2005 April 30 Saturday
Another Reason Genetic Privacy Will Be Impossible To Maintain

While reading a recent article about genetic testing services a thought occurred to me for another reason why genetic privacy will be impossible to protect: What is to stop someone from sending in a sample of DNA from someone else to a DNA testing service while claiming the sample is their own DNA?

In the past I've argued that the eventual development of cheap miniaturized DNA sequencers will spell the death knell for genetic privacy. Miniaturized sequencers will enable individuals to surreptitiously acquire DNA samples from other people and then secretly test the DNA on their own DNA sequencer machine. Given that microfluidic silicon chips will some day be as cheap as microprocessor and memory chips the idea of DNA sequencers operable by complete amateurs seems inevitable. Granted, such devices will not be offered for sale in the next few years. But 30 years from now (and perhaps much sooner) I'm hard put to see why such machines won't be cheap to build and easy to operate.

However, long before personal DNA sequencers hit the market commercial DNA testing labs will be competing to tell us increasing numbers of insights about our minds and bodies that will be gleanable from DNA testing. Once those services start producing useful information with predictive value about health and physical and mental performance what is to stop any person from impersonating another person and submitting another person's DNA as their own?

Consider that identity theft is a rapidly growing problem. Most cases of identity theft are discovered when someone gets a bill for a product or service they didn't buy. But identity theft for the purpose of getting DNA tested would not necessarily eventually result in the victim ever benig notified that someone else temporarily masqueraded as them. Why? Because when the DNA sample is submitted it would not have to be submitted with the real name of the person the DNA sample was stolen from.

I can envision a DNA testing regulatory regime whereby this method of privacy invasion could be made much harder. To catch such deception everyone in a society would have to be required to donate a tissue sample to a central database. Each sample could be converted to a DNA gel pattern that would be of the sort that law enforcement agencies use to check human tissue samples found at crime scenes. All samples submitted to DNA testing services would be compared against the centralized database to verify the identity of the submitters. If a submitted pattern matched an entry in the central database then the identity supplied with the submitted DNA could be verified against the identity associated with that pattern in the central database. A difference in name, age, or other characteristic could raise a red flag.

The hardest part of the authentication of the identity of the DNA sample submitter would be to verify that the person who submits the sample really is the person whose sample it is. One way to solve that problem would be to require DNA samples be taken from a person in a licensed clinic. Though bribery of clinic workers could defeat such a scheme. Another approach would be to store biometric data in the form of images of irises, fingerprints, and/or other biometric information into the centralized database. Then a person submitting a DNA sample at a clinic could have their eyes or fingers scanned by a remote computer to authenticate their identity during the DNA sample submission process.

I still see such an elaborate regulatory system as fairly easy to defeat. Even if civil libertarians didn't block the creation of a centralized DNA pattern and biometric informaton database and even if the regulatory requirements for sample submission were make rigorous enough to prevent DNA sample submitters from masquerading as someone else such a system would still be defeatable. How? Simply take the tissue sample to another country where the regulations are much more lax and submit the sample for testing in the other country.

Crooked employees of DNA testing labs would be another weak link in a regulatory regime aimed at protecting DNA privacy. Test samples could be tested basically off the books. How to prevent that? Require use of DNA testing machines designed to be unable to operate without being currently hooked via some sort of encrypted link to a larger database that would record all DNA that gets tested.

I do not expect to see such a rigorous regulatory system for DNA testing to be enacted in the next 10 years and probably not for much longer. By the time regulation starts to catch up I'm guessing that personal DNA testers based on microfluidics chips will be on the market and the regulatory regime for large DNA testing facilities which I outline above will therefore become irrelevant.

Currently some DNA testing services accept samples via the mail. Well, this makes cheating incredibly easy. On the other hand, the incentive to get another person's DNA tested is still very low and will remain low until a lot more information about a person can be gleaned from testing their DNA.

I'm still betting on the eventual death of DNA privacy. Incentives to learn about the DNA of others will be too great while regulatory regimes will lag technological advances and will be too easy to defeat. Technological advances, corruptible employees of testing services, the ability to fake identities, and the ability to travel to less regulated jurisdictions will combine to make DNA privacy impossible to protect.

Update: Medical Madhouse Madman points out that a person who lets their doctor or a hospital have a copy of their DNA sequence (or even some subset of that sequence) is also at risk of having their DNA information stolen by a clinic worker.

