Methods of effectiveness research: Intro

I'm currently in the last semester of my Master's in Economics degree and am completing an independent study related to types of Economic analysis with respect to health care. Evidence-based medicine (EBM) is an "it" topic when discussing health care reform, or health care policy. While I'm not directly studying EBM this summer-- my topic is related as effectiveness research not only looks at which treatments are effective (like EBM), but also looks into quantifying and valuing the costs associated with the treatments--and attempting to make decisions related to which treatments/ procedures are worth their costs.

I'm looking at four types of analysis-- Cost of Illness studies, Cost-Benefit analysis, Cost-Effectiveness analysis, and Cost-Utility analysis. I'm also examining issues and practices that arise in the practice of these types of analysis--such as sensitivity analysis, discounting, and challenges associated with data collection and data interpretation.

Many of these topics have recently received a lot of attention from medical and academic communities--especially cost-effectiveness research. To understand why this has become so popular you have to understand where this type of research has come from. Cost-Benefit analysis is largely considered to be the gold standard form of analysis as far as Economists and other professionals are concerned. It's a very simple concept-- you subtract the benefits from the costs (in a common unit-- like dollars) and if the result is positive (ie: benefits > costs) then the decision is considered to be favorable. We do this everyday without noticing. For example, do the added calories from an extra Oreo outweigh how happy eating the extra Oreo is going to make you? If so, you will likely choose to put down the Oreo because the net benefit is going to be negative.

As far as health care is concerned, do the benefits of a twice yearly pap smear outweigh the costs associated with going to the OB-GYN for a pap? Maybe--maybe not. It depends on who you are (do you have a history of cervical cancer? how old are you? have you ever had an irregular pap?). What are the costs of going to the OB-GYN for a pap? Obviously the cost you pay to the office is a cost-- but what about what the insurance company pays? What about your lost productivity or wages at work due to the office visit? What about the emotional stress of receiving a pap twice a year? These are the types of problem you run into with Cost-Benefit analysis when you start trying to analyze these types of procedures. It's difficult to quantify costs-- but even more difficult to quantify benefits (ie: how much money is it worth to prevent cervical cancer?) For this reason, there has been a movement towards Cost-effectiveness analysis.

Cost-effectiveness analysis tries to get around the issue of having to put a dollar amount on benefits. Rather, benefits are measured in another form-- such as the number of cases of cervical cancer reduced due to an increase in screenings, or days of regular blood sugar when taking a new Diabetes medicine, or the % reduction in the size of tumor due to a new cancer treatment. Once the benefits are quantified in this form you can take the cost (which, granted, is still hard to quantify completely) and divide it by the measured benefits (ie: the number of cervical cancer cases avoided). This gives you what is known as an ICER-- incremental cost effectiveness ratio. When other ICERs are calculated (using the same units and same measures) they can be compared.

Cost-utility analysis, which is a form of cost-effectivenss analysis-- an is often just called cost-effectiveness analysis, takes the analysis one step further. Cost-utility analysis uses QALYs (quality adjusted life years) to value benefits rather than trying to directly measure an outcome.

Simply put:

As you can see, the numerator stays the same in all forms, but the denominator changes. Often times people prefer to avoid Cost-benefit analysis because of the difficulties associated with valuing benefits. Cost-effectiveness analysis and the ICER are useful, but sometimes ICERs can not be compared if the denominator isn't being measured in the same way (ie: for a Diabetes treatment: one researcher may measure days within a range of healthy blood sugar levels for a non-Diabetic person, while another researcher may measure days within a healthy blood sugar range for people are are healthy, or pre-Diabetic.) Another benefit associated with using QALY's is that QALYs take into account, not only mortality avoided, but also morbidity avoided due to a treatment. Researchers often like QALYs because of their ability to account for the increased quality of life people achieve after going through a treatment--something not accounted for when measuring a specific health outcome like days of healthy blood sugar. QALY's are, of course, not without their own problems as is described in "Cost-Utility anaysis: Use QALY's only with great caution". In this article, McGregor describes how mis-measuring (or not understanding what is being measured ) with regard to QALY's can limit how powerful the explanatory or comparative power of results of a Cost-Utility Analysis are.

