In a physician chat group recently, a doctor who treats hospitalized patients made a recommendation to our group of 38,000 members that left me startled and alarmed. She shared her protocol for all COVID-19 patients admitted to the hospital: every one of them gets not only a chest x-ray but an entire battery of special tests, including a coagulation test, a leg ultrasound and a CT scan.

This was offered as her blanket standard of care. What it actually represents to me is one of the biggest problems with health care in America—because not every admitted patient needs all these tests. And this is not a new story.

Simply put, as physicians in the U.S., we overdiagnose and overtreat people. We order way too many tests, treatments and surgeries that you don’t need and that may actually harm you—and they cost money, lots of it. Incredibly, in a nation of 328 million people, we order approximately 15 million nuclear imaging studies, 100 million MRIs and CT scans and close to 10 billion blood tests on patients every year.

We’re not talking about a little pinch of “err on the side of caution” here. We are talking about runaway medicine, with patients aboard a system with no brakes, few guardrails and no one fully at the controls. Moreover, we as physicians don’t do a good job informing you of the downstream risk of these costly procedures, because we may not even recognize the mess we are creating in the first place.

Why all the tests? Well, we may order them because we are worried about missing a diagnosis and we want to get you healthy as soon as possible. Sometimes, it is to avoid getting sued. (When in doubt, it’s always easier to test.) Sometimes it’s because patients request the tests and we yield. Less innocently, we may order more tests because they bring in more money. We are doctors, but we are revenue generators too, and, for some, not necessarily in that order.

Let me walk you through two overdiagnosis/testing hypotheticals. Here’s one: You are admitted to the hospital with a cough and a low-grade fever. You test positive for COVID-19, and a chest x-ray demonstrates that you have a pneumonia. This is a fairly typical scenario. As a clinician, though, I also opt for a CT scan to make sure you do not have a blood clot in the lung (a pulmonary embolism), even though when you were admitted most doctors already would have put you on a prophylactic lower dose of a blood-thinner for this—to help prevent clotting.

The CT scan again shows the pneumonia; we knew that already from the x-ray. There is no finding of a pulmonary embolism. But the scan it does show a little something of which I’m not exactly sure—a possible nodule, or what we physicians refer to as an “incidentaloma.” This is not uncommon; incidental solitary lung nodules are found on CT studies from 8 to 51 percent of the time. Per an expert panel of the American College of Radiology, most incidentalomas found on imaging studies are benign. In fact, 99 percent of lung nodules found on chest CT are benign—that is, not cancerous. But their discovery can lead to extensive downstream testing, worry and potential morbidity, as we doctors start poking and prodding you, wanting to assure ourselves that the incidentaloma is indeed nothing.

I inform you of the nodule in your lung. You pepper me with questions and become worried, because, after all, sometimes nodules are cancerous. We may run more tests to assess it, but more likely we’ll monitor this nodule for months—years, sometimes—and do additional CT scans at various intervals to see whether it changes in size. These tests will expose you to additional radiation, which may increase your risk of cancer years down the road, and cost you money.

If we are more concerned about your incidentaloma, we may decide to do a lung biopsy, in which we stick a needle directly into the lung to get a tissue sample for evaluation. Not very often, but occasionally, we accidentally poke a hole in the lung, causing what is called a pneumothorax. This is quite serious and can make you immediately short of breath or sometimes even cause a drop in your blood pressure.

Now, not only have you received potentially dangerous radiation from the original CT scan, but you have undergone an invasive procedure and you’ve experienced harm. This particular complication will likely require the placement of a chest tube or a Thora-Vent device, both of which are invasive and can be very painful. If the pneumothorax is significant, it will mean a hospital stay in order to observe you. Key here, of course, is that we never needed to go down this path— but here we are.

Here’s another scenario, and one that is much more common. Instead of the incidentaloma, we find a small pulmonary embolism (PE). As your doctor, I breathe a sigh of relief and pat myself on the back for ordering a test that I didn’t really think you needed, but one that yielded this discovery.

