Previous month:
September 2013
Next month:
November 2013

October 2013

Reimbursement and Pricing Policies for Medical Devices

According to “Evolving Reimbursement and Pricing Policies for Devices in Europe and the United States Should Encourage Greater Value” by Corinna Sorenson, Michael Drummond and Lawton Burns, “Europe more formally and consistently considers value to determine which technology to cover and at what price [compared to the US]”. The article also describes the policies both the United States and Europe have introduced to “provide temporary coverage and reimbursement for promising technologies while additional evidence of value is generated,” and outlines further actions to be taken “to ensure wise value-based reimbursement and pricing policies for all devices.”

While European decision-makers have shifted their focus from cost control to value-based policies over the past decade, the United States has been slower in adopting value-based decision processes and those have tended to focus on pharmaceuticals rather than medical devices.

The European system works as follows.

  • “Market approval of a device in one country [of the European Union] should provide access to other markets through the Conformité Européenne marking process.”
  • But “institutional arrangements for financing differ among countries, which can result in divergent coverage, reimbursement and pricing decisions for a particular device.”
  • “Prior to making coverage decisions, European jurisdictions typically require that high-risk, innovative or costly devices… undergo a health technology assessment” and recommends for or against use, or use for only a subset of the population, based on “clinical trial evidence and cost-effectiveness data”.
  • When the available evidence is insufficient or inconclusive, some European countries “offer restricted coverage for patients enrolled in studies designed to collect better data.”
  • “Once coverage is determined, most European countries use prospective payment systems to determine reimbursement rates,” which may reflect value in some cases or may not, because payment systems are updated infrequently.
  • To address the issue of insufficient payment for new technologies, some countries have introduced “separate or supplementary payments [lasting usually two or three years] to provide partial or total reimbursements for potential beneficial devices until they are fully captured by the payment system.”

The article also describes the American system, and emphasizes that “with limited exceptions, CMS does not currently consider a device’s comparative effectiveness or its cost relative to alternative treatment options in its pricing.” It details the concepts of “add-on payments” (for inpatient care, mostly used for implantable medical devices) or “pass-through payments” (for outpatient care, more common), which were developed to mitigate the high costs of new devices in the US. Devices must be “new and high cost” and must “substantially improve the diagnosis or treatment of beneficiaries.”

It is worth noting that “devices are rarely directly reimbursed by private insurers.” Instead, they reimburse medical procedures or episodes of care “at a specified or negotiated amount that must cover the price of the device along with other items – such as supplies, labor and facility costs – that are part of the procedure or care episode.”

The authors then outline three potential policies to improve value-based reimbursement and device pricing:

  1. Fostering pre- and post-market evidence (“current requirements allow clinical evaluations of most new devices to be based on similar existing (predicate) technologies rather than the actual device in question, and the clinical data submitted to be based on a literature review alone”),
  2. Exploring new approaches for assessing value (“devices have particular characteristics that introduce unique challenges to measuring their value” such as “frequent modifications following initial development” and the “learning curve” of medical professionals),
  3. Linking evidence of value to reimbursement (“this approach sets different reimbursement rates for different levels of clinical effectiveness” and may also entail different copayments).

Debra Dunn on Corporate Social Responsibility

To continue on the Stanford d.school's theme (with recent posts on the d.school itself and Tina Seelig over the past two weeks), today's post focuses on a lecture of Debra Dunn, another professor at the d.school who had a distinguished 22-year long career as an executive at Hewlett Packard before joining the academic world, on corporate social responsibility. (It is also available as a podcast here.)

My favorite two videos of hers are posted on Stanford's e-corner (entrepreneurship corner) website, one on career advice and one on creativity vs control, both with her husband Randy Komisar of Kleiner Perkins Caufield & Byers. Students might find those videos particularly valuable. I'm curious to know what you think!

Creativity vs control

Career advice


Reimbursement Contracts with Price Rebates from Pharmaceutical Companies

Today’s blog post is on “International Best Practices for Negotiating Reimbursement Contracts with Price Rebates from Pharmaceutical Companies”, an article published in the April 2013 of Health Affairs and authored by Steven Morgan and Jamie Daw of the Center for Health Services and Policy Research at the University of British Columbia in Vancouver, CA and Paige Thomson of Vancouver Coastal Health, also in Vancouver, CA.