So why does any of this even matter? It matters because no one is able to spend infinite dollars on health care. When there are scare resources, decisions have to be made. Using these forms of analysis we can make more informed decisions--which desperately needed by the federal and state governments as more and more of their money is going towards health care expenditures. We need to make purchasing health care more like purchasing a refrigerator. If you were going to purchase a fridge, you would research the models you were interested and determine which bells and whistles you wanted, and then research how much each model cost in one present day currency (as in 2010 dollars). From there you'd pick which one best fit your needs and your budget. We need to approach health care the same way-- determine what the costs are and which benefits we desire-- and then weigh each one in a standardized way to determine which treatments and procedures we want to (and can feasibly) support.

Information Problems

Asymmetric information is a topic discussed in introductory Economics courses that describes a situation where one party has more information than another. For example, if you are selling your car to a passerby on the street, you have more information about the car than the passerby does. You know that you took your car to your trusty mechanic every 4,000 miles for an oil change, and that whenever anything seemed remotely wrong with your car you got it checked out immediately. Perhaps you’re charging $ 8,000 for you 8 year old car. The passerby is thinking “that’s a pretty steep price for this car… it looks good, but how do I know if something isn’t wrong with it… there MUST be a reason (s)he’s getting rid of it”.

Alas—we have an information asymmetry.

There is no way the buyer can be certain of the quality of the car until they purchase it and drive it around for a few months. The seller can attempt to tell the buyer what a good car it is, but how can the buyer be sure they aren’t just getting scammed?

This story is applicable to the heath care industry as well. Let’s say you go to the doctor with a persistent sore throat and the doctor diagnoses you with some condition--something you can't even pronounce with confidence. “Oh no” you think to yourself, “that sounds serious". The doctor goes on to prescribe you an antibiotic and perhaps some other medicines that will make you more comfortable. The doctor is giving you directions on how to take your medications and what the side effects are, etc. You’re starting to feel a little out of the loop and you’re getting blown away by all of the foreign terminology and directions. At the end of the appointment the doctor asks “do you have any questions”? You think to yourself quietly for a moment… you’re not even sure you could come up with an educated question if you had to. Despite feeling overwhelmed you respond, “no”.

You’ve just been a victim of asymmetric information (and if this situation has never happened to you, surely you’ve taken your car to the shop before and have agreed to all sorts of repairs you weren’t really quite sure about). Doctors go through an intense amount of education—both in the books and in practice. We can’t expect for them to convey all of that knowledge to us when we visit their offices, but we really ought to be better consumers. You wouldn’t buy a new dishwasher without doing at least a little bit of research (at least you’ll compare Home Depot and Lowe’s). However, when you go to the doctor, often times you go in with very limited knowledge about prescription drugs or treatments your doctor may prescribe. If your doctor tells you that you need a test that will cost you (or your insurance company) $4000 at the hospital that you usually go to, are you going to call the other local hospitals to see if you can get the same test for less money? Maybe you think that the test will cost the same everywhere—surely these things have to be standardized—right?

Wrong. The variation in cost between hospitals can be astronomical. We’ll wait for another blog post before we discuss this in full—so for right now you’ll just have to believe me when I say that procedures don’t always cost the same at every hospital.

So, what is the problem here? The problem is simple—you don’t understand that product(s) you’re consuming, and you’re likely paying too much. The bigger problem is that, if you have health insurance, you (and your doctor) probably don’t care too much since you won’t be picking up the bill yourself. This results in overtreatment at increased costs. What’s worse is that, if you’re taking unnecessary medications (perhaps you get prescribed an antibiotic for a common cold) you may actually end up hurting yourself in the long run. I hypothesize, that if we were better, more informed, consumers of medicines we would save money, but more importantly we could be healthier. Every day people are harmed by drug interactions that could have been avoided. Doctors are people too—so they will make mistakes. The informed consumer can safeguard themselves from medical mistakes by asking questions and attempting to learn more about the medications and treatments their doctors prescribe.

What can you do? Next time you’re at the doctor try and think of questions to ask—even if they’re simple. Your doctor has a lot of knowledge—you just need to finagle it out of them.

