Yesterday's White House meeting on health care reform re-floated the idea of taxing employer-provided health benefits to help pay for insuring the uninsured. Sen. Max Baucus (D-MT), chairman of the powerhouse Senate Finance Committee, told reporters after the meeting that the president "might consider" taxing some employer-provided benefits, even though President Obama "'made it very clear' that he preferred his own revenue proposals," according to the  New York Times.

The idea of taxing health benefits has drawn strong support from many progressives. Eyeing the potential to raise $680 billion in revenue over five years (the health benefits tax exclusion now dwarfs the home mortgage deduction, whose repeal would only raise $444 billion over the same period), the liberal Center for Budget and Policy Priorities issued a new report yesterday calling for limiting the health care exclusion because "universal coverage may be out of reach otherwise." Jon Cohn, writing in his debut Kaiser Health News, column, endorsed the idea earlier this week with a slap at unionized workers "whose employers give them blue-chip coverage."

But are there really a lot of Cadillac plans out there? That gratuitous slap ignores the reality of today's insurance marketplace. Employer-based plans pool risk for members of that plan only. Whom do you think has the high-cost "blue-chip" health insurance coverage -- a newly opened, foreign-owned auto assembly whose average employee is 35 years old and has been hired because of his perfect health, or a General Motors plant whose average worker is 55 and has suffered through the stress of multiple layoffs and multiple plant closing threats over the past two decades? The only thing "Cadillac" in the health insurance costs of that GM worker is the nameplate of the car rolling off the assembly line. His higher premiums are a direct function of he and his co-workers' higher claims, not more generous benefits.

In pushing for removing the tax deduction, the CBPP report at least pointed to the necessary adjustments that would have to occur to make the new tax truly progressive. High-cost groups would have to be protected by not allowing insurance companies to set higher prices based on either an employer's size (thus protecting small business, which usually has higher rates because of higher administrative costs) or the health status of a firm's employees. This is called community rating, which can only be enforced by a strong regulator.

The exclusion's removal would also have to take geographical variation into account. Making the insured pay higher taxes because they live in areas with high health care costs punishes the victim, not the beneficiaries of those higher health care expenses, which are hospitals, physicians and medical suppliers who collect the fees for the often useless procedures offered in high-cost areas.

The idea of removing the income tax deduction as a way of raising revenue for insuring the uninsured has just one compelling argument behind it. It's a form of progressive taxation. Because tax rates are higher on higher income, the tax exclusion for health benefits is much more valuable to high-income employees than low-income employees. Removing the exclusion only for those with high incomes would amount to a progressive redistribution of income from the upper class to the working class, good economics because of the unequal distribution of income in our society, but very bad politics. Taxing the rich to pay for a new entitlement -- universal health care -- may appeal to liberals and the left, but can be easily attacked by opponents of reform. 

Indeed, couple its redistribution effects with the likelihood that Congress will be reluctant to impose tight regulation of the insurance industry and cost-control measures to offset geographic variation, repealing the health benefits tax exclusion could engender an angry backlash from already insured workers. I can already see the next round of Harry and Louise commercials, funded by opponents of reform. The 85 percent of working Americans who are privately insured will be told ad nauseum that the only benefit they're going to get from health care reform is a higher tax bill on top of their already skyrocketing co-pays and deductibles.

Let's send up two cheers this morning for the front page New York Times story by Gina Kolata about slow progress in the fight against the more than 200 forms of cancer. She starts the story by expressing skepticism about President Obama's decision to ramp up the "war on cancer." The story goes on to show that many of the "breakthroughs" announced in recent years really haven't changed the prognosis much for people with metastatic disease.

“Avoid consensus views.” “Challenge conventional science.” The Bill and Melinda Gates Foundation wants to inspire Silicon Valley-style innovation among scientists working on neglected diseases of the developing world.

And based on the grand challenge research projects unveiled yesterday to mark World Malaria Day, they’re doing a pretty good job. How about an in-home electro-magnetic wave device that repels mosquitoes? That’s what Szabolcs Marka, an astrophysicist at Columbia Universisty, used his $100,000 grant to develop, at least through the conceptual stage. “Frequency, intensity, there’s still much to be worked out,” he said in his thick Hungarian accent.