The authors interviewed policy-makers in nine high-income countries to identify common practices in reimbursement contracts, which commonly involve confidential rebates paid directly from manufacturers to insurers. A simple rebate requires a manufacturer to pay back to the insurer a fixed proportion of the list price of the prescription drug. A more complex rebate may depend on the volume of drug use, the types of patients the drug is prescribed to, or patient health outcomes.

While reimbursement contracts involving rebates have been common in the U.S. in the 1980s, they are a novel practice in many other countries. The authors point that, “because many countries regulate drug prices based on what companies charge in other countries,… manufacturers are therefore increasingly using the confidential rebates… [to price discriminate, i.e.,] charge different prices to different buyers.”

 

Insurers’ motivation in using reimbursement contracts is twofold:

  1. “to achieve financial goals, lower prices and budgetary predictability,”
  2. “to address concerns about the cost-effectiveness of new medicines.”

The authors note that all of the seven countries that used reimbursement contracts (the two exceptions out of nine being Austria and the Netherlands) often used simple rebate contracts, which achieve the goal of lowering prices. To attain budgetary predictability, decision-makers should resort to volume-based rebates or schemes with spending cap.

The authors identified the following dark aspect of price rebates: manufacturers would ask health technology assessment agencies to review medicines at prices that were too high to meet the insurer’s conventional standard of price effectiveness, but would then propose confidential reimbursement contracts so that the lower-than-list prices would meet the cost-effectiveness thresholds. Also, health outcome-based reimbursement contract could “reduce clinical uncertainty” regarding new drugs with limited evidence of effectiveness, but are difficult to administer in practice.

In addition:

  • “Some policy makers believed that manufacturers submitted extraordinary prices to put upward pressure on conventional decision-making thresholds.”
  • There exists a risk that “once an initial reimbursement contract expires – which typically occurs within three years – it may be politically difficult to withdraw coverage even if the manufacturer stops offering rebates.”
  • Reimbursement contracts give drug manufacturers a “substantial information advantage: knowledge of the rebates offered to all other payers.”

The authors argue that reimbursement contract design must follow the three critical principles of pragmatism, discipline and transparency. Pragmatism means that simple rebate schemes should be used as much as possible, because of the difficulty in enforcing more complex schemes. Discipline means that policy makers must not make deals falling outside their established contracting framework, because this risks undermining manufacturers’ faith in the negotiating process. Transparency means that, although the contract details should be confidential, the existence, purpose and type of the reimbursement contract shouldn’t be.

The article also contains a short discussion of the bargaining power of large insurers vs small insurers and patients without a drug benefit plan. From the manufacturers’ perspective, “the goal of drug pricing through confidential rebates is to generate a range of prices across payers.” The authors advise that “to mitigate drawbacks from the use of pharmaceutical reimbursement contracts [such as information dissymmetry], policy makers must create policy structures that shift control in negotiations to the payer and ensure legitimacy in their decision making.”


Tina Seelig on Creativity

TinaseeligToday's post provides a few resources about Tina Seelig's teachings on creativity in the business world. Seelig, executive director of Stanford's Technology Ventures Program (and a neuroscientist by training), teaches creativity and innovation at the d.school, which I wrote about in last week's post. She is also the author of, most recently, inGenius and served as the director of the National Center for Engineering Pathways to Innovation from 2011 to 2013. I had heard about her book but hadn't read it and had never watched a video of hers until I came across her TEDxStanford talk below.

This podcast provides an extended version of that same talk given elsewhere. The only drawback is that the visuals she used aren't provided, but they are shown in the TEDx talk above (you definitely want to have watched the TEDx talk first, though!)

She makes excellent points for beginning to end, and trying to summarize her speech wouldn't do her work justice. (Thumbs up for the quilt-making analogy and the Möbius strip, though.) Let me just say that every moment of her TEDx talk is packed with useful information, and she knows how to tell good stories.

Another good speech of hers was recorded at a @GoogleTalks event.

Enjoy!


Healthcare Cost Containment

Two recent articles in my favorite publication, Health Affairs, provide ideas to tackle health cost growth. The first one, by Paul Ginsburg (Achieving health care cost containment through provider payment reform that engages patients and providers, May 2013 issue), advocates for reforms to provider payment along with stronger incentives by Medicare and Medicaid to reward such new providers’ contracts, discourage non-participation and engage beneficiaries, “perhaps through incentives for patients to enroll in an accountable care organization and to seek care within that organization’s network of providers.”