Here are some questions to ask that I found on about.com:

1. What is the name of my medication?
2. What does my medication do?
3. How/ when should I take my medication?
4. How long should I take my medication for? (this is important—sometimes doctors are not inclined to remove you from medications, even when you may no longer need them)
5. What should I do if I feel better and I don’t want to finish my medication?
6. Does this medicine contain anything I’m allergic to—or will it react with any of my other medications and/or supplements or vitamins?
7. What food, drinks, or activities should I avoid when I take this medicine?
8. What are the side effects? Are they common?
9. Is there a generic version of this medicine?
10. Is it safe to use this medicine if I am pregnant or breastfeeding?
11. How soon will this medicine start working?
12. Will any tests be necessary while I’m taking this medication?
13. Most importantly (arguably): What risks are associated with this medicine and do they outweigh the benefits?

Patents

I am no legal expert, but today I thought I would examine another one of the reasons why medical care is so expensive in the U.S.--the practice of "evergreening" patents in prescription drug. This is a tactic pharmaceutical companies use to extend the life of patents of money-making prescription drugs.

A patent on a prescription drug typically lasts 20 years. However, this can be extended for various reasons. Congress allowed for a 6-month extension of patents if the drug was one that needed to be tested in children. While child testing is important since children sometimes react differently to medicines than adults, this is a very costly provision. According to an article by Robert Weissman (called The evergreen patent system: pharmaceutical company tactics to extend patent protections (Patently Abusive)) It cost pharmaceuticals a bit less than $800 million to do these tests in children on selected drugs, but it resulted in an extra $30 billion dollars in sales. (That's right $800 million in costs, $30 billion (with a B) in extra revenue). I think there is something fishy about this situation. A patent lasts for 20 years-- why can't they do research on the affect of the drug on children in those 20 years? Why do they need an extra 6 months? Six months doesn't seem like much time to do substantial research--especially with medical testing where a lot of the testing revolves around affects from drugs on the body over time. It seems to me that enough research could be completed in 20 years, and that extending the amount of time to do research by 2.5% doesn't add much except excessive profits at the expense of those who are ill.

Again from Weissman:

Led by Public Citizen, consumer groups say the pediatric exclusivity has conferred a windfall on the drug companies. Public Citizen estimates AstraZeneca will earn more than $1.4 billion in added revenue for Prilosec, thanks to the pediatric exclusivity provision, with Pfizer also crossing the $1 billion threshold for Lipitor. Drugs such as Prozac, Celebrex, Zoloft, Claritin and Cipro will bring their makers more than $300 million in added revenue, due to the provision.

Another way pharmaceuticals can extend the life of a patent is by tweaking the ingredients in medications, or by claiming that the drug can actually cure or treat other conditions than originally advertised. For example, a depression medicine may be able to be re-billed as treating anxiety which will extend the life of the patent. Sometimes when a patent does expire, pharmaceuticals can again, tweak the ingredients, and then market the drug as a "new and improved" version of the drug. Since consumers don't know whether the new ingredients are really necessary or not, they may choose to take the new medicine (with the new patent), rather than staying on the old drug which has lost its patent protection.

I don't want to totally demonize pharmaceutical companies since they do a lot of good. I also firmly believe that we do need monetary rewards to help pharmaceuticals recoup the costs of research and development. We can't expect people to be extremely inventive if we're not willing to compensate them handsomely. However, the line needs to be drawn somewhere. The cost of prescription drugs is rising much faster than the value of our incomes and health benefits. At some point the cost will be more than we car bear, and we will have to either reform the way prescription drugs are priced, or we will have to do without prescription drugs; and for many people, the latter is not an option.

WSJ: Does a slower economy increase athleticism?

Slow Economy, Faster Marathons?

The Wall Street Journal online edition had an article that presented evidence of laid-off workers running faster road races. Overall, marathon times in the U.S. have gotten faster this year after slumping a bit in 2007 and 2008. The author suggests that this is because, during 2007-2008 people were trying to hold onto their jobs as the economy started to contract--which in turn meant that they sacrificed their early morning run for an even earlier arrival to the office. In 2009 when massive layoffs really started adding up more people had time to train for major athletic events, like marathons.

