The Incidental Economist

Zachary Marcum, Mary Ann Sevick, and Steven Handler in JAMA:

Medication nonadherence is widely recognized as a common and costly problem. Approximately 30% to 50% of US adults are not adherent to long-term medications leading to an estimated $100 billion in preventable costs annually. The barriers to medication adherence are similar to other complex health behaviors, such as weight loss, which have multiple contributing factors. Despite the widespread prevalence and cost of medication nonadherence, it is undetected and undertreated in a significant proportion of adults across care settings. According to the World Health Organization, “increasing the effectiveness of adherence interventions may have far greater impact on the health of the population than any improvement in specific medical treatments.” How can adherence be improved? We propose that the first step is to view medication nonadherence as a diagnosable and treatable medical condition.

Whoa! That was unexpected. It ought to spark conversation at the next cocktail party, that is if you happen to attend a cocktail party for doctors and health policy wonks.

I don’t have a lot to add other than some questions: Are behaviors themselves medical conditions? Or are behaviors usually symptoms of the thing we call a condition? For instance, selling all your possessions to get high is an unhealthy behavior, but it isn’t the condition. Drug addiction is. Which do we treat? (That’s rhetorical.) Is medication nonadherence more like drug addiction or more like the thing drug addiction makes you do? If the latter, what’s the real, underlying condition?

Related to all this is, why don’t drug companies do more to promote medication adherence? Don’t they have a business interest in people using more of their products? I think someone once wrote me about some legal obstacle to drug companies promoting adherence in some way, but I don’t recall what the issue was.

Also, isn’t the hospital that is now at the center of an ACO much more interested in medication adherence than it used to be? Pre-ACO, the hospital earned greater revenue if its patients didn’t take their meds and ended up back in the ED or OR. Now, as an ACO, perhaps the hospital has a greater incentive to treat the condition more cheaply. If that’s possible with drugs, adherence should be paramount.

Your thoughts welcome.

@afrakt

Levy and Meltzer’s update of their 2004 book chapter on the effect of health insurance on health is not new. It was published in 2008, but I’ve never blogged on it. It’s a great and concise resource. Here’s an ungated PDF.

Their Table 1, below, rounds up the best evidence available at the time of publication (click to enlarge).

Most, but not all, of these studies find that expansions of health insurance result in health improvements. The fact that only some of these studies find an effect on health illustrates one important limitation of this type of study: The results of natural experiments may be specific to the population studied. As a result, different natural experiments may yield different conclusions. For example, Card et al. (3) show that the transition onto Medicare at age 65 does not reduce mortality at that age, but Currie & Gruber (4, 5) show that Medicaid expansions reduced child and infant mortality. These varying results are not necessarily contradictory, however, because they apply to different populations. But these differences underscore the fact that the question “How does health insurance affect health?” is complicated, and the answer will depend on (among other things) what we mean by health insurance and whose health is being considered.

Does health insurance improve people’s health? You’ll find Levy and Meltzer cited by both those who answer “yes” and “no” to this question. Of course, as they plainly write, the real answer is nuanced. It depends on which type of people you examine and what you mean by “health.” Just mortality and all Medicare beneficiaries, no discernible effect. Other physical health outcomes and/or other (perhaps sicker) populations, the effect is apparent.

Are those benefits worth the expense? That’s a completely different question and a subjective one. Your answer to it cannot change the objective facts about whether insurance improves health.

@afrakt

Rep. Markey (D-MA) has studied the problem carefully (NECC is in his district) and has filed a thoughtful bill in Congress today (one pager here; section summary here; full text here). Simple compounding is left to state regulation; sterile compounding and commercial scale operations get stricter federal oversight. It’s a good bill, but I have some comments:

• The transparency provisions should be stronger: inspection reports should be published, like restaurant inspections.
• The bill requires warnings on the label, but that won’t directly impact patients since they never see the labels for sterile injectable drugs. I suggest making patient disclosure a condition for billing the drug.
• Instead of “one-size-fits-all” regulation, the bill allows states with demonstrated regulatory capacity to share regulatory responsibility.
• The bill also avoids constitutional pitfalls, which doomed the last attempt to regulate this industry.

On the state level, the Massachusetts Joint Committee on Public Health should release their bill soon.  It is expected to also propose an enhanced “intermediate” level of regulation for the higher risk forms of compounding. If the federal bill becomes law, some adjustments will have to be made at the state level, but Massachusetts shouldn’t wait for the feds. (Congress has many issues to face). The most prominent example is the state regulation of out-of-state compounders. This is inefficient, but necessary, at least until we have a real federal law.

Both bills avoided the temptation to over-regulate the traditional practice of compounding.

@koutterson

Note:  I’ve talked with many government officials and others after the NECC crisis about compounding, but always pro bono.