In contrast with current efforts in the private section such as value-based insurance design and reference pricing, “Medicare has rarely used incentives to direct beneficiaries to more efficient providers. Hospital deductibles – a form of cost-sharing – are uniform nationwide. Coinsurance amounts vary only slightly across providers… Widespread use of supplemental coverage protects [beneficiaries] from these costs.” It also “lacks levers to shift volume to lower-cost providers, and there is no opportunity for providers to gain volume through greater efficiency.” The author mentions a proposal by Steven Lieberman where beneficiaries would enroll in, as opposed to be attributed to, a specific accountable care organization, and “could be offered lower Medicare Part B premiums and lower patient cost sharing for using providers in the ACO’s network”. They would also be “offered a share of the ACO’s savings as a premium rebate.”

The second article is the actual proposal by Steven Lieberman, published in the July issue (Reforming Medicare Through ‘Version 2.0’ of Accountable Care). It would enhance ACOs’ ability to manage care while requiring them to accept increased accountability and financial risk. The proposal would make it ACOs more likely to succeed, compared to the current situation where “they do little to engage or reward beneficiaries, and they give providers only modest incentives to participate, permit ACOs to avoid responsibility for overbudget spending, and restrict mechanisms that can lower costs and improve quality.” The new system would encourage widespread participation by providers and beneficiaries and eliminate the looming cuts in physician fees for the physicians participating in ACOs. Adjustments to fees paid to ACOs would occur on a quarterly rather than annual basis.

An ACO could choose between two possible methods to pay providers: (1) the Centers for Medicare and Medicaid Services (CMS) would pay providers directly using the applicable reimbursement, or (2) CMS would pay the ACO, which would then distribute payments based on provider contracts and shared-savings rules. The second option would require CMS’s pre-approval and would only be applicable for ACOs satisfying certain conditions, such as enough capital to bear the proposed level of financial risk and an adequate network of providers. There would be special rules for patients living in areas with at most one ACO, so that the patients would not penalized for selecting fee-for-service if they have little other choice. The article further explains how to make the proposal budget-neutral.

Stanford d.school

Dschool logoI've been researching Stanford d.school (the Hasso Plattner Institute of Design at Stanford) for a part of my book on engineering innovation, and it seems clear to me that, in the short time since its inception in 2005, it has established itself as a hub for creativity and innovation open to all Stanford (mostly graduate) students. MIT's Media Lab has its own mystique among the techies, but the d.school has a broader appeal - it aims at teaching everyone to be creative and even holds executive education programs - and it has very quickly become a media darling. I wasn't sure what to make of design thinking at first, but after browsing through some of the resources online (and the school distinguishes itself by its excellent website and its willingness to put free resources available to all) I am becoming more and more convinced that design thinking is an important concept many of us would benefit from using in our line of work. I would be surprised if it doesn't turn into a buzzword, the way "paradigm shift" used to be a few years ago and the way "innovation ecosystem" is now. 

The first resource I'd recommend is the Bootcamp Bootleg, a pdf file gathering many of the ideas behind the design thinking process. The following steps will be relevant to a wide range of people, but in particular consultants (and definitely not just designers):

  1. empathize,
  2. define,
  3. ideate,
  4. prototype,
  5. test.

Empathize is about seeing the world through the end user's eyes to better identify his need (which will be defined in step 2) and understand how he will use the product. I continue to be amazed by the number of engineering consultants who create a product, such as a software system, without being in close contact with the people who will ultimately use it - talking frequently to other people at the company, perhaps, but never bothering to chat with the end users. And then they wonder why their amazing software is left unused. It also seems to me that many college students are rarely trained in engaging with people outside their peer group, especially customers or end users of their work. Defining the need (step 2) and exploring solutions (step 3) are also musts in the consulting world.

I wish prototyping (step 4) were more common. I always tell students to start small when we work on optimization models but I'm not sure if any of them, once in the real world, keeps the same habit - perhaps through no fault of their own, since they will follow the guidelines of the project manager, and not everyone understands why starting small is important. (In addition, if a project is running behind schedule, the temptation to cut short the prototyping part is very strong.) Testing (step 5) is also something I hope is done thoroughly by consultants before they submit their deliverable, but its importance is perhaps more intuitive and obvious than that of step 4 and thus it is less likely to disappear from the actual process.

