Thursday, September 19, 2013

A Surprisingly Cost Saving Program--Based on Incentives

Here is an interesting story from NPR about what happens when a simple change in incentives is brought to a population.  The gist of the story is this.

In Cardiff, Wales, a lot of people were reporting to the emergency department with injuries and not reporting to the police.  The hospital did gather information about the location of the injuries and what caused them.  The leaders and people in the local area noticed how much of a toll the violence and injuries were taking on the population.  The hospital changed one thing to help police.

It did not report names.  It only reported locations of incidents and types of injuries.  The police took it from there and focused their prevention and enforcement efforts in areas that were hot spots.

The effect--spending of less than $400,000 saved over $11 million in medical and legal expenses.

That is pretty amazing.  There are a few things to think about.

First, what actually happens to the resources that are freed up?  In other words, what else are they used for?  Hopefully something more productive, but they don't really say in the story.

Second, why does such a simple incentive work?  Hospitals could not report names.  And it is somewhat amazing to think that individuals are actually sufficiently rational to choose whether to engage in violence in bars and pubs as a function of the likelihood of being caught.

Of course, this is not the first study to show that people respond to incentives--even about crime prevention.

Perhaps the idea of using simple changes in incentives with for public health measures--where the possibility of enforcement is understood, not everyone gets caught, but there are significant penalties for those who do--could encourage better public health oriented behaviors.  

Thursday, September 12, 2013

Prescribing Fruit and Vegetables

This morning I heard an interesting piece on NPR's morning edition.  The idea is that physicians in New York City can prescribe fruits and vegetables and the individuals to whom they prescribe these will get "Health Bucks" to use at the farmers' markets.  The health bucks are allocated as $1 per family member per day for up to four months.  So, a family of four would get $480 to spend at farmers' markets over four months.  In addition, they get extra education.

Is this a big deal?  Well, on a per-person per-month basis this adds $30 that can be spent only at farmer's markets.  That may be a big deal for some families.  They may be able to access quality fresh fruit and vegetables they would not be able to otherwise.  But, the $30 alone might not do much.

What would happen without information?

Are families able to make trips to the farmers' markets on a regular basis?  There is still an opportunity cost of time.

Maybe it is just enough to get them to try fresh fruit and vegetables after which they engage in better dietary behavior themselves.

There is testimony from one participant and from one physician.  Does that prove the program is successful for everyone?  Well, no program is successful for everyone.  It does offer some proof of concept.  And perhaps for $30/person/month this is a relatively inexpensive addition to a weight control program.  It would be interesting to see how well this generalizes and what options other than farmers' markets could be used when there are fewer available or the travel distances are greater.

Wednesday, September 4, 2013

Observation Status

This morning, I heard an interesting piece on Morning Edition on NPR about hospitals using "observation status" rather than admitting individuals as inpatients (the story can be found here:  I listened with interest and thought about unintended consequences--the heart of so much of the study of health economics.

The key message here is that the government (the Center for Medicare and Medicaid Services, abbreviated CMS) has a program in which it reviews old hospital admissions (up to three years back) and determines whether the patient should have been an inpatient or outpatient.  If the patient was admitted as an inpatient and should have been (based on the CMS standards) an outpatient, then the hospital will have to repay the reimbursement it received and will likely get nothing back (according to the story).

Holding a patient in observational status (which we would generally expect would be just for a day or less, perhaps two days) avoids the determination of whether the patient needs inpatient care.  Some patients have been held in observation for even longer.  If their care is essentially inpatient and they recover, they can then be sent someplace else.  So far, if they do not need the full inpatient care this would seem to be good for CMS, good for taxpayers, and good even for the patient to the degree that the patient has copayments for which she is responsible.

The hang up (and the unintended consequence is this).  Patients are only eligible for Medicare to pay for nursing home services if they are discharged from an inpatient setting.  So, now patients who still need fairly skilled care but who do not need hospital care are not eligible for having the nursing home care covered by Medicare.  Who pays?  The family in most cases.

