In the future, [privacy] concerns will become more sensitive as genetic testing becomes more predictively powerful. Yet, at the same time, as that era blossoms, it will bring the risk-pooling benefits of universal health-care to the fore. As long as people do not have to share genetic data with private insurers, as is the case in the United Kingdom until at least 2014, those who anticipate bad health will do well to buy insurance cover. The genetically lucky, meanwhile, might as well save money and rely on the state. This will squeeze private insurers, suggesting that the [UK National Health System's] golden period may be yet to come. [my emphasis]It's not a new argument by any means, but I was a little surprised to hear it in Nature.
Essentially, the argument goes, allowing consumers to engage in genetic testing while simultaneously barring insurance companies from accessing the results will lead to an inequality in the distribution of information: consumers will be able to predict their risk of certain chronic (and expensive) diseases much better than their insurers. As the predictive power of genetic testing increases so will this disparity, until it maxes out at the point where all remaining disease risk is non-genetic in origin.
In countries with a reasonable public health system, such as the UK, Australia or most of Europe, private health insurance is a luxury rather than a necessity - so people with a low genetic risk of chronic illness won't bother joining*. This will then increase the average risk to insurance companies of each new customer, forcing them to raise their premiums, and thus driving the next group of low-disease risk individuals out of the system. Repeat this for a few cycles and the whole system would ultimately collapse, or at least contract, leading to a widespread exodus to the public health system (which the editor somewhat bizarrely describes as a "golden period").
So, will accurate genetic testing, combined with legislation protecting customers from disclosure of genetic information, lead to the downfall of the private health insurance industry? I have too little economic knowledge to even hazard a guess, but this seems to me like something I really should know more about - so I would welcome informed suggestions in the comments.
* Obviously for readers living in countries without effective public health systems (such as Somalia, Tajikistan, and apparently the USA - or so I was told when I suggested that GINA might not be all good) the equation will be different.
Subscribe to Genetic Future.
11 comments:
Even worse,
What about the public system? In the US that system is what will ultimately take the hit....
I imagine a few corporations going out of biz isn't as bad as an entire country going out of biz.
Or those in the national healthcare systems can do what they already have been.....come to the US.
This next 20 years will change the face of insurance, healthcare and medicine.....I am certain of that.
-Steve
www.thegenesherpa.blogspot.com
speaking as an american, one thing that has come to mind are gene-environment interaction effects. imagine for example if there's an allele which is only a risk allele if one engages in a particular dietary regime. would it not be useful for insurance companies to eventually track one's diet through urine tests so as to reward those who carry the allele but take the appropriate mitigating lifestyle steps?
I wondered about the same thing last year.
The answer is yes.
Either insurance companies will not be able to use information derived from genetics, in which case individuals will opt in and out for only the specific care they need, or companies will, and will be able to offer care only to those who don't need it.
Consider long-term-care insurance, where I think the effects of genetic testing may be felt first.
ApoE testing is simple, currently-available, and predictive of the risk of developing sporadic, late-onset Alzheimer's disease (the common form) (link to Science news report.)
About half of dementia is Alzheimer's. Your risk of developing AD is about 10%-15% if you have two copies of the commonest ApoE allele, ApoE E3 (sorry--%s are approximate and I don't have links handy).
If you have one E3 allele and one bad E4 allele, your risk rises to about 30%-35%.
If you have two E4 alleles, as about 2% of the population does, your risk goes up to about 60%.
This was the gene test that James Watson didn't want to know the results to.
Obviously, long-term-care insurers should be very interested indeed in the ApoE status of applicants. Life insurers and I assume LTC insurers take a blood sample to check serum cholesterol. In the pre-GINA era, were any ever tempted to extract some DNA and run a simple SNP assay? I'm not even certain that GINA precludes this action on their part.
