The following originally appeared on The Upshot (copyright 2020, The New York Times Company)
Much of the care provided in the United States is unnecessary or too expensive. Cost sharing is one way to try to push patients to be more thoughtful consumers of medical care. The blunt way we use it, however, often does more harm than good.
Cost sharing is a blanket term for things like deductibles, co-payments and co-insurance. If patients are spending money “out-of-pocket” — their own money — they might think harder about whether care is worth it.
Recent news has highlighted one aspect of cost sharing in the United States that seems to make little sense: the notion of making it start over each year. Because of the way deductibles work, you pay them in the early months of the year, when the flu (and now coronavirus) are hitting.
Since few people have met their deductible yet, many will think twice about getting care now — right when the medical community most wants them to. Recognizing this, Gov. Andrew Cuomo recently announced that New York insurers should waive cost sharing for coronavirus testing. But his powers apply only to Medicaid, where cost sharing is already minimal. Those who have private insurance through their jobs or Medicare — which is most people and also those people (older Americans) who tend to be at highest risk — are under federal regulation.
Craig Garthwaite, director of the health care program at the Kellogg School of Management at Northwestern, says the biggest problem is that we don’t discuss often enough what we’re trying to accomplish: “There are two main reasons to consider cost sharing to be efficient. First, we might dissuade excess use of care, and second we might move people across products to more cost-effective options. Otherwise, we are just decreasing the benefits of insurance.”
There’s plenty of evidence on the amount of excess care. A study in PLOS One in 2017 asked doctors across the country how much care they thought people received that wasn’t necessary. They reported that more than 20 percent of it wasn’t. They thought that more than 10 percent of procedures, more than 20 percent of prescriptions and about a quarter of tests weren’t needed. A review of the literature in JAMA last year estimated that overtreatment or low-value care was responsible for $75 billion to $100 billion in waste a year.
Of course, none of those things happen without a doctor’s permission. Maybe we need patients to help in saying no.
But it turns out that patients aren’t very good at differentiating between necessary and unnecessary care. This isn’t entirely their fault. If there aren’t enough in-network specialists available, for instance, all the monetary incentives in the world can’t make patients choose more wisely.
Further, some care is clearly necessary. When expert panels deem something worthy of universal screening, for instance, it seems odd — not to mention counterproductive — to then put up barriers to obtaining it. Doctors don’t want patients to second-guess the decision to get these screenings. We want everyone to do it.
That was one of the main reasons that, as part of the Affordable Care Act, all screening with an A or B rating from the United States Preventive Services Task Force would be free without cost sharing. When it’s deemed necessary for everyone, all barriers are removed.
In many other facets of necessary care, though, the U.S. system isn’t nearly as thoughtful. When people have asthma, for instance, and they’re told they need a preventive steroid inhaler, that care is not “optional.” Doctors are not looking to have anyone question whether it’s worth “their money” versus an insurance company’s money. We want them to get and use the inhaler.
The same holds true for many chronic conditions. People with high blood pressure must take their medication — likewise, people with sickle cell disease, ulcerative colitis and, of course, cancer. When cost sharing is imposed for these diseases, the evidence suggests it leads fewer patients to take their drugs, which leads to worse outcomes.
In no disease is this clearer and more prevalent than diabetes. People with insulin-dependent diabetes, especially Type 1 diabetes, must take their medication, often many times a day, or they risk serious complications or even death in a very short period.
This is where the second reason to use cost-sharing comes into play. Perhaps there are cost-effective options for insulin. In that case, making sure one of them is available for low or no cost seems a good idea.
As things are set up now, when we look for care, there are usually almost no options available with low or no cost sharing. Everything costs money out of pocket, no matter what insulin you choose.
Alternatively, the system could recognize that some diseases or conditions need more care, not less. For these cases, cost sharing could be eliminated or significantly reduced.
In France, people with 32 specific chronic illnesses (making up 13 percent of the population) are exempt from co-payments for treatments for those diseases. In England, cost sharing is minimal, but even so it’s waived for people with cancer and other long-term conditions. In Germany, those with chronic conditions have their cost sharing capped at half the level of others.
In these countries, the percentage of people 55 and over who skip doses of drugs is much lower than in the United States. (People often skip doses because of the cost.)
One idea would be to change the types of cost sharing to be more strategic. Deductibles, for instance, make much less sense than co-payments or co-insurance. Deductibles just try to make people second-guess their spending, no matter what it’s on.
A better model might target people who choose high-cost options, increasing their cost sharing. Under a reference pricing system, for instance, insurers commit to pay the full cost for some lower-cost (but high-quality) care, and patients must pay the difference if they decide to go to providers that cost more.
“Reference pricing is smarter cost sharing because it puts the pressure not on the use of the service, but on the choice of a high-cost facility,” Mr. Garthwaite said. “In this way, it is less likely to decrease necessary care while still controlling costs.”
Cost sharing — which would be eliminated in Bernie Sanders’s “Medicare for all” proposal — isn’t inherently bad. But right now, it doesn’t appear to have a clear purpose, and the novel coronavirus reminds us how urgent a matter this can be. We seem to be getting all the harm without reaping much benefit.