So, the government implemented a policy to control costs.  This led to a rational response by hospitals to avoid losses.  This leads to a lack of ability to discharge to nursing homes and/or get high quality nursing home care as patients and their families cannot pay for it.  This was not part of the intended cost control but is part of the impact.

The solution--perhaps allow Medicare to cover any rehabilitation-related care regardless of prior location.  But what unintended consequences might this have?  I will leave readers to ponder. 

Thursday, August 29, 2013

A University Giving a Strong Incentive for Better Health Maintenance

Penn State University (by disclosure my undergraduate Alma Mater from which I received an award as alumnus of the year from the Honors College and for which I have mentored numerous undergraduates during my professional career--so I may not be completely unbiased in my assessment in this post) has recently announced that it is making a major change to its health insurance plan.  There is a good article in the Pittsburgh Post-Gazzette that can be accessed here (

The gist of the plan is this--the University has set up a program called "Take Care of Your Health."  Employees are being asked to fill out some health information (they call it a wellness profile), schedule a physical, and schedule a series of lab tests to obtain "biometric" information (using the University's language) including lipid levels, blood pressure, body mass index, and other information.

What is the incentive?  Often times organizations will give their employees a discount for compliance.  And that discount is often rather small but enough to get people to do something as long as their preferences do not lead them to feel like they should not do it (perhaps employees would feel like this is too much information to share with their employer despite communication that the purpose is not to share it with the employer). However, this is not a positive incentive.  This is a very negative incentive.  Employees who do participate will go on paying the same premium they would have paid already.  Employees who do NOT participate will have to pay extra.  And not just a little extra.  They will have to pay $100 extra.  And not just once.  But every month.  That is $1200/year for each adult member--employees and spouses.

That is a big deal.  That would be more than a 1% decrease in income for a large number of university employees.

That raises a series of questions.  First, why are people not getting physicals and lab tests now?  Probably a two part answer there.  I don't have direct access to information on what Penn State's health insurance benefits pay for, but the physical may not be free and the labs are likely not free.  There is a money cost.  Also, there is a time cost.  Making the appointment. Getting to the doctor.  Waiting.  The time for the exam.  Getting home.  There may be child care costs.  There may be costs to coordinating with a spouse or partner.  The list goes on.  Both time and money are limited resources.  And if a person feels he or she is generally healthy and at little risk of an adverse outcome, then he or she is unlikely to feel the need for a physical.

The same thing goes for lab exams except even more in some ways.  While many health plans will cover one wellness driven physical per year, there is almost always a charge for lab tests.  These take time.  And they can sometimes be inconvenient--in terms of being required to fast beforehand, for example.

Second, is the $100/month enough to get people to change their behavior.  I can't predict for certain, but that would seem like a sufficient change in price (not just at an absolute level but probably in relative terms) to get many people to change behavior.  I would bet that a lot who end up paying the $1200/year will probably do so because they forget to take care of things in time.

Third, i have to ask myself what will be done with the $1200?  Or, conversely, what does having the self-reported and testing information give the University that they would not have otherwise. One statement is that they are not trying to be invasive into everyone's life but to save money overall by finding individuals who are likely to experience inordinately expensive medical care events based on a predicted probability.  The premise of requiring any program like this (where even if the individuals did not think that the $100/month charge was unreasonable or everyone filled out the form) is that the cost of requiring the form to be filled out, the data storage, the data analysis, and the data safety would be offset by avoided events.  I am not convinced that this has been shown to be the case in all examples that have been studied, but it is something to think about.  The $100/month is not necessarily an indication that the University predicts that the individuals who do not participate are more expensive.  In fact, they may be some of the healthiest, lowest cost individuals to the system.  However, the university is implicitly suggesting that for individuals who are not willing to share information, the university is going to shift some of the cost of transferring the risk back to the individuals.  Why do I phrase it that way?  The university is self-insured.  In other words, it pays money from its account and from employee paychecks into a fund it holds to later pay health care bills.  If they make good predictions the are just about right every year in predicting expenditures.  The university pays some and the employee pays some of what goes into the fund.  The university will still pay some for the employees who do not participate in the health maintenance program--just less.  The employee is not bearing any more risk--they know exactly how much more they will spend on health insurance.  They are just paying more from their own pocket to transfer the risk.

