Health Politics

After reading about self-experimentation on Seth Roberts's blog, I've been thinking about the connections between self-experimentation and self-efficacy. Self-experimentation presupposes that it's possible to change one's circumstances, and before starting any self-experiment, the "subject" needs to believe that it's theoretically possible for them to change their circumstances and needs to put serious thought into the reasons for their problems and possible solutions, already more than most people do.

Take the perennial problem of the graduate student: procrastination. Nearly all graduate students and many academics procrastinate: e.g., the dead professor whose desk I used while researching quantum computing made small sculptures from paperclips and dental floss. Some view procrastination as an inevitable and discouraging part of academia, but it's also possible to try to find solutions and feel in better control. Viewing procrastination as an area of self-experimentation (as Seth Roberts discusses in his blog), that forces the subject-investigator to formulate potential solutions, try them one at a time, give them a shot at working, and assess them.

It seems to me that any reasonable intervention to improve procrastination in which a protocol is followed would improve procrastination: most of the problem with procrastination is that it's mindless, disordered, and makes people feel helpless. Following any experimental protocol is a step in the right direction, and if people start to believe that the intervention works, their feelings of self-efficacy could snowball as they begin to truly believe that they have control over a part of their life which felt disordered.

As I enter my final semester of graduate school and think about the work habits that I would like to improve, and the interventions that I would attempt, it seems that anything including watching Eretz Nehederet or dancing to a hip-hop song after writing a certain number of pages or taking a 15 minute walk first thing in the morning could work, as long as I put enough effort to increase my sense of self-efficacy.

New York City has recently required restaurants with uniform menus to post calorie content on their menus with a font size equal to the prices. This initiative may not decrease obesity, but if we're able to gather good data, posting calories on menus could help us better understand how people choose food.

Currently, we don't have a good understanding of how people choose what they eat. Observations of people's food choices through nutritional surveys and food diaries tell us only what people will admit to eating. Laboratory experiments tell us how people who volunteer for psychology experiments choose foods in a new environment, but may not generalize to larger populations in real life situations. Non-laboratory experiments with vending machines have found that people will buy more healthy foods when healthy food is "subsidized" and when less healthy food is "taxed", but nutritional information is not immediately available to subjects even in these experiments: the foods which were manipulated were pretty obvious candidates for healthy and unhealthy foods such as carrot sticks and potato chips.

We also don't know how much knowledge about food people have: when someone chooses a high calorie food, we don't know whether they have chosen that food in ignorance of its calorie content or despite its calorie content. Putting calories on the menu in a visible way gives consumers information which is more readily available than on food packages, and reduces the second problem: some people will read the calorie content of their food when making their choices, and the calorie content may influence their choices.

If calorie information becomes widespread, we could even begin to discuss an elasticity of demand according to both the price and calorie content of the food, as well as a willingness to pay for fewer calories. Just thinking about the McDonald's menu, people can minimize the number of calories they eat by choosing either the least expensive (basic hamburger + fries) or the most expensive items on the menu (salads, grilled chicken).

Some have speculated that posting calorie information on the menus won't affect behavior at all because people choosing to eat at places with unhealthy food can't expect lower calories, but that seems naive. After all, even people shopping at expensive stores are somewhat price sensitive, and all retailers go to lengths to make people feel as though they are getting a bargain.

The inclusion of calorie information on menus gives a tremendous opportunity for social scientists, if only we can get sales data suitable for a quasi-experiment (pre-post with control). Any ideas?

Original copy w comments

I've been inspired lately by the high-quality blogs around me: yesterday I read a random selection from James Grimmelmann's six years of blogging, so decided to dust off this device originally for helping me study for generals.

Andrew Gelman wrote this April about his friend Seth Roberts's diet book. The theory is that eating flavorless foods makes your brain less interested in eating, so you're less hungry, as is described with high-tech graphics in this CBC mini-documentary on the diet. Seth is a Berkeley psychologist who does self-experimentation in his spare time: lately, he seems to be researching methods for overcoming procrastination through structured procrastination.

I'm not used to the idea of drinking oil, but I'm about to go traveling for 3 weeks without kitchen access. Oil and water are plentiful and cheap, as long as I bring a tablespoon to measure the oil. In keeping with the self-experimentation ethic, I want to produce some data. I can't travel with a scale, but measuring tapes are light. Look for dimensionless scatterplots at the beginning of February!