In early April I offered an analogy to illustrate reference pricing. The post was provocatively titled “Copayments are stupid.” Perhaps for that reason it garnered a considerable number of comments, to which I respond in a new post on the AcademyHealth blog.

@afrakt

The use of psychiatric medications with young children has increased greatly in the US over the last 25 years, including 0.3% of US children under 6 (perhaps 100,000 kids) who take stimulants such as Ritalin. These children are given these drugs because they have or are thought to be at risk for attention/deficit hyperactivity disorder (ADHD). They get this diagnosis largely because they are aggressive or unruly, or at least are not being successfully managed by their parents or caregivers.

Is it a good idea to give these drugs to preschoolers? There are reasons to think not. The drugs are expensive. More importantly, we should not expose developing brains and bodies to powerful medications without a very good reason. So it’s important to ask whether these drugs benefit such young children.

In Pediatrics, Alice Charach and her colleagues review current evidence on ways to help preschoolers judged to be at risk for ADHD. The two primary treatments are programs that train parents in how to manage difficult kids and stimulant medications. Charach et al. searched the literature and found 55 empirical studies of ADHD treatments for preschoolers. But after applying rigorous criteria they found only eight ‘good’ quality studies that looked at parent behavior training. These studies consistently showed that parent behavior training leads to a moderate improvement in child behavior for many children.

Charach et al. found only five ‘fair’ and one ‘good’ quality studies of methylphenidate (Ritalin). These studies reported a mix of benefits and harms for children.

All studies noted improved ADHD behaviors (ie, inattention, hyperactivity, impulsivity) on active treatment… Adverse events were more common and of greater intensity at high than low doses. Poor appetite, social withdrawal, lack of alertness, stomach ache, irritability, and rebound were increased on medication relative to placebo.

In the one ‘good’ study of Ritalin, whether children benefited depended on whom you asked.

methylphenidate resulted in a small positive effect for teacher- but not parent-rated ADHD symptoms and social competence, no improvement in parental stress, and moderate worsening of parent-rated child mood. In contrast, clinicians rated children as improved with moderate to large effect size.

So, I don’t think that treating preschoolers with stimulants is a good idea. There is no strong evidence that stimulants help very young kids, whereas there is good evidence that parent behavior training can help.

The Charach literature review is another example of what Aaron calls “comparative effectiveness goodness.” But the not-so-good thing about this example is that after Charach purged out the weak studies, the eight remaining strong studies on parent behavior training included only 424 children. And the single strong study on Ritalin involved only 124 kids! This is a shallow evidence base for treatment decisions that affect tens of thousands of children. This will sound trite, but we need both more research and higher standards for research in this area.

@Bill_Gardner

I’ve written many times on how it’s rational for wealthier countries to spend more on health care than poorer countries. I’ve even graphed GDP vs health care spending to show this relationship. In this week’s NEJM, Victor Fuchs makes a new chart to look at this relationship in the US over time:

His thoughts:

An examination of data from the past 60 years for the economy as a whole and for health care expenditures indicates that there has been a robust relationship between the two. It seems premature to dismiss the sluggish economy as the major explanation for the spending slowdown of recent years. In the line graph, the economy is represented by the gross domestic product (GDP), which is the total value of all goods and services produced in a given year or its equivalent, the total income received by all contributors to production (e.g., labor, management, and capital). The GDP and national health care expenditures are adjusted for population growth and general inflation. Between 1950 and 2011, real GDP per capita grew at an average of 2.0% per year, while real national health care expenditures per capita grew at 4.4% per year. The gap between the two rates of growth — 2.4% per year — resulted in the share of the GDP related to health care spending increasing from 4.4% in 1950 to 17.9% in 2011. Most experts believe that a gap of close to this magnitude over many future years would have catastrophic consequences for the federal government and the U.S. economy.

Overall, there’s a decent relationship between GDP and health care spending. And, up until the 1990′s, it seems that (for the most part), health care spending growth was always a bit higher than GDP growth. Then came the 1990′s.

It was shocking to me how much of an effect that time had on health care spending. Say what you will about the backlash and the way it was handled, but health care spending growth over that period was held way lower than it otherwise might have been. Why? Managed care.

For those of you unaware of what happened then, there was a big push away from open networks and fee-for-service system towards insurance companies signing contracts with hospitals and physicians, with fees and reimbursement negotiated ahead of time. That was when we starting getting patients to pay more if they sought care out-of-network, and also when physicians began to be reviewed for their actions, and potentially sharing in the insurance risk if they spent too much and gains if they held spending down.

If that sounds familiar, it’s because much of it still stuck around, and more of it is incorporated into Obamacare. Look closely at the chart. What’s also interesting is that since the 90′s, in the last decade or so, health care spending growth has been back in relationship with GDP, but below it.