The d.school offers a free 80-min virtual course online called the d.gift, because it aims at redesigning the gift-giving experience. It is something you want to do with a partner - watching the video will bring you nothing, since the workshop leaders provide only brief instructions and then much of the time on the video will be spent watching a clock count down the time on the screen while participants interview their teammate, sketch a prototype and provide feedback. Here is the video.

I also enjoyed 3 smaller pdf files called "mixtapes":

If I were a graduate student at Stanford, the d.school is where I would spend all my days. But since I'm not, perhaps I can bring design thinking ideas to the world of higher ed. What about you?


Lehigh's Presidential Scholarships

Lehigh University offers something called Presidential Scholarships, also known as the "free fifth year", to undergraduate students who graduate with a GPA of 3.75 or higher, so that they can work toward a second undergraduate degree, a graduate degree or even a non-degree program at Lehigh without paying tuition that extra year.

Students can register for as many credits as the university allows during each semester of their scholarship, but they only have one year to enjoy the free tuition. Needless to say, this motivates many students to do Master's programs that other students take one year and a half to two years to finish in a year sharp, and in fact often two semesters if they are doing internships during the summer. We are talking about fifteen credits or five graduate-level courses a semester when the regular full-time load is three courses or nine credits. (A Master's is thirty credits. To give some perspective, students on teaching or research assistantships aren't allowed to take more than ten credits a semester.) Because the Presidential Scholars represent the best Lehigh has to offer, they usually manage to complete their degree on time, but they often talk about the hectic-ness of it.

Such a tight schedule has many disadvantages, the first of which being that, if they pick a Master's in a discipline different from their undergraduate degree, they have to go through the interview process only weeks after they've started in the new program and thus are unable to make the most of the new credentials they are about to gain. (I am thinking for instance about students who majored in industrial and systems engineering as undergraduates and decide to pursue a Master's in analytical finance for their free fifth year, but I imagine the situation is the same if they pick the Master's in Technical Entrepreneurship, or in Healthcare Systems Engineering - something close enough to their undergraduate major that it makes sense for them to pursue as a Master's degree, but not in direct continuation of their previous studies.)

The second disadvantage is that it is a loss of opportunity for Lehigh not to require the Presidential Scholars to choose programs where they have to do at least a capstone project (or even a thesis where applicable, or a two-semester project). Lehigh trained those students and they have proved to be exceptional. From Lehigh's perspective, the whole point of encouraging them to stay for a fifth year should be to help the university shine at the graduate level, which they are in a clear position to do - you won't find better students than that. In fact, MIT created its Master of Engineering program in Electrical Engineering and Computer Science, reserved to MIT EECS undergraduates, with precisely that end in mind.

It is very nice to let Presidential Scholars get free all-coursework degrees, but it really doesn't particularly help Lehigh showcase what its students can do - especially its best ones - at a level often overlooked by universities: the Master's students. (Master's students always get short-changed because they end up spending much less time on campus than undergrads or doctoral students. A pet theory of mine, however, is that the ones who get a Master's in engineering play a critical role in enabling engineering innovation.)

Thankfully many Master's programs have a project or thesis requirement - however, not all of them. This is most likely because programs with project or thesis are hard to scale, and in terms of revenue generation, having a project requirement hinders program growth and revenue growth, since you have to find projects for all the students you enroll. (It is worth pointing out that Master's students besides the Presidential Scholars have to pay the full price, in general without financial aid. This is something universities care about.) Administering capstones certainty is time-consuming, but overall letting the best undergraduates take a free fifth year to add more courses isn't particularly helping Lehigh raise its visibility among its peer institutions or showcasing the quality of its innovative research.

It makes little sense to me to force the students to complete their Master's degree in one year (technically, they're not forced - if they're not done in a year they simply have to pay tuition out of their own pocket, after four years of doing just that, and that is something most students very much want to avoid, especially if they have prior loans). A year and a half would be so much better. And while tuition wouldn't be known for the second year, universities like George Washington charge undergraduate students a known tuition rate depending on the year they enter, so making a commitment for eighteen months instead of twelve shouldn't be the end of the world. Besides, Lehigh always takes some level of risk because it doesn't know how many students will choose to enroll in the program. I can't see any clear argument in favor of one year, assuming that the maximum number of credits students can take is capped.

