Mandates don’t stay modest

A tactic used by insurance companies to deny expensive behavioral
therapy to autistic children has been deemed illegal by a Los Angeles

In a preliminary ruling, Los Angeles County Superior
Court Judge James C. Chalfant found that Kaiser Permanente's refusal to
pay for a child's autism treatment because the provider was not
licensed by the state runs counter to California's Mental Health Parity
Act. That act requires insurers to cover care for mental and behavioral
problems at the same levels they do for physical illnesses.

Here is the full account.  Three different (but not unrelated) takes on this story are:

1. Whatever you think of occupational licensing, as a matter of social status it seems odd to apply it to dog doctors, or for that matter toilet and sink doctors (i.e., plumbers), but not to those who treat autistic children.

2. These treatments can cost $50,000 a year or more and there is little reliable RCT evidence that they actually work.

3. Yesterday I saw two separate television ads, on two separate channels, campaigning for the Virginia State legislature on the grounds that one's opponent had opposed mandatory insurance coverage for autism treatments.  The ads simply take it for granted that such coverage would be a good thing.  (Rest assured I do not usually watch TV, or its commercials, but the first was in a restaurant at Eden Center and as for the second it was the first day of the NBA season.)


Yeah, the political process is subject to this sort of crap, but what's our alternative? Is this not the equilibrium? Even if at this margin we push with all our might to stop this health care bill, these questions won't go away. And people have demonstrated that they are happy to let the government get involved in these questions when it involves something they want.

In other words, "citizen special interests" (as opposed to the typical "corporate" special interests) in combination with everyday voters' gullibility, are driving a lot of the crap here. Put those two together and I'm not sure how we can hope for a different outcome in general.

Of course you could say that about anything. These posts are useful reminders to people of a certain political persuasion about the relative merits of different political stances at this margin.

This is an example of whether something should be covered by private insurance, or by social insurance.

Social insurance is social security disability insurance. The cost is assessed on everyone through taxes. If you spread the costs of treating autism across a large population though social insurance, the cost is infinitesimal.

It also should be considered appropriate as a social insurance matter because untreated autism becomes a social cost later if the person cannot work or needs additional governmental support.

I do not believe in making many things social insurance, and I am strongly in favor of periodically reviewing social security disability status. I am not a bleeding heart. But, if you look at total costs and benefits, we're better making this a matter of social insurance, just as we do with learning disabilities.

We should not be unmindful of cost or the risks of sloppy or self-serving diagnosis. But, treating this a different way makes more sense rationally than treating this as a private insurance matter which assesses costs on a narrower population.

When we are all born, there are some dice that are cast. Your parents didn't know, and you won't know about your prospective child.

Speaking of licensing...

Question: Isn't there little reliable evidence of most new and experimental treatments. Especially those conducted in such small samples? The Howlin study had a sample of only 44 subjects widely varying in condition).

I'm just a simple country economist, but I wouldn't see that sort of sample as scientifically determinative no matter the result.

The upshot of this is neither a call for better statistical collection and analysis nor for more scientific standards about what should/n't be covered. Rather, I look at this and ask how, in a world where the customer doesn't pay, we could ever hope to collect robust data about what works and what doesn't?

That is, I can imagine a libertarian sort of world where most medical expense is paid out of pocket in something like the high deductible/HSA model. Since many treatments for things like autism are already paid out of pocket, it's obvious more new treatment approaches would be attempted, and (if we collected data well), we'd generate a larger sample with which to evaluate outcomes.

On the other hand, when I imagine a socialist sort of world where most medical expense is directly paid by third parties at indirect expense to beneficiaries (through taxes or insurance payments) I don't see how it's even possible to generate the large samples needed to evaluate new treatments since new treatments are often paid out of pocket and generally few can afford this.

1. Agreed, but that's a matter for a legislature, not the court.

2. The article doesn't say whether the insurer argued this point or not. Insurers definitely should have the right to deny coverage for treatments that don't work, but then the lawsuit would have to actually resolve the question of effectiveness. If the insurer doesn't argue that point it would be extremely unusual and possibly illegal for the judge to raise it on his own.

3. If the treatments are effective, then covering them *is* a good thing. Approximately no insurance buyers will understand the limits of coverage at the time of buying (even if you assume they have a choice about which insurance to buy, which most don't, and I don't recall you sticking up for Wyden-Bennett). The case for uniformity of the product is very strong.

Competition is only effective at improving product quality when the differences in the quality of different products are known to the consumer. The Pinto did just fine in competition for years.

Also, people making judgments in a limited amount of time in a field they're not professionals in are very very bad at tail risk (sometimes even professionals screw it up) and almost all health insurance is about tail risks -- a few people needing very large amounts of health care. Allowing insurance companies to cover only the cheap stuff would amount to allowing them to sell something that looks like insurance but isn't.