P.S. I suspect that there is a strong element of self-selection in this self-experimentation business. The people who are absolutely nauseated at the idea of oil or sugar water, who find that they are actually nauseated by oil or sugar water, will never stay long enough to register even as a failed diet participant.

When the IOM noted that as many as 40,000 people die per year from medical errors, it provoked a firestorm of outrage and provided new impetus for quality of care research and quality improvement initiatives, led by employers and consumers, such as the Leapfrog Group. Some have even said that the health care crisis is not about increased costs, but decreased value: consumers are willing to pay current health care prices, or even more, but are outraged that prices are so high for low-quality care.

The movement for quality improvement (QI) has been misinterpreted as being critical of individual physicians or even all physicians or as highlighting rare isolated mistakes. In fact, QI is aimed at improving the systemic flaws in the health care system, which currently relies overly much on human memory, error correction, and multi-tasking, is fragmented and disjointed, and is not designed to ensure that patients understand their doctors' instructions. Rather than casting aspersions on the skills, compassion, or abilities of physicians, QI tries to direct these towards patient care rather than extraneous issues.

Anecdotal reports don't teach us about prevalence, but they do give examples of the types of medical errors that can occur in hospitals.

At the same time that the public seems to be at least moderately concerned with health care quality, the leading source of information for consumers about their health care is still friends and family, rather than any of the many surveys of quality available, such as the federally sponsored national surveys of health care quality in specific health plans.

The National Committee for Quality Assurance presents a several summary measures, each of which summarizes dozens of measures. They check some of the quality indicators that are not visible to patients, such as compliance with professional care guidelines (e.g. eye and foot exams for diabetics, follow-up visit within a month after mental hospital stay, postpartum visits within 8 wks of delivery, flu shots for the elderly), measures of patients' abilities to understand their doctor's instructions, etc.

For California a ratings site uses the same data, but displays more details about the summary measures.

Like many parts of the pharmaceutical industry (which I will abbreviate as pharma, which is also the name of the pharmaceutical lobby group) many people are at least vaguely uncomfortable with direct to consumer advertising (DTCA), as it highlights the fusion of health care and profit. We all know that Chocolate Frosted Sugar Bombs (R) do not save lives and accept the commercials as pure attempts at consumer manipulation, but we know that pharmaceuticals save and markedly improve lives. Is pharma heroic for taking enormous financial risks to develop drugs even for rare diseases, or are they profitting from the sick and gullible?

This discomfort makes it way into the policy debate about DTCA, but without being articulated well. People opposed to DTCA generally argue that health care costs and spending are increasing fast enough as it is; in fact, after the FDA loosened advertising requirements in 1997 and DTCA began in earnest, by coincidence, health care costs began rising at a faster clip (at about 12% per year, after several years of 6-8%), since this was also when health insurance purchasers began turning away from HMOs to less restrictive plans such as PPOs. Since 1997, consumer advertising spending for drugs has more than doubled, now on the order of $3 billion, as has total spending on drugs, now about $150 billion, and drugs which are advertised most heavily have seen the greatest increases in use. Certainly, not all of these increases can be attributed to DTCA, but even if we attribute only a small fraction to the ads, pharma's advertising dollar has been well-invested.

Some people feel uncomfortable at hearing these figures, perhaps at the great influence advertising has in the societal definition of "health", perhaps at the fact that people are learning health information from a source that stands to profit from convincing people that they are sick, or perhaps because television advertisements shatter that Norman Rockwell doctor image more than an HMO ever could. To some degree, DTCA may only be a reflection of the changing health information marketplace, rather than a cause: consumers cite diverse sources that influence them to talk to their doctor about specific health issues, and these sources are mostly distant from the health care system, such as friends/family (51%), broadcast media (40%), print media (34%), pamphlets in doctor's office (33%), and the Internet (16%). 33% said they were influenced by other doctors, and 17% by their pharmacists.

Attempting to look objectively at the DTCA issue, I can make a few statements.

First, as a society we value advertising when it advocates socially desirable goals, and conceptualize it as "public service" advertising, even if it's paid for by an industry, such as milk advertisements. Prevention of osteoporosis is a socially valuable goal, and all the better if the milk industry rather than a limited government public health budget will pay for advertisements to reach those who might not otherwise know about osteoporosis. People perceive DTCA as helping pharma, but not ordinary people.

