Issue No. 14: Misplaced Incentives Raise Cost of Administering Drugsin Hospitals Sky-High

While most of us get our medicines at a pharmacy, about 30% of all drug spending pays for medicines that are administered by a physician, typically in a hospital or a private-practice or infusion clinic. Only a tiny proportion of all patients get their drugs this way – usually by injection or through an intravenous drip. Because a medical professional is required to administer the drug, reimbursement comes through the medical benefit part of an insurance plan, not the drug benefit.

These drugs – usually biologics, or complex, protein-based medicines -- often treat complex conditions like cancer, rheumatoid arthritis or multiple sclerosis. Costs are concentrated in a small number of high-priced medicines. Just 25 drugs are responsible for driving the majority of costs for both commercial and Medicare insurance plans, according to the Magellan Rx Management Medical Pharmacy Trend Report. 

Costs are even higher when such medicines are administered in hospitals. Everyone has heard stories of hospitals charging $15 for a Tylenol pill, but the prescription-drug charges are a more serious concern. The main source for high costs, incredibly enough, is the insurance reimbursement system. Hospitals are reimbursed far more generously than doctors who administer drugs in their offices. That fact is not only driving decisions on where to administer drugs, it is also encouraging mergers of physician practices into hospitals.

Avastin: $14,100 in Hospital, $6,620 in Doctor’s Office

Immunospuppressant drugs are far from unique. An IMS Health study found enormous differences for administering cancer drugs. The oncology drug Avastin, for example, cost $14,100 if administered in a hospital and $6,620 in a physician’s office; Herceptin, for breast cancer, was $5,350 in a hospital and $2,740 in a doctor’s office. 

A paper published last June in The Oncologist journal, stated that United Healthcare, a large insurer, paid physicians a premium of 28% above the average sales price to administer cancer drugs in their offices but a premium of 152% above the average sales price for precisely the same drugs administered in a hospital. 

This discrepancy is one reason that drug delivery is migrating to hospitals. That’s where the money is. In 2004, more than 90% of Medical Pharmacy Benefit (Part B) drugs for commercially insured patients were administered in doctors’ offices; by 2014, the figure had dropped to around 40%. In just four years, from 2011 to 2015, the cost of such drugs, per member per month for commercial insurance plans, rose 72% in hospital settings and 37% in doctors’ offices. For Medicare, costs rose 18% in hospitals and fell 2% in offices.

Why the Large Discrepancy? 

The big question is why. Insurers have clout, so why would they allow hospitals to take advantage of them this way?

The answer is that insurers need hospitals to participate in their networks in order to get patients to enroll in their plans and employers to choose those plans. Lee Newcomer of United Healthcare says that the hospitals tell insurers, in effect: “If you want our beds, you have to take our prices for oncology treatment.”

As expensive as medical pharmaceutical treatment may be, it pales in comparison to overall hospital costs. In effect, drug reimbursement becomes a loss leader, enticing hospitals to sign up with a health insurer.

Concerning Part of the Story: 340B Discounts

Another concerning part of the story involves the 340B discount program, intended to give a break in drug costs to hospitals that serve a large share of high-need patients. Hospitals, however, do not have to pass the discount along to the patients but may use the savings to offset their other costs. As Adam Fein writes on the Drug Channels blog: “The 340B program is highly controversial, partly because the 340B legislation does not specify or restrict how covered entities utilize funds generated by the program.”

An unintended consequence has been that hospitals that qualify for 340B discounts are buying up physician practices. These practices can then dispense drugs that the hospital purchases at the discount price, and the hospital pockets the savings – even if the practices are in affluent neighborhoods.

The 340B program has thus grown enormously. In 2016, it accounted for more than 50% of all hospital drug purchases, up from 30% in 2013. 

And no wonder. One study found that for a $1 drug, including delivery to the patient, a 340B hospital gets more than $2.50 while a non-340B hospital gets about $2 and a doctor’s office gets around $1.25. It’s no wonder, then, that independent physician practices are vanishing. In 2010, some 56% of oncology practices were independent, but by 2015, after the advent of 340B discounts, that figure had dropped to 43%.

An article in the  Charlotte Observer described the situation:

Large nonprofit hospitals in North Carolina are dramatically inflating prices on chemotherapy drugs at a time when they are cornering more of the market on cancer care, an investigation by the Observer and The News & Observer of Raleigh has found.

The newspapers found hospitals are routinely marking up prices on cancer drugs by two to 10 times over cost. Some markups are far higher.
It’s happening as hospitals increasingly buy the practices of independent oncologists, then charge more – sometimes much more – for the same chemotherapy in the same office.

The article quoted the president of Carolinas HealthCare System, a $7 billion chain, as saying, “The drug itself may just be the vehicle for charging for the services that are provided [elsewhere].” A cancer institute owned by the hospital chain collected $4,5000 for a dose of Irinotecan, a treatment for colon cancer, the article reported, even though “the average sales price of that amount of the drug [was] less than $60.”

​Low Incentives to Use Biosimilars

Biosimilars offer protential to save billions of dollars in healthcare costs. These are drugs with active properties similar to those of biologics, which in turn are complex large-molecule medicines. However, biosimilars are similar but not exact copies of biologics, hence the reference Biosimilar. But biosimilars aren’t used as much as you might expect. Many insurance plans are simply unfamiliar with the safety or efficacy data of biosimilars. In addition, some plans have contractual arrangements with the maker of the original biologic and find it in their business interest to restrict access to the biosimilar.

A RAND study estimates that biosimilars will reduce spending on biologic drugs by $44 billion by 2024. But that will only happen if there are biosimilars to use. Faced with short-sighted reimbursement policies, drug manufacturers will find it risky to invest in biosimilars to bring them to market. Unlike small-molecule generics, biosimilars are not cheap to produce.

Another policy that artificially inflates drug costs is the infusion of medicines in hospitals compared to physician offices. As hospitals acquire more and more physician practices, they bill the infusion services through the hospital system which at time is up to 300% more expensive than providing that same medicine in a physician’s office. These types of practices artificially inflate drug spending and increasing premiums while doing nothing at all in improving health.

Selective Transparency

A new study by the  Institute for Policy Innovation tackles the issue “selective transparency” across the spectrum of health care. Merrill Matthews and Peter Pitts explain:

While there is some pricing data available for various health care sectors, most of it is irrelevant or hard to access. For example, consumers may be able to find the list price for a prescription drug, but that price doesn’t tell them the price the insurer or PBM actually paid for it after discounts. And that discounted price may not include rebates that go back to the insurer or PBM—and that very seldom go to the patient.

Unlike pharmaceuticals, there is very little data available on health outcomes for hospitals and physicians—and what is available is difficult to access. And even that data may be misleading because of what is measured. For example, measuring hospital outcomes by the number of readmissions doesn’t actually say much about the quality of care. And measuring physician utilization rates also doesn’t tell us much about outcomes.

As much as prescription drugs get faulted for lack of head-to-head effectiveness comparisons, the reality is that there is much more data comparing drugs to one another than there is comparing Hospital A to Hospital B or Physician A to Physician B. Doctors can often tell a patient the likelihood a specific drug will help based on scientific evidence; that’s almost impossible to do with hospitals.