This is a complex and somewhat indirect way to get people to change their health maintenance investments and behavior.  It will be interesting to see whether this "big stick" approach works when the stick encourages behavior that helps predict big expenditures but affects behavior at a point several steps removed from the behaviors that actually produce large expenditures.   

Tuesday, August 27, 2013

MBA Creativity

Since my last entry in this blog, I have moved from being a faculty member at the Johns Hopkins Bloomberg School of Public Health to the Vice Dean for Education at the Johns Hopkins Carey Business School.  Last Friday, we had the Student Resource Fair at the Business School's Harbor East campus in Baltimore.  There was one organization called Creativity by Design that represents the efforts of students in a joint program that includes the Business School and the Maryland Institute College of Design.  I introduced myself to many of the student organization leaders that afternoon and this organization's leaders were no exception. So, I asked the person with whom I spoke to tell me what his organization was trying to promote.  He said that one of his biggest goals is for other MBA students not to introduce themselves as not being creative.

I found this an interesting comment.  Clearly, not everyone is artistically creative.  Not everyone is musically creative. Not everyone likes to write creatively.

But I found it hard to imagine that anyone in training in a business school (particularly in an MBA program with a goal of being entrepreneurial) could be uncreative.  Everyone who wants to be an entrepreneur has to have some type of creativity.  To invent a product.  To bring the product to market.  To market the product.  There are just so many ways to be creative.

In the process of producing and selling a product, the person who is not creative would either (a) never have a product to offer, (b) never be able to figure out a unique angle from which to market in the crowded marketplace that exists today, or (c) never be able to develop a new way to deal with age old business problems that include how to address the customer or client in the most effective way.

Would some people likely have more creativity than others?  Of course.  But no creativity?  Not likely.  And if a person truly had no creativity they likely would not be very successful and someone who is more creative would likely adapt an idea and take it to market in a way to make a lot of money.

So, I agree with the assertion that no one in the MBA program we offer ought to be introducing himself or herself as not creative.  Instead, each student should find a way to make the most of the creativity they have in benefitting themselves and benefitting humanity.       

Wednesday, February 20, 2013

Helping Others Change Their Behavior

A lot of what I blog about related to health economics is about people changing their own behaviors or the government/insurers/employers giving people incentives to change their behaviors.  In this morning's Johns Hopkins Bloomberg School of Public Health newsfeed, there was a link to a New York Times blog piece about wanting to change others' behavior.  In particular, this entry was about changing parens' behavior when "parents" are in their 60's (or older) and are not taking good care of their health.

The blog piece was written almost exclusively from a psychological change point of view, although there were a few references to things that could be done inexpensively.  (For example, a trip to the park with grandchildren that would involve a lot of time walking.)

However, most of the steps could be interpreted hyst as easily in an economic context as in a psychological context.

People make decisions based on the information they have available.  One thing that people may need is more information.

There are varying ways of providing additional information.  Some have a high disutility.  This could be from the way in which the information is shared (a patronizing tone of voice) or the time cost of obtaining the information (here, read this book).  The less disutility from obtaining the information the better.

There are varying costs of engaging in healthier activity.  If a healthier diet could begin with fixing a meal for older parents or taking them out with the family (or just grandkids) for an activity they like anyway that happens to have a health benefit, then the marginal cost of engaging in the healthy behavior is diminished.

So, some simple economic principles--increase information in ways that do not create large disutilities and that involve minimal marginal costs in terms of time or money.  With any of these, people would be expected to make a shift toward the now less costly healthy behaviors.  The key question is how much the behaviors will change in response to the changes in incentives.  And for that, we would need empirical evidence (like elasticities of demand for healthier activities or meals when the time and monetary prices change) to supplement the economic theory.

The key here is, as I raised at the beginning, that instead of complete self-motivation or governmental inposition, we are now relying on private citizens who are willing to use some of their own resources to benefit others.  Thinking from a societal perspective, the final key would be to determine how to avoid having the government (if it also wanted to play a role) crowd out what individuals are willing to do for each other.  