Magiati et al. (2007) has the largest experimental group (N = 28) among published controlled trials investigating early intensive ABA-based autism interventions. Also, unusually among the few major trials (prospective, controlled, multi-year, N > 10) in this area, it was conducted by individuals who do not have relevant conflicts of interest.

The solitary RCT (Smith et al., 2000, 2001; you have to read the authors' errata) is very small (experimental group N = 15) and reports largely poor results, especially for children whose specific diagnosis is autism (many of the children had a subthreshold PDD-NOS diagnosis).

California's health partity law falsely and expediently redefines autism, a neurodevelopmental disability, as a major mental illness.

"California's health partity law falsely and expediently redefines autism, a neurodevelopmental disability, as a major mental illness."


What is the significance of this in your mind? Not knowing much about California law, I don't see why its important.

I am curious. What happens to today's autistic adults? Are they mostly working and caring for themselves? Who cares for them if not? The only two I have ever met were completely incapable of taking care of themselves (this was in the last 5 years, and they were 19 and about mid 20s old), and they were described to me as mildly autistic by their caretakers (parents in both cases). Does the new therapy change this outcome?

So far, there are no existing data connecting early intensive ABA-based autism interventions to adult outcomes, regardless of the long history of research into these interventions. Claims that these interventions save money in the long run are speculative. Ganz (2006) found evidence for the costs of these interventions but not for their benefits.

According to published controlled trials, there is no evidence that "earlier is better" for ABA-based autism interventions. All findings reported re age (at beginning of treatment) and outcomes show no relation between the two (Magiati et al., 2007; Eldevik et al., 2006; Lovaas, 1987; Eikeseth et al., 2002, 2007).

There are anecdotes from parents claiming great success for a huge range of currently-marketed autism interventions. The autism literature with respect to treatments and interventions features strong placebo effects when there is good experimental design--the kind of design that is so far scarce in the area of ABA-based autism interventions.

As I've written elsewhere, the best adult outcomes reported in the literature, at least so far, belong to autistics who grew up before the era of early autism interventions and who, as children, met the narrowest, strictest autism criteria ever devised (Kanner et al., 1972; Szatmari et al., 1989; Farley et al., 2009).

In response to Joe, autism is not a major mental illness any more than Down syndrome is. Making decisions, including legal decisions, about a group of people based on totally false information is not good.

There are differing schools of thought about autism. Some think we should let them be the way they are. In my experience, my 5 year old autistic daughter, after months of (high quality) ABA, said to me "I used to scream a lot. I'm happy now."

Autism, like other brain-based conditions/syndromes/spectra, is diagnosed descriptively and not by some sort of objective brain scan or lab test. (Mental illness is the same way; hence some of the enormous controversies surrounding the DSM-V.)
This gives a certain leeway to practitioners in their diagnoses - at least at the margins. This means that as soon as you mandate autism coverage, doctors will be under pressure to categorize other developmental disorders or issues as autism to get that coverage...
An interesting commentary on this phenomenon:
This is the same problem, IMHO, with mandated mental health coverage. Psychiatry is essential to quite a few people. But it is also easily overconsumable (particularly with the "lifelong diagnosis" model of many mental illnesses).


I don't think we're quite on the same wavelength here. Obviously, autistics deserve greater ethical consideration than we give tomatoes. My point is that you seem to think the label in a statute book has far reaching negative consequences. To me, this is a label in a statute, it has no effect outside the statute - the question is whether the law treats autistics inappropriately.

What are the consequences you see? We know it means that insurance has to cover treatments - do you think the insurance mandate is sending too many children to quacky/oppressive treatments? Do you think that the treatments are a way of forcing autisics to conform to society, when they shouldn't have to? Are parents reading statutes and being misled into thinking their children are like schizophrenics instead of developmentally disabled? Am I missing something and the statute changes how doctors treat autistics?

@ Rob

Could you kindly provide a sufficient number of studies, say 5 or 10, from the "more than a hundred years of data" that confirm thay psychology and psychiatry (two VERY different fields by the way) cause great harm to patients. Or maybe ask your friends L. Ron or Mr. Cruise for some copies?

To be fair, if these christian scientists are claiming prayer treatment, they’re probably not consuming a lot of other health care, so can’t we just throw them a bone? I mean really, it’s the least we could do since they so lower the burden on our HC system.

I mean, maybe we could grant them a religious exemption from the insurance mandate altogether? Like, we’ll pay for your prayer treatment but you have to sign off on not showing up in our emergency rooms if the lord doesn’t work.

Conversely, we should probably be encouraging Christian scientist missionaries, you know, to bend the curve.

Comments for this post are closed