AD: Sutent and the other three drugs we looked at for treating end stage renal cancer were at the opposite end of the cost scale from the Alzheimer’s drugs. They extended life, but at a very high price. We would have to give up a lot of other health care to buy them. We initially said no, but we were uncomfortable about it because of the extra life that people could get. So we looked at our approach to valuing treatments which extend life at the end of life, and did a bit of recalibration. That enabled us to recommend one of the four drugs, Sutent.

TL: Don’t most new treatments offer something new?

AD: Yes, most things have something new to offer. The issue is not whether something is new or different, but whether that difference has any real value to patients, and whether it’s worth the other health care we could buy, but which we will have to forgo. In practice, we find sufficient value in most of the things we look at to recommend use in the NHS.

TL: What kinds of recommendations have you made?

AD: Of the 350 individual recommendations (in 170 technology appraisals), 84 percent recommend use of the intervention in whole or in part. It’s a pretty good rate. Obviously the companies selling the new products would like us to say “here’s a new product, let’s just use it the way its license suggests it should be used.” The intelligent approach, though, is to take a look at what’s being offered and see what value it adds and then optimize its use by targeting at those patients who can gain the most benefit.

TL: Can you be specific about the types of recommendations you make?

AD: Twenty-nine percent are unrestricted recommendations. That means the interventions can be used across the full range of the licensed indication. For 55 percent of the recommendations, we find that the new treatment benefits some patients—for example, those with advanced disease or that several drugs do the same job—so buying the cheapest makes sense. So we optimize the advice to reflect this.

TL: What about the other 16 percent of your recommendations?

AD: In six percent of the cases, the treatments have some promise but need more research. Our recommendation is that it should be used only as part of a clinical trial. Nine percent are simply not recommended because they show no benefit, and in one percent of the cases, the manufacturers did not make a full submission.

TL: What would happen if you spend more money on ineffective treatments?

AD: It would mean that money isn’t available to spend on other things that are more effective.

TL: Do patients understand that?

AD: It’s difficult for patients on one level to accept that. It’s terribly difficult for people if they think there’s something missing in their treatment. People can become angry and take their case to the media. It’s entirely understandable. People do want their taxes used as wisely as possible by the NHS, but they don’t like the idea of cost being taken into account in their own care. It’s a very human contradiction.

TL: How does the decision-making process work?

AD: We follow a standard process. We gather all the evidence in the public domain and ask the sponsoring company for the results of other unpublished clinical trials, although they are under no legal obligation to provide them. Then all the evidence goes before an independent advisory committee. Industry people are on the committee, and they are joined by economists and statisticians and people who use the NHS, as well as patient advocacy groups and NHS professionals. They look at the evidence and interpret it according to a framework which leads through a standardized decision process. Then they formulate a recommendation.

TL: What happens next?

Trudy Lieberman is a fellow at the Center for Advancing Health and a longtime contributing editor to the Columbia Journalism Review. She is the lead writer for The Second Opinion, CJR’s healthcare desk, which is part of our United States Project on the coverage of politics and policy. Follow her on Twitter @Trudy_Lieberman.