I also believe there should be a tight network of Lehigh's Presidential Scholars. We hear sometimes in the university newspaper about the Rossin Junior (Undergraduate) Fellows, who organize activities for the whole group, help underclassmen understand their major and represent Lehigh. Why aren't the Presidential Scholars a tighter cohort? They combine advanced, graduate knowledge with impressive academic credentials, and they usually have outstanding people skills, because that is the sort of people Lehigh attracts as undergraduates. We should help them meet each other and create strong ties, because the Presidential Scholars in the College of Engineering will undoubtedly be tomorrow's leaders in engineering innovation.

In summary, Lehigh should make sure the Presidential Scholars' graduate days at the university leaves them amazing memories and gives them the opportunity to shine (it's good for them, but it's also good for Lehigh!) Their worth is abundantly clear - graduating with a 3.75 GPA at Lehigh is difficult, to say the least. So why aren't there more policies designed to help them have a wonderful, productive fifth (and sixth!) year where they can give back to the university through their cutting-edge research work? Those are the young alumni that companies will rave about. A free fifth year is nice, but Lehigh isn't getting as much out of the program as it could, and the scholarship recipients aren't either.


The effects of reference pricing for orthopedic surgery

The August 2013 issue of Health Affairs had a fascinating analysis of the impact of reference pricing on the use of and prices paid for knee and hip replacement surgery. The authors, James Robinson and Timothy Brown, both affiliated with UC Berkeley, compared usage and procedure price data for the California Public Employees’ Retirement System (CalPERS), which started using reference pricing in January 2011, with data for Anthem Blue Cross, which does not use reference pricing and served as a comparison group. The period studied was 2008 to 2012.

CalPERS identified 41 value-based purchasing design (VBPD) facilities, that is, facilities that satisfied the following criteria: (i) procedure prices were less than $30,000 (for the hospital’s allowed charges only, not including physicians’ fees, but such fees were found to show little variation and thus were not a primary focus of the study), (ii) quality was acceptable and (iii) the group of hospitals provided sufficient geographic dispersion. Employees selecting a VBPD hospital still had to pay 20 percent in co-insurance, up to $30,000, but if they selected a non-VBPD hospital they also had to pay the amount in excess of the $30,000 threshold.

The authors reach the following conclusions:

  • “In the first year after implementation, surgical volumes for CalPERS members increased by 21.2 percent at low-price facilities and decreased by 34.4 percent at high-price facilities.”
  • “Prices charged to CalPERS members declined by 5.6 percent at low-price facilities and by 34.3 percent at high-price facilities.” Specifically, “the average price charged by non-VBPD facilities as a group declined dramatically, from $43,308 in 2010 to $28,465 in 2011.”
  • “There was substantial variability across individual hospitals in their pricing strategies after the implementation of reference pricing. Almost half of the hospitals not designated as VBPD by CalPERS… continued to increase prices in 2011, while half reduced their prices. However, the average price reduction was more than twice as large for the facilities that reduced prices ($11,048 per patient) as the average price increase for those that increased prices ($4,097).”
  • “In 2011 reference pricing accounted for $2.8 million in savings for CalPERS and $0.3 million in lower cost sharing for CalPERS members.”

The authors do raise the question, at the end of their paper, of “whether the observed reductions in prices for knee and hip replacements were offset by price increases for other services,” although interviews with CalPERS and Anthem staff suggested they had not. The price trends may also be modified as time goes by. Price data should be gathered for longer periods of time to truly ascertain the impact of reference pricing. But it is undeniable that this cost-sharing method (identifying high-value hospitals but letting plan enrollees choose any hospital provided that they are willing to shoulder the cost difference) has significant potential and holds more promise than high-deductible plans, which unwittingly incentivize not receiving preventive care.

I found this article highly informative and well-written, and highly recommend it to anyone interested in reference pricing.


Malcolm Gladwell on Elite Institution Cognitive Disorder

Malcolm Gladwell, of Outliers, Blink and The Tipping Point fame, has a new book out this week: David and Goliath. He recently talked about a central concept of his book, applied to the number of students who get STEM (Science, Technology, Engineering and Mathematics) degrees. Once you make it past the first 2 minutes where he comments at length on not being paid by Google to give the speech, his talk - on the counterproductive desire to associate with elite institutions - is downright fascinating.