Maybe I’m being too pessimistic about the ability of these types of actions to slow down the growth of spending. It appears we may have learned some lessons from the 1990′s, and kept applying them into the last decade. Maybe reinforcing them in the coming years will continue to keep growth under GDP growth. Time will tell, but it’s possible.

@aaronecarroll

Michael Hiltzik’s article in the LA Times doesn’t pull any punches:

But you’d be hard-pressed to find a campaign against the ACA as narrow-minded and dishonest as the one mounted by medical device manufacturers.

This campaign has been largely a data-free zone:

The industry can’t cite a single objective study that supports its contentions that the tax will suppress innovation in the field and make U.S. manufacturers globally uncompetitive.

Worth reading the entire thing.

@koutterson

In this week’s JAMA Internal Medicine, “Comparative Effectiveness of Intensity-Modulated Radiotherapy and Conventional Conformal Radiotherapy in the Treatment of Prostate Cancer After Radical Prostatectomy“:

Importance  Comparative effectiveness research of prostate cancer therapies is needed because of the development and rapid clinical adoption of newer and costlier treatments without proven clinical benefit. Radiotherapy is indicated after prostatectomy in select patients who have adverse pathologic features and in those with recurrent disease.

Objectives  To examine the patterns of use of intensity-modulated radiotherapy (IMRT), a newer, more expensive technology that may reduce radiation dose to adjacent organs compared with the older conformal radiotherapy (CRT) in the postprostatectomy setting, and to compare disease control and morbidity outcomes of these treatments.

Design and Setting  Data from the Surveillance, Epidemiology, and End Results–Medicare–linked database were used to identify patients with a diagnosis of prostate cancer who had received radiotherapy within 3 years after prostatectomy.

Participants  Patients who received IMRT or CRT.

Main Outcomes and Measures  The outcomes of 457 IMRT and 557 CRT patients who received radiotherapy between 2002 and 2007 were compared using their claims through 2009. We used propensity score methods to balance baseline characteristics and estimate adjusted incidence rate ratios (RRs) and their 95% CIs for measured outcomes.

Some patients benefit from radiation therapy after a prostatectomy. But the original method, conformal radiotherapy (CRT) required a fairly large amount of radiation. Intensity-modulated radiotherapy was supposed to fix this. It cost a lot more, though. Nonetheless, it went from being non-existent in 2000 to more than 82% of radiation therapy in 2009.

Is it better though? Does it reduce morbidity and mortality? Wouldn’t it be nice to know?

That’s what this study did. It used existing data to look at what happened to patients who received each type of therapy for three years after their prostatectomies. What did they find? There were basically no difference in the adjusted number of events between IMRT and CRT for each 100 person years with respect to the diagnosis of gastrointestinal events (9.4 vs. 9.9), urinary incontinence (11.8 vs. 12.0), or erectile dysfunction (11.7 vs. 13.8). There was also no significant difference in treatment for recurrent disease.

So why are we paying so much more for IMRT? Why is it being used in more than 80% of cases?

@aaronecarroll

I’ve written about fluoridation in my books. Putting fluoride in the drinking water has been called one of the ten greatest public health achievements of the 20th century.  Cavities are still the most common chronic disease in kids 6 to 19 years of age. Still, battles rage to reverse this policy.

Sarah Kliff has a nice summary of the history of this fight. Go read.

@aaronecarroll

How would you debate a wing nut? Cass Sunstein has a very encouraging and interesting answer.

For a positive answer, consider an intriguing study by Philip Fernbach, a University of Colorado business school professor, and his colleagues. Their central finding is that if you ask people to explain exactly why they think as they do, they discover how much they don’t know — and they become more humble and therefore more moderate. [...]

Interestingly, Fernbach and his co-authors found no increase in moderation when they asked people not to “describe all the details you know” about the likely effects of the various proposals, but simply to say why they believe what they do. If you ask people to give reasons for their beliefs, they tend to act as their own lawyers or public relations managers, and they don’t move toward greater moderation. The lesson is subtle: What produces an increase in humility, and hence moderation, is a request for an explanation of the causal mechanisms that underlie people’s beliefs.

Interestingly, what we do on this blog — provide evidence — is not what the investigators suggest works. I don’t know if they tested this approach, but it is well-known that evidence is confirming but not convincing. If the Fernbach study is right, what one should be doing is a lot more listening and asking than telling.

Sunstein’s column is worth a full read. Before commenting, at least do that  much. If you disagree, I ask that you explain exactly why and how much you know about this subject. The published study is gated, but what appears to be an ungated working paper version is here (PDF). It’s in my pile.

@afrakt