Right now, it seems that the LTC insurance business is running in an age of innocence--potential insureds should be determined to find out their ApoE status, and even more determined to prevent insurers from doing so (unless the results are good, of course--but then the urgency of buying the insurance is much diminished).
razib,
Good point re: gene-environment interactions - your recent post on the Icelandic cystatin C mutation provides a good example. amac notes that life insurers already test customers for serum cholesterol (and I note they also routinely test for cotinine to detect closet smokers), so there are precedents for basing premiums on testing results. In parallel, I expect industries would spring up offering supplements to people promising to fool the insurance test (as is currently the case for cotinine testing).
amac - excellent points. I guess the age of innocence is now drawing to a close: one of the major attractions of direct-to-consumer personal genomics seems to be that by leaving doctors out of the loop customers have one fewer avenue by which their results can leak out to insurers. But how much of an impact will this really have on the survival of the health insurance industry? Will the average consumer be able to make sufficiently sophisticated decisions based on genetic information to be able to scam insurers?
Oh, and you're right about GINA not barring life insurance companies from using genetic test results to make coverage decisions - GINA only affects health insurance and employers, as far as I can tell.
amac,
"ApoE testing is simple, currently-available, and predictive of the risk of developing sporadic, late-onset Alzheimer's disease (the common form) (link to Science news report.)"
It is not that reliable unless you have family history. The test is confusing and genetics guys like Rudy Tanzi don't recommend testing for this.
I think it could be facilitated with the help of someone who is trained in this type of testing....But still, I agree with the Gene-gene and gene-environment stuff....that IS the future....not SNP scans.
-Steve
www.thegenesherpa.blogspot.com
Steve Murphy MD wrote --
> [ApoE testing] is not that reliable unless you have family history. The test is confusing...
Could you clarify? Do you mean:
-- it is technically difficult for a CLIA High-Complexity Lab to determine the alleles at the two SNPs that distinguish genotypes of ApoE E2, E3, and E4?
-- genetic, epigenetic, and environmental factors well beyond ApoE genotype make SNP test results uninterpretable to many or most testees?
-- it is difficult for those knowledgable in the field to interpret SNP test results and express the relative risk of developing sporadic AD for the individual being tested?
-- many or most lay people will fail to properly interpret their SNP test results, in the absence of intensive, personalized guidance?
If a link or reference is handy, it'd be appreciated.
Yes, private insurance is going to have a hard time dealing with this, and that's not a bad thing. Might as well bemoan the loss of the carriage-building industry. That's why SNP profiling is such a game-changer. Healthcare reform is coming, too, so to me it seems like there's no better time.
In the end, more people will know more about what it takes to keep them healthy, which is never a bad thing, and less spending on healthcare overall could ease the burden of providing insurance for those who have known higher risk.
One interesting thing to consider is the assumptions founding this hypothetical situation (insurance co.s going out of biz because only the potentially sick people choose to buy coverage):
1)Everyone (or the majority of the market) will soon know exactly what their predictive genetic future is, and
2) All these people use that information to make a very logical choice.
It's pretty hard to believe those assumptions completely. We all know that with the exception of some Mendelian traits, genetics are indicative, not predictive, and as several have rightly pointed out, the gene-gene and gene-environment interactions are not at all well understood for most conditions. Plus the market is a long way from the majority of people having genetic scans done.
Also, humans don't make particularly logical choices all of the time. It seems just as reasonable to assume that once everyone is having genome-wide analyses done, the vast vast majority will realize they are at increased risk for something, and are then likely to be more interested, not less, in additional insurance coverage. I don't think there will actually be that many genetic "winners" in this game. Really, most of us are walking around with potential disease-causing mutations.
As to the GINA act, I have some issues with the potential safety implied with its passage and the way it might create false security in the consumer mind. See the http://blog.wired.com/wiredscience/2008/05/genetic-protect.html
for my more complete opinion. Raising awareness of this is really important to me. A more educated and informed populace is what we need to help navigate the new and often intimidating realm of personal genetic health.
Thanks for posting about this.
-Tera
http://www.genedream.net/wordpress/
In the Wired interview Tera linked (7/9/08 4:39am), she wrote:
"[GINA] does not prevent genetic discrimination against people when they are applying for life insurance, or long term care and disability insurance."
That is certainly relevant to the ApoE/Alzheimer's scenario discussed above.
Post a Comment