From the New York Times (Cutting Dosage of Costly Drug Spurs a Debate)
When a drug can cost more than $300,000 a year, the right dose becomes a matter of public debate.
The drug in question, Cerezyme, is used to treat a rare inherited enzyme deficiency called Gaucher disease. Some experts say that for most patients, as little as one-fourth the standard top dose would work, saving the health care system more than $200,000 a year per Gaucher patient.
“It is economic malpractice to give a much higher dose of an expensive drug than is required,” said Dr. Ernest Beutler, an authority on Gaucher disease at the Scripps Research Institute.
Some other Gaucher specialists argue otherwise, saying that skimping on the medicine could endanger patients.
The economic stakes are high, but a big part of the problem is how drugs are priced in the first place. Infused and injected drugs (which are also usually the most costly) tend to be priced by volume or vial. Use twice the dose of Cerezyme and it will cost twice as much.
Interestingly this is not the pricing model generally used for pills. Higher strengths of drugs like Lipitor don’t tend to cost much more than the lower strengths, if they cost more at all. That’s the reason that pill splitting is used to cut costs.
I’d like to see someone try pricing drugs like software. After all, it costs very little to manufacture software or drugs. Most of the cost is in R&D. Therefore a licensing model could work well. Rather than charging a Gaucher patient twice as much if they use twice the dose, why not just charge by patient-month or patient-year? Same thing with less expensive pills like Lipitor for chronic conditions. Charge a certain amount per patient-year and let the patient and doctor adjust the dosage to the right level.
Shifting to a licensing model would require some changes. For example:
- Shifting from prescription-by-prescription co-pays to a monthly or annual license. This probably wouldn’t be so hard, and existing PBMs could handle it
- Putting in place safeguards to protect against product diversion. This wouldn’t be a big deal if everyone went to a licensed model, but of course that won’t happen. In a mixed system patients with drug licenses might sell or give surplus drugs to others. This would need to be addressed, although arguably such diversion doesn’t have to be cut to zero to make licensing worthwhile