But here’s the catch: From the time CT pulmonary angiography was first used about 30 years ago, we have dramatically increased the number of patients in whom we look for and diagnose PE. “There is good reason to believe that this hasn’t done anything good,” says Jerome Hoffman, professor emeritus at UCLA Medical Center and an expert on overdiagnosis.

In an interview, Hoffman told me that doctors used to look for PEs “only when patients were clinically very ill," and the PEs they found were indeed very important; they killed about one in five people with this diagnosis, he says. Now, “because it is so easy to look with CT, we find a lot more PEs, but the same number of patients die—meaning that your chance of dying from a PE today is drastically less.” Hoffman calls the lower case-fatality rate an optical illusion, in that we’re now finding so many more PEs, the vast majority of which would be clinically insignificant.

In our hypothetical, though, we don’t have evidence from randomized controlled trials to conclude that ignoring these small clots is safe, so I feel compelled to treat you. Treatment is controversial: some guidelines suggest it’s not required, some advise to consider it on a case-by-case basis, and others recommend treating most cases. I choose to prescribe you a blood thinner medication—but sometimes this has harmful side effects, like serious bleeding. This might lead to you needing blood transfusions, which, of course, carry their own risk. You’re now transferred to the intensive care unit. And so it goes.

Technology becomes the problem rather than part of the solution. And, says Hoffman, “This will only get worse as our wondrous technology ‘improves,’ and we find ever tinier abnormalities that we otherwise wouldn’t have known existed, and wouldn’t have caused any harm.”

The cost is staggering. At least 30,000 deaths in the U.S. each year are linked to mistakes and injuries caused by superfluous medical treatment. Meanwhile, unnecessary testing and surgeries add up to more than $200 billion in extra spending per year in the U.S., according to the Institute of Medicine.

In a 2017 survey, doctors from the American Medical Association (AMA) said that nearly a quarter of all the tests they performed were unnecessary, along with more than 20 percent of the prescriptions they wrote and every 10th surgery. Researchers estimate that among cancers detected by screening, some 11–19 percent of breast cancers and 20–50 percent of prostate cancers constitute overdiagnosis.

Too much medical care is a true public health crisis in this country. But such treatments continue to be pushed as part of what is already a $3.8 trillion domestic health care market—and hospitals, insurers, device makers and big pharma are scooping up the profits.

Let’s get back to your COVID-19 admission. What’s that going to wind up costing you? FAIR Health analyzed data from over 30 billion private health care claim records, using revenue codes associated with influenza and pneumonia as a means of estimating these costs. Their finding: The average charge for a hospitalized COVID-19 patient stay would be $73,300 if the patient were either uninsured or seeing an out-of-network provider.

That estimate is significantly lower for a privately insured patient with an in-network provider, but it still clocks in at a breathtaking $38,221. (How much you’d actually pay, of course, depends on your health plan deductible and the cost-sharing policies of your insurer.) Throw in some extra testing charges, and it’s little wonder that nearly one third of working Americans have some level of medical debt, with 28 percent of those owing at least $10,000.

As physicians, we all went into medicine to help patients, not bankrupt them. But medicine has become a business replete with overdiagnosis and overtreatment—and skyrocketing medical charges. It’s clear that our health care system needs significant reform.

A broader discussion of that change is beyond the scope of this article, but here are some ideas to kick-start the process. First, rein in hospital and pharmaceutical price gouging via market forces, price caps or regulation. Use value-based or bundled care, rather than fee-for-service medicine, to reduce the impetus to order more. Initiate tort reform, to help alleviate the practice of defensive medicine. Push for cost transparency so that prices are knowable to patients ahead of time, thus allowing for informed decisions. Eliminate some of the enormous administrative excesses and various middlemen who siphon off profits in our medical supply chain, spiking costs.

You should absolutely see a physician for your health concerns, especially if you are sick with COVID-19; it’s a dangerous virus. But don’t ever be afraid to speak up and ask why tests are being ordered. Ask why specialists are being consulted, why procedures or surgeries are being scheduled, or why you are being admitted. If it’s an elective situation, feel free to get a second or third opinion.

Let’s take our health back while keeping our wallets intact. As we’ve seen repeatedly, it is often true in medicine that less is more.