Not to mention what would be the power source in impoverished regions where malaria is endemic, or why there is reason to believe it would be more effective than bednets.

There's a big difference between the two. A warranty guarantees results if the product doesn't work. A discount simply lowers the price.

This morning's story in the New York Times suggesting some drug companies are starting to offer "pay for performance"-style guarantees to drug purchasers caught my eye because it confuses the two. Take one of its examples, for instance, which involved Actonel (risedronate), an osteoporosis prevention drug jointly marketed by Procter & Gamble and Sanofi-Aventis.

To fend off generic competition, the companies are offering to pay for any fractures suffered by women who take the drug. This could prove costly for the companies, right? It costs $30,000 to treat a hip fracture and $6,000 for a wrist fracture, according to the story.

But now let's look at how many older women would still suffer fractures even after taking the drug for a year. According to this analysis out of Great Britain, about 0.4 percent of older women who took risedronate still suffered hip fractures (compared to 0.6 percent who took another drug). That's about one hip fracture for every 250 women who took the $100-a-month drug.

About 2 percent of the women on the drug suffered far less costly non-vertebral fractures, some of which need treatment but many of which do not because they only show up on x-rays. But let's assume all 2 percent suffered wrist breaks.

By doing the math on the fractures, it's easy to see from the company's revenue perspective how the income from the drug in this program clearly outweighs the cost of paying for the treatments. It costs 250 patients $1,200 per year or $300,000 in revenue to the company to prevent one hip fracture, which costs $30,000 to treat, and five wrist fractures, which cost another $30,000 to treat).

So the bottom line is the company spends $60,000 in treatments for every $300,000 in revenue it takes in. Such a deal.

Now, from a health system perspective it could be argued that preventing the pain and suffering of the one woman in 500 on risedronate who didn't suffer a hip fracture or the one woman in 330 who didn't suffer a "non-vertebral" fracture is worth the cost (these are the numbers needed to treat to prevent one of each event). However, several cost-effectiveness analyses conducted in Europe arrived at the opposite conclusion.

This German analysis, for instance, pegged the cost of gaining a single quality-adjusted life year from risedronate treatment at 50,000 Euros (about $65,000 at today's exchange rate). That is higher than the generally accepted benchmark of $50,000 per QALY for what constitutes "cost effective" care.

So while it is nice to see that P&G and Sanofi-Aventis have offered to pay for fracture treatments, let's be clear about the nature of the program. It is not an "outcome guarantee," which was how it was described in the Times story. Rather, it's a minor price discount that brings the price of the drug only slightly more in line with its true value.

(I've updated the math from an earlier posting to include all events and ascribed the highest cost possible to those events.)

The Wall Street Journal's Health Care blog this afternoon carried an analysis of drug firms' pay-for-performance programs that was similar to my own analysis this morning, both of which played off the same New York Times story. So why no hat tip to GoozNews? Maybe it was just a case of great minds thinking alike and they don't read us over there. Sigh.

Posted at 4:26 p.m.

What does it say about global priorities if it takes nearly a century to develop a better vaccine for combating a disease that kills nearly two million people every year?

For the first time in 80 years, a vaccine aimed at protecting young children from pulmonary tuberculosis will advance to a proof-of-concept second stage clinical trial. The booster shot, a small dose of a recombinant antigen 85A found on the surface of Mycobacterium tuberculosis, stimulates the production of antibodies that supplement the classic BCG (Bacille Calmette-Guerin) vaccine, which, is given to an estimated 100 million children every year.

Alas, while the BCG vaccine is somewhat effective against meningitis-related brain TB and blood-borne miliary or "galluping consumption," which combined still kill an estimated 40,000 to 70,000 children worldwide every year, it is largely useless against pulmonary TB, which accounts for most TB cases and the nearly two million deaths from the disease. Researchers, largely operating on funds provided by charities like the Wellcome Trust, have been in a desperate race to develop booster shots that would improve the efficacy of the classic BCG shot.

