Last week, I wrote about the counterintuitive ways competition can work in the pharmaceutical industry: New drugs typically enter the market with higher price tags than existing medications, and then all the prices rise. In the case of Gleevec, a breakthrough cancer drug approved in 2001, the advent of pricier and more potent competitors seemed to  speed up the older drug's list price increases.

What remains frustratingly opaque, however, is precisely what the prices are. The drug industry commonly points out that list prices are essentially fiction; a drug's real price is a secret number that pharmaceutical companies develop in negotiations with the firms that provide prescription drug benefits. List prices are a proxy, a starting point against which rebates are applied -- but drug companies refuse to disclose the rebates and discounts they give to insurers, leaving the exact, "real" price basically mysterious.

Insurance companies get the benefit of these discounts and rebates; consumers pay a portion of the drug cost in the form of a copay.

Since discounts are private, a legitimate unknown in our story was how big those rebates were -- could Gleevec's hefty price hikes have been offset by increasingly large discounts?

To help answer that question, Richard Evans of SSR Health and his team offered to calculate the approximate value of the rebates over time for Gleevec and four other drugs made by Novartis. To estimate the average size of these discounts -- a calculation that includes factors other than rebates, such as fees and changes in wholesale or retail inventory -- SSR Health compared the gross sales of the products collected by health care data firm Symphony Health Solutions and the net sales Novartis reported each quarter. They also tracked the number of units sold and the wholesale list price of the drug.

First, Evans and his team found that the rebates given to Gleevec (the blue, dotted line below) appeared to be more modest than those given to other Novartis drugs. (Interestingly, the largest discounts in the sample they examined were to Tasigna, one of the drugs competing with Gleevec.)

Then, SSR Health looked at whether these discounts offset the average price increases. They did not. According to their analysis, the net price increases -- after the discounts were applied -- were sizable. Since the end of 2008, the average net price increase from year to year increase was 15 percent.

Then, Evans's team examined how the amount of drug sold varied over time, to see if the increase in revenue was driven by more pills being sold or by price increases. Their analysis found that the number of units sold remained relatively flat over time. Thus, it was Novartis' price increases after rebates that drove the growth in U.S. sales, helping turn Gleevec into Novartis' biggest drug by revenue last year.

"The evidence suggests that Gleevec’s relatively rapid list price growth has not been offset by growth in average discounts, meaning net price growth for the brand has been quite rapid," Evans wrote in an email. "Typically, U.S. brands’ list pricing gains are being offset by larger and larger rebates, meaning net price growth is typically much less than list growth – but as it turns out this appears not to be the case with Gleevec."

Novartis spokesman Eric Althoff said in an email that the company does not comment on rebates because they are proprietary.

Another report, by Credit Suisse, estimated the size of the discounts for a number of drugs in 2014 (hat tip to Joseph Walker of the Wall Street Journal). That report showed that, at least in 2014, Gleevec had one of the smallest discounts of any of the drugs in the sample examined -- just six percent. In contrast, the average discount was 29 percent.

Gleevec's price trajectory was undoubtedly affected by rebates and discounts, but the evidence strongly suggests it was price increases -- large ones -- that fueled the drug's growth over the past eight years, even as new competitors were introduced.

The debate over list prices and rebates also draws back the curtain on a huge problem: If drug companies are going to insist the list prices are a poor proxy, they should disclose the real ones.

Before the story on Gleevec was published, Holly Campbell, a senior director at the pharmaceutical industry trade group PhRMA, wrote in an email that she couldn't respond directly to questions about Gleevec's price, but said that generally speaking, competition and rebates work to control drug prices.

"It appears that you might be looking just at increases/decreases in a  list price, which does not take into consideration the significant – and typical – negotiations that occur," Campbell wrote. "Generally speaking, while the price of a medicine may increase or decrease over its lifetime, prices do fall dramatically as competition occurs among brand-name medicines -- and even further with the introduction of generics."

Brand name competition didn't seem to rein in Gleevec's price, and it's too soon to understand the full effect of the generic, which was introduced in the United States in February. Sun Pharmaceutical Industries, the generic maker that has exclusive rights to sell the generic in the United States for six months, said the price will be at least 30 percent off Gleevec's list price. But one patient tweeted a photo of her bill for name-brand Gleevec in December and compared it with her March bill for the generic version, imatinib. The reported price (not including rebates) that her insurance paid for the generic was just six percent less than the price it paid for Gleevec.

"The price of Imatinib Mesylate charged by us is considerably lesser than prices indicated in the tweet. We have neither visibility nor control over prices charged by providers," Frederick Castro, a spokesman for the generic maker Sun Pharma wrote in an email.

As with drug prices in general, the bill is likely an obfuscation -- hidden rebates mask the true price.