft. And demonstrating a screening test is helpful is often easier than proving that it’s useless.

Not all of the preventive medicine procedures I recommend have scientific evidence of benefit. Probably the most important of the unproven “tests” is having a regular history and physical examination.

Insurance coverage of preventive medicine is all over the map. Many private insurers (like Blue Cross, Aetna, and Blue Shield) say they cover preventive care, but in fact their coverage is spotty. Often you get the best reimbursement for the preventive exam itself, but not labs. Most insurers cover mammograms and colonoscopies. Medicare covers many preventive medicine examinations and procedures. The Affordable Care Act extended health screening to virtually everyone with insurance, but not every health insurance plan complies with this requirement.

With that background, the most definitive list of preventive medicine procedures is at the U.S. Preventive Services Task Force, a US government agency. You can what’s recommended and what’s not and why at their web site.