Is this a spurious relationship or not?

At first the economist in me said, this is probably just a fluke in the data. The evidence the author provides is really just aggregate level stuff that says marathon times have gotten faster this year. Then the author proceeds to say that this is because a lot of those unemployed people are actually competitive marathon runners... seems a little far-fetched. I could believe once people are laid off they take up running as a hobby--hence why marathon registration has increased, but I don't know about their times being significantly faster statistically. Wouldn't once ''regular Joe's'' have to be training a more than 6-9 months to get really, really fast? I would think so.

I can more easily believe the student athlete story that goes like this: Athletes who graduated recently are more likely to try and run competitively this year than in the past because they know that the job market is so unfavorable. The cost-benefit calculation has really changed for new grads. The cost of pursuing a hobby like competitive running is less costly because you're probably not giving up a salary at some high-paying job (however you are giving up health insurance, potentially). The benefit, is that you don't have to go through the pain of searching for a job during a recession, and you get to do what you love--run races (and potentially get paid for it)! I can see how this would skew the data towards faster times, if all of a sudden you have a huge pool of competitive runners being dumped into the marathon pool.

In the article, the author interviews a few recently laid off workers-turned competitive marathon runners. Their anecdotes fit his story about the evidence. I thought to myself, maybe this isn't completely spurious. I even thought about my own situation. This summer I'm taking enough summer classes to keep me busy, so I'm not working a part-time job like I usually would. I've always been a recreational runner-- but my pace has always been very, well, recreational--at around 10 minutes per mile for longer races. This summer I'm training for my third marathon this fall (which, first of all, may only be happening because I have enough time to train for it since I'm not working). I've been running a few fun runs around town--and strangely--my times are substantially lower than they usually are. I'm down to about 9:00 minutes per mile now for longer runs. I'm not sure if this is because I have all of a sudden developed some fast twitching muscle fibers, or if it's due to the fact that I have time to do quality runs. Last summer when I was taking 6 credit hours and working 40 hours a week, my running really suffered and I ended up missing out on the Fall marathons. Even when I would get out for a run in the morning it would be rushed--and if I waited until after work and school I was too exhausted to do anything but jog around the block a few times before crashing into bed. I think a lot of people probably have the same experiences-- it's tough to come home from work and find the energy to go outside for a speed workout on the track. However, when you're not working you have ample time to pursue your hobbies. So maybe the effect of laid-off workers on marathon times is really there-- at least a little bit.

Pay for Perfomance

I’ve had a businessweek article marked to read in my browser for days now, and finally tonight I got around to reading it. It discusses how President Obama spoke to the AMA (American Medical Association) and discussed some of the changes he believes should be made in the current health care system. Many of the changes would result in lower incomes for doctors—so the president was not exactly “preaching to the choir”. Doctors likely see the effects of high medical costs every day, but surely don’t want the cost reductions coming out of their paychecks.

A lot of health insurance policies in the United States are structured in a fee-for-service fashion. This means that as a doctor performs more services he/she receives more money. It’s not hard to see the effect this would have on the number of services provided by doctors. Obama noted that this fee-for-service changed the way medicine has been practiced—and it needs to change.

"It is a model that has taken the pursuit of medicine from a profession—a calling—to a business," said Obama. (taken from businessweek article)

This incentive structure provides another reason health care costs are higher in the United States than in other countries. How much higher--now that is a good question. According to the BusinessWeek article"it is doctor payments that consume one-third of the nation's $2.4 trillion in health-care spending." I think we would need to verify this with more data, but if it's even remotely true, this topic ought to be at the top of the list of necessary health care reforms.

There is no question regarding the power of incentives. Perhaps the solution here is to align incentives with both our financial and health goals, rather than having them compete. For example, what if doctors were paid bonuses when a patient quit smoking—I’m not aware of any policy like this currently in the U.S. Currently doctors get paid to advise people to stop smoking, and they treat the illnesses that occur because of smoking—cancer, emphysema—both very costly (physically, emotionally, and financially) illnesses. If we could persuade doctors to get patients to quit smoking with additional income, rather than just the happiness a doctor gets from getting a patient to quit, I think we would see a decrease in smokers, and eventually a decrease in diseases caused by smoking

Pay for performance. It’s not a new concept—but it is an effective one.