Second, as a society, we value health care when it saves lives or improves quality of life. The classic critique of DTCA is that it causes basically healthy people to consume more drugs than they otherwise would: a person, after a bad day, will go to their physician for Prozac.

Pharma and other proponents make the claim that DTCA brings health information to people and brings them to health care where they receive diagnoses for conditions that would otherwise go unnoticed, which is surely a valid point. Any advertising by the companies who make the drugs and equipment for procedures on the lists of quality measures (e.g., HEDIS, AHRQ, JCAHO) would be a public service: for instance, if flu vaccine companies would advertise extensively to the elderly and raise their vaccination rate, they would literally save lives (influenza/pneumonia is consistently one of the top ten causes of death in the US for all age groups). Similarly for any other procedure with universal professional acceptance which is currently being underused.

Given this criterion, advertising for statins (cholesterol lowering drugs of which Lipitor is the most popular) may be socially beneficial; people at risk for heart disease may know their risk status, but not know that statins may help them until they see advertisements. In fact, a recent study showed that even after DTCA increased statin use in a study population by 60%, the new users were judged to need the drugs as much as the pre-DTCA users did. One might think that the type of person who would ask for statins would be a person with marginal need for the drugs, but this study showed that advertising caused people with significant heart disease risks to seek the statins. (R.W. Dubois et al., “Growth in Use of Lipid Lowering Therapies: Are We Targeting the Right Patients?” American Journal of Managed Care 8, no. 10 (2002): 81–86.)

For the drugs advertised with more marginal benefit or advertised with less accurate information, the advantages (if any) are more marginal, both in terms of cost-effectiveness and risk of over-prescribing. The advantages are also extremely difficult to quantify or even describe, especially given the paucity of high-quality data.

First, which drugs are being advertised? While physician detailing covers all drugs, 60% of all DTCA dollars are spent advertising 20 drugs (Rosenthal MB. Berndt ER. Donohue JM. Frank RG. Epstein AM. Promotion of prescription drugs to consumers.NEJM. 346(7):498-505, 2002 Feb 14. ). The classes of drugs with the most advertising investment highlighted by Rosenthal et al were (in order): antihistamines, nasal sprays, cholesterol-lowering drugs, proton pump inhibitors, and antidepressants.

A survey of doctors released today (n=643) on how DTCA has affected their practice reveals a few things that I found surprising. A quarter of visits related to DTCA result in a new diagnosis, of which a substantial fraction (30%) were considered "high priority" (a designation used in MEPS for chronic and high cost conditions such as ischemic heart disease, diabetes, and affective disorders). The survey also shows that the industry is doing well with these advertisements: about 40% of the time that a patient is spurred to see their doctor because of an advertisement they saw, they walk away with a prescription for that drug, although physicians also suggest lifestyle changes (39%), another drug (22%), over the counter drugs (12%), suggest a diagnostic test (9%), or refer to a specialist (6%).

It has also been speculated that DTCA has resulted in more uniformity of drug use than there is in other treatments. Medical treatments have been shown subject to enormous geographic variation which cannot be explained by demography or geographical patterns of disease: on average, there is about a 4-fold difference, with some treatments having a 12-fold difference. Drug prescribing does not have such significant variations with less than a two-fold difference on average, and only a 4-fold difference for controversial treatments. It is speculated that DTCA is responsible for increasing uniformity, although it is still unclear whether this uniformity is appropriate use, or whether we have uniform levels of overuse. There is no causal evidence, but being able to connect two seemingly disparate subjects is something that health policy wonks cry huzzah for.

There is much more information, but not now.

The less publicized part of the new Medicare drug bill is the ability for people to create the Health Savings Accounts that Cato et al have been drooling over for at least the past decade. They seem so sensible --- letting people use their own money to buy day-to-day health care, and let insurance cover only the parts which are true emergencies. Again, it's not so simple. I have a few questions, and again no answers:

1. The true emergencies are covered by catastrophic health insurance, which pays for everything after $N. These insurance policies should, in theory, be inexpensive and have predictable cost, but what is the rate of premium increase for them? I have heard anecdotal accounts of the rate of premium increase being higher than the general health cost inflation of ~12%, which might make HSAs unaffordable in the future.