"There is still a long road ahead, but this marks an important milestone toward the goal of a more effective TB vaccine," said Jerald Sadoff, president of the Aeras Global TB Vaccine Foundation, which has six TB vaccine candidates under development. Others organizations involved in producing the leading candidate include Oxford University, where the vaccine was initially invented by Dr. Helen McShane and her co-investigators; Emergent BioSolutions of Rockville, Md., which will manufacture the vaccine; Isis Innovation Ltd., the Wellcome Trust and the University of Cape Town in South Africa, where the second-stage clinical trial will take place.

The vaccine has so far passed safety tests involving 245 healthy adults and 250 healthy children, which took place in The Gambia and South Africa. The second stage trial will test the vaccine in 2,784 children in a province about 100 kilometers from Cape Town. The BCG vaccine will be given at birth and the booster shot will be given at 18 weeks.

Officials at the Aeras Foundation estimate various philanthropies have invested about $16 million to bring the recombinant booster shot to this point. Aeras, which is largely funded by the Bill and Melinda Gates Foundation, will invest another $15 million for the second stage trial. If it is successful, which won't be known until 2012 at the earliest, the vaccine's developers estimate it will take another $120 million to run the four-year, 40,000 infant third stage clinical trial.

If all goes according to plan, the world could have a new vaccine for combating tuberculosis by 2016. The BCG vaccine, made from live attenuated virus, was first used in humans at the Pasteur Institute in France in 1921, but did not come into widespread use until after World War II.

Food and Drug Administration deputy commissioner Joshua Sharfstein last week told agency scientists he's “already burning the candle at both ends to keep up with this job” as he asked for a one-page memo outlining complaints about the device devision. There will be an all-hands-on-deck meeting today to discuss the troubled Center for Devices and Radiologic Health, the New York Times reports.

There was one point on which nearly every special interest at Monday's public meeting on posting clinical trial results in the government's data bank, ClinicalTrials.gov, agreed: for the time being, it's best to leave patients in the dark.

At issue is whether clinical trial sponsors submitting data to the results database, which will be up and running 18 months from now, ought to include narrative summaries aimed at the lay public, most of whom do not have the skills to interpret the reams of data that will be housed in the publicly available technical section of the database. The 2007 Food and Drug Administration Amendments Act called for including interpretive materials for patients and consumers, but there are no agreed upon standards for how sponsors can write those narratives without appearing promotional.

Industry officials appeared more than happy to leave out those summaries. The Pharmaceutical Research and Manufacturers Association, even as it promised to register results from clinical trials for products that never reach the marketplace, something not required by the new law, warned that its non-technical narratives for approved drugs might be misinterpreted. "Pilot test a limited number of non-technical summaries," Jeffrey Francer, assistant general counsel of PhRMA, suggested. "PhRMA offers to help facilitate such a test with its member companies."

"Use focus groups to determine what will be useful to patients, and then employ an objective third party to review information prior to posting to ensure it's not misleading or promotional," advised Katherine McCarthy, director of science and regulatory affairs at the Biotechnology Industry Organization.

Note that both groups assumed that their member companies would write the narrative summaries.

The most definitive objection came from the New England Journal of Medicine. "The results should not include narrative summaries," declared senior deputy editor Edward Campion. "Even conscientious, dedicated professionals sometimes slip into a subjective, even sloppy narrative about their own work." In his experience, academic researchers and industry scientists were equally prone to hyping their own trial results when they submit manuscripts for peer review and editing.

Journal critics have been making that point for years, although their argument is that slanted conclusions often get published even after the peer review and editing process.

So, if the medical journal model can't save the patient-oriented narrative, what other model might there be? Bill Vaughan of Consumers Union, argued that non-technical summaries written by trial sponsors should be abandoned. "The current system is so distorted that narrative summaries cannot be trusted and are often little more than an advertisement," the consumer advocate wrote.

