An interesting feature of writing for DDNews is receiving kudos, pithy complaints and questions. Here I include some of recent questions. The answers have no clear validation. Opinions, thankfully, don’t require that.
Q: Why are you so circumspect about large data sets?
A: Because I continue to get invited to give plenary talks at nuclear engineering meetings and to join the editorial boards of polymer chemistry journals. I know nothing at all about these subjects. The artificial intelligence that invited me clearly needs further work. My backup gal, Amazon Alexa, is getting smarter and smarter each week. She listens and learns, but has yet to invite me to a conference in Paris or Davos.
Here is my current business plan in verse: ”Genotype guides, phenotype decides, N=1 gets the CDS job done.” Massive amounts of data are often derived from invalid (or at least inconsistent) methods testing poorly characterized samples. My academic colleagues frequently complain about a lack of access to well-documented samples, collected and stored to maintain their integrity. I then suggest it is preferable to have fewer well-characterized samples then 100,000 of dubious origin. If you can’t distinguish health from disease in 10 patients vs. 10 controls, the data is likely to be useless to a physician. What would be gained by 50,000 patients and 50,000 controls? As a rule, for quantitative data, the bigger the data file, the less consistently it was collected. We are perhaps on the cusp of using clinical decision support (CDS) in the form of information technology to augment (not replace) judgment. This will bring big data together with small data. The FDA issued a draft guidance on CDS on Dec. 8, 2017 (titled “Clinical and Patient Decision Support Software”). Clearly there is opportunity for bias in developing algorithms and a challenge to keep them up with the latest science. We need to be very careful of settled science as it tends not to be settled for very long. The guidance nicely deals with these issues.
Q: So few academic projects seem to help patients at the end of the day. Why is that?
A: This is a correct observation, but why single out academics? Most research careers in business or government don’t appreciably change the world either. I believe that academics want to have impact beyond educating students (a great thing itself!). It’s not that they don’t; it’s just that it is hard. Impact is rare. It takes thousands trying things to get even a modest hit. Professors are themselves part of a library, perhaps “a library of librarians” to be screened against targets. Hitting the bullseye on the target is very rare. Those who don’t get a hit perhaps provide a service, making those who do hit more prominent. Hiring and tenure decisions are human activities. We try, but we have no reliable biomarkers of success. I’ve seen researchers who really struggled with tenure then hit on a result that helped millions of patients. I’ve seen others sail through the system and then get stuck in an area that became a backwater. They didn’t escape from a comfort zone of what they knew from grad school. I’m not saying they were wrong to go narrow and deep, but that they didn’t get lucky. Most of us don’t. Planning is overrated and luck is underrated. The late Indiana medical device maven Bill Cook was fond of saying “ready, fire, aim.” That’s good advice.
Q: What are your thoughts about the tax bill signed into law last December?
A: Like me, it’s imperfect. There are potential benefits for many in the short term and significant liabilities for our grandchildren unless some discipline on the spending side also comes to pass. I’d prefer to reduce corporate income tax to zero and fund government with contributions from the individuals who would thus benefit, including employees, shareholders and customers. If 21 percent is better than 35 percent, 10 percent or zero percent would be even better. Ireland likes 12.5 percent. Nevertheless, the reduction has already allowed for rewards for individuals and early signs of more investment here in the United States. Animal spirits have been set free for only a bit over a month.
Reduced regulatory burdens are also helping. Perceptions matter and they have improved. Congress takes advantage of busy voters and business alike. Taxing an inanimate object (a business) is a means to fool many people much of the time, the very same people who buy from the business, own the business with their retirement funds and work at the business to support their families. The perfect should not be the enemy of the good and what we have in this new year looks good. Biopharma will benefit and that means patients, employees and shareholders.
Q: Several nonprofit hospital systems reportedly are looking at manufacturing generic drugs. Is that good?
A: It’s interesting. For one thing, this is now more easily accomplished than was once the case. We have a very rich ecosystem of CROs and CMOs who know the drill very well and can make this happen without those hospitals investing in fully integrated manufacturing. The hospitals are responding to a shortage of drugs and recent generic pricing kerfuffles. This is a free enterprise response to a problem and I like that. What I don’t like is the notion of nonprofit businesses competing with businesses expected to pay taxes. No business lasts if its expenses exceed revenues for very long, whether hospitals, banks, universities or pharmaceutical companies. Why should any of these pay income taxes? I bank at a credit union and that’s perhaps not fair to the commercial banks I also use. Only people pay taxes, although businesses, nonprofits and government units are all expected to collect some which are visible (see your pay stub) and then hide others from view. Now to the question: A large consumer products company in Cincinnati, Ohio, thought they should be in the drug business as did a photographic film company in Rochester, N.Y., and a chemical company in Wilmington, Del. How well did that go?