2. How many catastrophic health insurance policies are true insurance, and how many have caps, such as lifetime maximums of $1 million? Nearly all insurance policies have such caps, but HSAs are claiming to be "true insurance", back end coverage for really bad disasters, rather than front end after a high deductible.

Currently, the fifty states and various territories of the US have different insurance regulations, which is a barrier to the entry of new companies into the states with small insurance markets and irregular regulations, requiring the insurance companies to learn and comply with an entirely new set of rules. In the case of price regulation, the companies may need to conduct entirely new actuarial studies of whether entering the market would be cost effective. The result of this patchwork of regulation is that small insurance markets have few players, which should in theory make health insurance less accessible to individuals and small groups buying insurance on the market.

Three weeks ago, I attended hearings by the House Committee on Financial Services on a proposal which would centralize insurance regulation and lower barriers to entry of new markets. As an aside, the hearings were boring, and filled with unctuous insurance industry lobbyists who paid professional line standers to stand in line for them starting over an hour before the hearings were to begin; worse was the whining from the unfortunates whose line standers didn't arrive in time or didn't arrive period. And No One would share their hand by telling me what they think of the proposal.

The aspect of this proposal which is being discussed is the one above --- making markets more competitive and so hopefully more efficient, which in turn should help the uninsured gain access to insurance --- is an aspect that no one could oppose. A fundamental part of American culture is a hatred of bureaucracy, and the idea of 50 different sets of regulations should make any good American recoil with disgust. If the centralization would be mandatory, effectively repealing state insurance regulations, the question becomes more complicated.

The proposal raises a few questions to me:
1. Do small insurance markets cause uninsurance --- that is, we can't prove causality, but are there more uninsured in small states with unusual regulations than in larger states? This question may not be answerable because small states have different demographic and economic profiles, so we may not be able to compare their rates of uninsured with the size of the insurance market. e.g., if the former states are rural states with few minorities and immigrants, weak unions, and many small companies, and the latter are more urban states with large industry, strong unions, and many minorities and immigrants, there are too many factors correlated with lack of insurance.

2. What, precisely, are these regulations that are barriers to entry?

3. State insurance regulations are generally made for political rather than economic reasons, so some regulations could be well-intentioned, but have negative effects on access for the uninsured. For instance, about five states have regulations that ban catastrophic-only insurance coverage, causing otherwise healthy people to need to choose between being totally uninsured and paying higher premiums for more comprehensive health insurance than they want --- making the market inefficient; on the other hand, such a regulation can be seen as protecting the consumer, who might not realize exactly what their insurance should cover. Can we grossly oversimplify and come up with some assessment of what proportion of state insurance regulations are valuable and which are pernicious?

4. What regulations would be put in place by a federal system, and how would these compare with the current state system? Politically, it sounds like there will be no price controls, for instance: while price controls can be seen as a market inefficiency, they might be protective against a price death spiral.

5. Are any state regulations necessary in order to be responsive to local market conditions? That is, is it possible to create a federal system that manages to apply to all of the insurance markets?

6. What would the effect of this system be on the self-insuring inter-state companies which are currently exempted by ERISA from following state insurance regulations, but tend to voluntarily follow many of them just for good relations with their employees? That is, would this system apply to these companies? Conversely, would these companies change their coverage somehow in reaction?


All of these questions are so general as to be meaningless since they cover such a vast system which, as much as it is studied, probably has more unanswered questions than answered ones. In addition, it's hard to know which of these questions are relevant. e.g., Self-insured companies are currently not involved in these questions of state insurance regulations at all, yet it's hard to imagine that changing the regulation structure wouldn't affect them at all. Nu?

In the continuing decennial cycle, the health care system is now perceived to be in crisis, but when people discuss this crisis they mean different things --- their own insurability (at a reasonable cost) on the individual/small group market, their ability to form a relationship with a doctor who smiles and knows their name, their own rising out-of-pocket costs, health care cost inflation making businesses reluctant to create new jobs, the rate of medical errors, threats to public health such as bioterrorism and previously unknown infectious diseases, inter alia.

Whether or not any of these are actually problems, we are privileged that health policy has again emerged as a key issue at a level of interest perhaps not seen since Wofford's surprise election in 1991. With this interest comes much discussion and coverage. Kaiser does a stupendous job of cataloging and highlighting the daily news in health policy, but I want to talk back to the news and explore the issues at greater depth, leaving the ephemeral issues behind.