Thursday, February 14, 2013

Baby Boomer Health

The Washington Post published an article earlier this week presenting data that Baby Boomers are in worse health than their parents were at the same age.  In an era of every increasing medical care expenditures this is an interesting phenomenon to observe.

It is important to note that the study is based on survey research.  The survey question is about rating your own health on a scale of excellent, very good, good, fair, and poor.  One thing that may be happening is that people may have higher standards for what "excellent" health in middle age means.  Maybe excellent health in the generation of Baby Boomers' parents was about functionality and work.  While today's Baby Boomers see some of their peers running marathons and participating in other activities that require very high levels of health and fitness.  So, perhaps the fact that this is a relative and not absolute measure obscures some of what is going on as people may say, "I could never do that, so while I'm in pretty good health I must not be in excellent health, so I'll mark very good."  I have no data to back this up.  I am simply pointing to a difficulty with survey research.

Now, there are other important indicators here.  10% more of the Baby Boomers in middle age are obese compared with what their parents were at the same age (39 percent versus 29 percent as reported in the Washington Post article).  Could that make a difference?  You bet!  Does that mean that the Baby Boomers are any less happy?  Perhaps not, because some of what leads to obesity are choices that we make about career and family.  Different careers.  Different family activities.  Different amounts of time spent playing sports rather than watching sports.  Different amounts of time spend in the car rather than walking.  All have an impact on health.  Their impact on happiness is not as clear.

The Washington Post article also pointed out that the study noted different prevalence of diabetes and use of a cane or walker at middle age.

The article also mentioned that this may lead to a longer life with not so healthy extra years.  This is not necessarily the desired outcome.  There has been an enormous amount of work done to try to figure out how to compress morbidity.

The economic choices that we make before middle age have a great impact on out health.  Our health in middle age will shape the economic choices of the rest of our lives.  What lessons will younger generations be able to learn about planning for their future health and economic well-being?  

Wednesday, February 6, 2013

Insurers Covering Contraception

Last week's Wall Street Journal had a piece on new contraception opt out regulations--as a follow-up to what was originally required coverage under the Affordable Care Act. (I am not sure if the link I have provided is for an article free to the public but there are many sources of information on this new regulation.)

Contraceptives have multiple medical uses.  But clearly one use is to prevent unwanted pregnancies.  When it comes to whether this should be a required part of an insurance plan we can set all the religious arguments aside and look at some basic economics.

Insurance is meant to provide reimbursement when the unexpected happens.  Contraceptives are intended to prevent something from happening.  And, while we may have a discussion about just how much control each person in a relationship feels he or she has, pregnancy will only happen if two people choose to have intercourse.  Thus, the thing that is to be prevented is under the control of the individual.  This is different form exposure to the influenza virus where the person can take steps to reduce exposure and to reduce transmission even if he has been exposed, but he cannot completely avoid exposure.  A woman could completely avoid the risk of pregnancy and she has control over how much of that risk she is taking--even without contraceptives.

So, we are asking for insurance to cover something over which we have control.  Let's make an assumption that all women want the contraceptives.  (Debatable, but let's run with it for a moment.)  Then, what are we doing.  We are asking an insurer to cover the cost of something that all women will use.  This involves a layer of administration that would be unnecessary if the women paid for it themselves.

So, as a society we would spend less if women paid for it themselves.  We are setting up a policy that could lead to more spending as a result.  One might ask--aren't there better things to use societal resources on than administering a program to pay for contraception?  Or perhaps more economically speaking--if the government has decided that it is important to subsidize the purchase of a preventive good or service, is there a more efficient way to do that than to wrap it into health insurance.  And, when I think about efficiency, I am thinking about also including the costs of getting it passed in the first place and then regulating it--neither of which has been negligible.

If we want to spread the costs of paying for contraception to those who have more resources available there are other ways to do it.

If I understand what the new regulations say, if an employer opts out, individuals will be able to get an additional insurance policy at no out of pocket costs.  I am not sure of the logic on this one, but it seems like whether people in the employer are paying for it directly (if the employer chooses not to opt out) or indirectly (as I would have to think that a product with no premium would be wrapped into the administrative costs for everyone else), the employees of employers who opt out would still end up with at least some share of the cost for this--it would just be even more diffused.