His group's solution? "The FDA and Congress should consider raising the fees on drug applications so as to cover the cost of hiring truly independent scientist-analysts to provide an honest narrative statement based on audited data," he said.

Good idea.

People volunteer to take part in clinical trials for a variety of reasons. Some want access to experimental therapies. Others believe they’re advancing the cause of science. A few do it for the money.

Whatever their reasons, most participants would agree that the results of those trials ought to be made public. How else will other researchers learn from their experience, not to mention other patients who suffer from the same malady or the physicians who treat them.

Yet only passage of the Food and Drug Administration Amendments Act in 2007 required clinical trial sponsors to publicly post trial results. However, there are many loopholes in the new law and many vagaries in what and how to report. It will take careful regulations to insure that the clinical trial results database now under construction at the National Institutes of Health is useful to independent researchers, patients and physicians.

Those regulations will be the subject of an all-day hearing at NIH’s suburban Washington campus on Monday. The agency hopes to publish final rules later this year so it can meet the law’s requirement that the results database be up and running by September 2010.

Officials at the National Library of Medicine, which runs NIH’s ClinicalTrials.gov, are already wrestling with quality of the data submitted by trial sponsors under temporary rules. About 350 new trials are being registered each week and about 500 trial sponsors have already submitted results.

An official I spoke with earlier this month said the quality of the results reports is all over the map. Moreover, the agency doesn’t have enough adequately trained staff to properly vet the submissions. “We’re not doing peer review, but we don’t want garbage up there,” the official said. “We couldn’t even envision the errors we’re seeing. It takes a lot of curation.”

The National Library of Medicine (NLM) clearly needs to develop uniform reporting requirements in a form that will enable independent researchers to pool multiple trials for analysis. Serious safety concerns from widely used drugs often only emerge from such meta-analyses.

The NLM also needs to establish rules for how the data and its interpretation will be presented to the public. The last thing anyone needs is another forum for ghostwritten narratives about industry-sponsored trials that hype marginal benefits or gloss over troubling adverse events.

The agency should also consider asking sponsors to provide the public with easily understood benchmarks. For instance, the reports could include how many people need to be treated before one person benefits (the so-called number-needed-to-treat analysis). Risk reduction or benefits should be reported in absolute percentages, not relative percentages. And sponsors should be asked to create composite risk measures, instead of reporting each side effect individually and dismissing them all because each one on its own is statistically insignificant.

But the biggest roadblock to making the results database most useful lies is what won’t get reported. The law excluded early stage safety trials (so-called Phase I trials). That exclusion shouldn't be turned into a loophole for Phase I/II trials, which are designed to test not just safety but look for early signs of efficacy. Efficacy data from a Phase I/II trial could ultimately wind up being significant when pooled with larger Phase III trials, especially when the total pool isn’t large as in many rare diseases.

Of course, the biggest loophole of all is the “confidential commercial information” exclusion. This covers the thousands of trial results that already exist inside the Food and Drug Administration and will never see the light of day. The reform law said the FDA must publish results from the trials for approved drugs, biologics and devices, but only after having received at least three freedom of information act requests.

And it is prohibited from publishing any trial results for drugs, biologics and devices that fail to pass the agency’s safety and efficacy requirements, which is about half of all applications. “The argument that such data must be withheld so that potential competitors do not benefit runs counter to the principles of ethical human research, which require that risks to human volunteers be minimized and that human participation in research leads to generalizable knowledge,” Alastair Wood, managing director of Symphony Capital and a Professor of Medicine at Cornell University’s Weill Medical College in New York, wrote recently in the New England Journal of Medicine.

The myopic industry view that disclosing the results of failed trials will benefit competitors and therefore must be avoided helps explain why industry’s research and development departments have been so unproductive in recent years. Every company wastes time and resources repeating mistakes already made by others.

Every firm -- and the public -- would be better off by sharing that information.

The HealthNewsReview gave a big "F" to this week's New York Times coverage of the Dendreon press release claiming Provenge had proved successful in treating prostate cancer. The story "fails readers in every important way," the review said.