A good talk on Health Economics in the United States

Here is a link to the former CBO Director, Peter Orszag's blog. This post links to an interesting talk on the problems with Health Care costs in the U.S. as well as the slides from the talk. Currently Orszag is the Director for the Office of Management and Budget. He is the youngest member of the Obama cabinet.

This lecture describes some of the health care challenges the U.S. will face in terms of costs. Orszag discusses how medicare and medicaid costs are projected to grow rapidly in the future, and how costs vary geographically.. Run time is a little over 30 minutes.

The Economics of Health Care is one of the areas of Economics that I find most interesting not only because some type of reform is exigent, but also because everyone interfaces with the health care system at some point in their life.

Some may not agree that reform is necessary, and that the system works fine just the way it is. However, I would suggest they do a bit of research before drawing such a conclusion. According to the Economic Report for the President (2009 http://www.gpoaccess.gov/eop/2009/2009_erp.pdf see CH 7), Americans spend about $8000 per person on health care per year. This is expected to grow rapidly as technology advances.

The chart below illustrates approximations of spending as a percent of gross domestic product (GDP) in the U.S., Canada, Japan, and France. Canada, France, and Japan all have universal health insurance—which one would think would be more expensive. As you can see, the U.S. spends the most as a percent of GDP, in 2000 it was around 13%-- now it is closer to 15% I believe.

Some would stop here and say—“okay—yes we spend more than other countries on health care, but we have the best health care in the world, so the extra expense is worth it.” Sadly, as with everything else in life, more is not always better. Our life expectancy is lower than the other countries in the above chart and out infant mortality rate is higher—two leading indicators of aggregate health status. So now it looks like we’re spending more, but not getting more—which, anyone can tell you—that’s not a good thing.

(This thought also assumes that you have health insurance as a means to battle health costs, which is not true for about 16% of Americans. More on this at a later date…) Economists call this phenomena diminishing returns—meaning that with each additional dollar invested in health care, at some point less health is received. It’s really best to illustrate this one with a handy graph. Part of the health care debate is really between the two (poorly drawn) icons below.

The question is: Are we on the star or the circle?

If we’re on the star, this means that as we spend more on health care we get a positive increase in our health status—a good thing! If we’re on the circle it means that as we spend more money on health care we don’t get an additional health from that spending (like trying to ride your bike to the grocery store, except your bike happens to be a stationary bike). Worse, if we’re at a point beyond the circle it means that as we spend more money on health care our health status actually gets worse—yikes.

Obviously we know there are big gains to some basic health care spending—immunizations, annual exams like paps, breast exams, prostate, etc. However, are there the same gains when a patient undergoes every MRI, X-Ray, and other expensive test necessary in the name of “just in case”? This is where people get divided because the answer to that question usually depends on whether or not anything useful results from the battery of tests. It probably isn’t worth it to give someone an MRI for a stress headache, but it is certainly worthwhile to give someone an MRI for a headache that is caused by a brain tumor.

This brings us to our first point of discussion in this series of things that contribute to the high cost of health care in the United States: Defensive Medicine.

Doctors are supposed to only run tests that they believe are likely to produce useful results. However, doctors often practice what is known as defensive medicine to protect themselves from expensive lawsuits. The JAMA noted that 90% of doctors say they have practiced defensive medicine—and who would blame them given the costs of a malpractice case—both financial costs, but also costs to his/her reputation.

Clearly defensive medicine raises medical costs as patients are subjected to more tests than are necessary. This is costly not only due to the money it costs patients, but takes time away from work/family, adds stress to the life of the patient undergoing the test, and adds to the crowding of hospitals with test equipment.

Sometimes the tests themselves can be lead to health problems for patients. More on all of this at KevinMD—a physician’s blog: http://www.kevinmd.com/blog/2007/04/defensive-medicine.html

What are some solutions to this? How do other countries get around the costs of defensive medicine? This is perhaps a topic for a future post, but what are your thoughts?