In this, I am not taking a stand on whether we should be subsidizing contraception for low income mothers.  That is for policy makers, religious, and ethicists who think about the meaning of contraception to decide. What I am trying to emphasize is that this may be more costly than just having women pay for it themselves and we are wrapping what appears to be a public health issues (or at least a population health issue as we look at the health of the population including mothers and newborns) into a private health insurance mechanism that may or may not be the most efficient way or helping to facilitate access if that is the goal.   

Sunday, January 27, 2013

Hospice "Under-Use"

The end-of-life care system in the United States is not part of the health care system that often gets high reviews.  We have a system of hospice care that can be used to facility what some call a "high quality dying experience," although whether it is important to have a high quality experience of approaching death or to having on until very end and have what some consider to be a "lower quality dying experience" is a matter of opinion.  In other words, what constitutes a high quality dying experience is a matter of opinion.  Not everyone's utility (or their family members' utility) is maximized by the same choices.

Still, many think of hospice care as potentially less expensive, providing better management of the condition approaching death, and providing an opportunity for family of the dying individual to manage their own interests better.  In that case, the quantity of hospice care demanded is often perceived to be "less than what would be expected".  For a long time people have asked why.

Many reasons have been given.  Perhaps people don't understand hospice.  Perhaps patients of their families have a disutility of even dealing with the hospice decision as it means that they are admitting that they are close to death.  Perhaps providers have a disutility of discussing the issue with patients.

Or, perhaps, according to a news piece featured in the Johns Hopkins Bloomberg School of Public Health news feed last week (the news piece can be found here: the reason is an even more basic economic one.

The news piece reports on an article in Health Affairs (a very good health policy journal) that was worked on by a colleague of mine, Colleen Barry.  The work found that many hospice care facilities have rules that limit enrollment opportunities for patients with more complex needs.  This is more likely to be true in facilities that are smaller and for-profit.  Some, one may question the profit and the capacity of some hospice facilities.

This raises the interesting question about whether a focus on demand side policy interventions is misplaced.  In other words, should we be educating families more?  Should we be educating physicians more?  Should we be working on improving the decision making process?  Or should we, instead, be focused on changing incentives for the supply side?

No definitive answer here, but it does give us something to think about.  The news piece says that the Health Affairs article suggests that Medicare could consider raising the per day reimbursement rate for complex patients.  The big question--would that solve the problem or create new ones by making it easier for more people to enroll.  

Thursday, January 24, 2013

Incentivizing Good Behavior Among Hospitals

The Johns Hopkins Bloomberg School of Public Health newsfeed yesterday carried a version of a piece from the Wall Street Journal.  The article, headlined "Return Patients Vex Hospitals" was written by Laura Landro and appeared on page A6.  The article noted that there are new policies being put in place for Medicare to penalize hospitals up to 1% if the rate of readmission for discharged patients is higher than expected.

There are many interesting questions about this from an economic perspective.

First, how is the expected rate defined and could an individuals hospital do anything to "game the system" and push the expected rate upward, making it easier to stay under the rate, without harming patients?

Second, the article comments on how part of the problem may be the transition between the hospital and care in the community.  While there is no reason to suspect that those providing care in the community would want to provide anything other than the highest quality care, does this create a need for the hospital to control care in the community better?  Does this then lead to an incentive to form an integrated network?  And even if the hospital is successful at hospital care will it then be successful at providing care in the community?  Presumably the incentives might line up better if the hospital was responsible for care to the conclusion of the episode of care but would that give the organization too much market power and, again, would the management work well.

Third, the article mentions that thirty day readmission rates may not be the right thing to measure.  What if a hospital were to lower its readmission rate but then have more patients die in the community?  Would an incentive to decrease readmission rates then, perhaps in an unintended way, lead to more deaths.

Public policy to manipulate economic incentives must be ever vigilant for unintended consequences if the target of the policy/incentive change finds a way to respond that was not anticipated.