Although he gives it the fancy-sounding name of Elite Institution Cognitive Disorder (EICD), this is really the Big Fish in a Small Pond (BFSP) syndrome: apparently, a lot of people are more likely to stick to their career plans - such as getting that STEM degree - if they can compare themselves favorably to the people around them. In other words, they're better off as Big Fish in a Small Pond. Gladwell has some fascinating data regarding the distribution of STEM degrees depending on students' math SAT scores at institutions of varying caliber (top college vs middle-of-the-road). In particular his data suggests the distribution is rather stationary, that is, doesn't depend on the institution, although the students at the top institution are supposed to be "smarter" than the students at the not-so-well-known college. This is why Gladwell argues that people make their decisions by comparing themselves to the people around them, and lose track of the big picture.

Before I make any further comments, I'd like to encourage you to watch his talk, which lasts about 20 minutes.

Gladwell's thesis, at least in this short talk (I haven't read his book), seems to be that most people are better off away from elite institutions. Perhaps there is some truth to what he is saying about academia. I sometimes read papers that make, at best, marginal ("epsilon", as we like to say in my line of work) improvements about the state of knowledge - papers that no one at the university I got my PhD from would have dared consider research papers. And yet their authors submitted them for review, and they got published somewhere. Not in the best journals, but somewhere. They most likely have tenure and while not leaving their mark on their field in the way an academic at a top institution would, they might enjoy a comfortable lifestyle and perhaps a reputation as a good teacher or a valued committee member. Because they might have been the very best students in their program, they may have benefited from the strong support and encouragement of faculty members whose counterparts at a better-ranked institution would not even give such students the time of day. For them, staying away from people who would've diminished their self-esteem definitely paid off.

The bad part of all this, of course, as Gladwell point out, is that not-so-great PhDs from top institutions - who'd be the cream of the crop elsewhere - may decide, judging -- from their more successful peers' examples and their adviser's lack of encouragement (their adviser, after all, is unlikely to know of openings in the second-tier schools that would be a good fit for the student, and even if he did, unlikely to care) -- that they are better off leaving academia altogether. They do not realize that plenty of academics with worse credentials abound and even thrive, having found other strengths to contribute to their workplace. I guess we can call this the false negative of the career path read: some people draw wrong conclusions about their own potential, career-wise, from their immediate environment.

But what about people who rise to the challenge? The 15% of the bottom tier, using Gladwell's numbers, who do hang in there and get their STEM degree, although people with better math abilities (assuming we can judge math abilities by SAT scores, which is highly debatable) drop out of the major? Perhaps they succeed precisely because they were well-aware they didn't have the best SAT scores - because they showed resilience and persevered while the better students, unused to struggling in class, became discouraged and picked another path.

The issue, to me and hopefully to a lot of people, isn't that more people should be advised to "play it safe" and join groups (whether universities or companies or even groups of friends) where they won't feel their self-image so challenged. Pretending employers do not care which school students graduate from, or which company they worked for before, is either naive or disingenuous. The real issue, to me, is that we need to develop long-term resilience and a sense of perspective in students, which they will hopefully take with them when they graduate. But a system where students, worried about their GPA, will take classes with an easy A, and where professors, worried about their evaluations or unwilling to deal with struggling students, allow their course to turn into an "easy A" or "automatic A" course, does not do those things very well. There is also, naturally, a competence issue, where getting a C at Harvard just isn't the same as getting an A at Noname College. Perhaps the rise of MOOCs (Massive Online Open Courses) will allow students unsure of their prospects a different perspective before they change majors. As online education is exploding, it seems more and more doable for universities to adopt a more flexible approach.

Although Gladwell seems to suggest these students should pick a college where they will face less competition to get their prized STEM degree with flying colors, what matters just as much as the degree is the social network - the people students choose to associate with, their goals, their ambitions. If you opt to become a Big Fish in a Small Pond, you might feel very pleased with yourself, but you'll never know what you would have been capable of if you'd pushed yourself to your edge instead of spending your weekends getting drunk in small-town bars.

Being a Small Fish in a Big Pond opens horizons. It is also, of course, immensely unsettling, and it should come as no surprise that many would-be actors, for instance, leave New York City after a few years to pursue other endeavors, whether acting-related or not. (A number of them are now finding work as readers of e-books for the Audible service.) Yet it would not occur to most of them to forgo the obligatory move to New York to give the big leagues a try. The behavior of the students changing paths to succeed in the big leagues (switching majors but remaining at Harvard) makes a lot of sense to me. Now we just have to convince those students that they're still in the big leagues if they struggle to get their degree.