Q: How did healthcare expense turn out in 2017 and where did the money go?
A: This was the first quiz question I always ask my class of very bright juniors and seniors. Many are already accepted to medical school and others to Ph.D.programs. I use the same question every year with the same result. They have no idea. How many readers of DDNews know the answer? One smart student responded with a question, “How could we find out? We could go Googling.” That brings me back to big data and whether it can be trusted. The number for 2017 that I see most often is $3.5 trillion (12 places beyond the decimal point). Pharmaceuticals (prescription) are reported to be 12 to 14 percent of this by most accounts, with low numbers from the pharma industry and higher from payers and complainers. The econometric calculations are complicated (at least for a chemist). The supply chain is incomprehensible to mortals and there are many special deals. What are drug prices? It’s hard to know. The distribution system is a spaghetti bowl of variable pricing, rebates, discounts and charity. You can pick a number fitting your narrative. Drugs are either a fabulous money-saving tool for preventive medicine at any price, or a greed-driven means to charge a lot for little gain. It depends.
In a parallel with university tuition, there is a list price, but many don’t pay that or anything at all. There is an analogous problem with in-vitro diagnostics, which most put in the range of 4 percent of healthcare spend. Once again, the bill is very different based on who does the work, where it is done and how fast it is done. Attempts are made to set a reimbursement rate at the lowest possible level without taking into account these process features. For example, the cost of a measurement in a single sample with LC/MSMS is a lot different than the cost in a batch of 1,000 samples where speed is of limited utility. Hospitals use more than $1 trillion of the healthcare spend. I suspect the thought of vertically reintegrating to manufacture pharmaceuticals for their captive use is tempting. Remember that Procter & Gamble saw wisdom in going beyond toothpaste to prescription drugs. A lot was learned.
A: I like him. He arrived in spring of 2017 and got right to work. The ball was moving, and he picked it up and kept it going. He took the 21st Century Cures Act seriously, implementing the new user fee decisions from Congress (FDARA of 2017), and is working to speed up both generic and new molecule approvals. Gottlieb has the right combination of experiences as a physician, cancer patient and venture investor. He also had prior in-house FDA experience. The FDA is a Health and Human Services agency, and thus Dr. Gottlieb has a new boss, just confirmed in January. Alex Azar is also well-credentialed in government and more recently as a senior executive at Lilly. He’s a lawyer where a lawyer can be helpful. I’m anticipating that this team will perform well together, but we’ll have to see.
Both leaders recognize that developing drugs is very costly. Both promise to explore improved efficiency. Add that to the 21 percent tax rate and we have optimism for change. In addition, the excise tax on medical devices expected to crank up on Jan. 1 has now been delayed again for two years. Good, but not good enough. When the support to eliminate the tax is so broad, delaying is not a substitute for eliminating. This is important for DDNews readers since more combinations of devices and drugs are on the horizon. From Davos, Gottlieb suggested he’s looking at devices and diagnostics—particularly companion diagnostics—to experiment with the regulatory framework to get them synchronized with drug approvals. Distinct categories are now overlapping more frequently. Azar has budget of $1 trillion. Those many DDNews readers with NIH grants share in a budget of $0.032 trillion. I hope he’ll have time for you.
Q: $850,000 for a gene therapy?
A: My colleague Jeff Bouley gave thoughts on the topic in his January issue editorial. His theme included a desire for both more transparency on R&D costs and the economic benefits of treatment. Neither (cost and benefit) are measurable with much certainty. Both can be made to fit a “high” or “low” narrative by the categories selected. The published price is for a single course of therapy, an injection in each eye. I am assuming here that the cost of administering the drug and following any progress is not included. This reminds me of other small-volume medical procedures. For example, a heart transplant has a sticker price estimated at $1.5 million, and the number is in the same category as anticipated for treating this rare category of blindness. Cardiac assist devices, immunosuppressant drugs and advances in surgery and organ preservation procedures have all helped make the impossible possible. We do run out of hearts. It’s early for a gene therapy. How will the cost really end up and evolve over time?
Jeff referred to the popular notion that “corporations are greedy and heartless at their core.” Anthropomorphizing corporations is convenient, but wrong. All who say this are not paying attention to their own greed with respect to their pay, their retirement funds, their United Way contributions and the medical benefits they derive from these mean and unprincipled beasts. Integrate over time. We will not know how gene therapies turned out for another couple of decades. We were not meant to fly, either, but we did!
Peter T. Kissinger (who can be reached at firstname.lastname@example.org) is professor of chemistry at Purdue University, chairman emeritus of BASi and a director of Chembio Diagnostics, Phlebotics and Prosolia.