Personal medical information is protected by law today yet, one’s medical records can easily be accessed by someone working in a medical clinic, a hospital, or a nursing home. In order to obtain someone’s private medical information one would simply bribe a clinic worker, same as the solution suggested in the above entry. The question remains, if the information is not used for the purpose of ultimate profit, of what use it would be?

Might I add that if the hospital is part of a chain with a large central database then the number of people who will be able to get access to that information is even larger.

As for the potential uses of that information: Insurance companies looking to more accurately measure risk are one big obvious set of users. Note that not only medical insurance companies will find DNA sequence information useful. I think it exceedlngly likely that specific genotypes will be found to contribute to dangerous driving habits and slower reflexes. Car insurance companies would therefore find DNA sequence information useful for assessing accident risks.

Companies looking at potential hires would like better measures of mental abilities, level of motivation, ability to handle stress, and other qualities that will be at least partially predictable from DNA sequence information. Companies will also seek to hire employees more resistant to infections, less likely to become drug addicts, and less likely to suffer from other medical problems. Again, DNA sequence information will provide information that will allow companies to reduce the risk of hiring people at greater risk of these and other problems.

DNA sequence information will also be useful in personal relationships. A woman looking for Mr. Goodbar will want to check out the DNA of some guy she meets. She'll looking for both desirable genetic qualities for children and for genetic tendencies for preferred types of behavior in her potential spouse. For example, genetic variations influence the tendency to remain sexually faithful or to cheat. A woman who is looking for a monogamous guy is going to want to steer clear of a guy who has genotypes that code for promiscuity.

Similarly, intelligence agencies and militaries are going to want to know about genetic variations that occur more often in people who become traitors. A genotype does not have to cause a given behavior 100% of the time in order to be useful. If only 1% of the carriers of a genotype engage in a behavior while a mere 0.1% of non-carriers do that fact will have enormous value to businesses, intelligence agencies, law enforcement agencies, and individuals.

Once DNA sequencing costs fall by a few more orders of magnitude scientists will discover functional effects of hundreds of thousands of genetic variations by compariing the behavior and medical records of millions of people. DNA tests will become cheap and then not long after DNA tests will gain considerable predictive value. At that point people will start surreptitiously testing each others' DNA.

By Randall Parker    2005 April 30 09:27 AM   Entry Permalink | Comments (13)
2005 April 19 Tuesday
British DNA Crime Database Ups Crime Clearance Rates

In Britain more crimes are being solved using a national database of DNA sequences of criminals and suspected criminals.

One year on from legislation permitting police in England and Wales to collect and retain DNA samples from those arrested, a New Scientist investigation of the effect this is having on policing has revealed new data on the law's consequences.

Launched on 10 April 1995, NDNAD holds DNA profiles from almost 3 million people (see "DNA database: the facts"). "From an investigator's perspective it's a powerful tool," says Paul Stickler of Sussex police. In a typical month, the database churns out hits for 15 murders, 45 rapes and sexual offences and 2500 car, theft and drug crimes. With DNA evidence, the average crime clear-up rate increases from 24 per cent to 43 per cent.

That is a dramatic increase in clearance rates. With a total population of 60 million people in Britain the 3 million DNA database entries make up about 5 percent of that population. Future database growth will no doubt make it an even more effective tool for catching criminals.

DNA samples from suspects from previous investigations are proving fertile ground for discovering perpetrators of other crimes.

Since it began in 2001, the practice of retaining profiles of suspects subsequently acquitted has added 175,000 extra profiles to the database. Of those, more than 7000 have since been connected with crimes, including 68 murders, 38 attempted murders and 116 rapes.

The trend toward keeping biological samples and not just the DNA profiles produced from those samples is driven by expectations of future advances in DNA testing techniques. As DNA testing becomes more powerful the original samples from crime scenes which do not match with samples of known individuals in the database will be reanalysed in order to derive more information about race, ethnicity, eye color, skin color, hair color, height, facial features, and other features. Existing DNA analysis techniques already can provide quite a lot of racial and ethnic information. As larger numbers of DNA locations are deciphered more characteristics will be inferrable from DNA sequences.

Privacy advocates who oppose DNA databases have had little success in slowing their spread. In November 2004 liberal California passed Proposition 69 to collect samples for a DNA database not just for felons but starting in 2009 anyone arrested for a felony crime. The proposition received a 62% Yes vote.

In California Even those arrested for misdemeanors qualify for DNA sample collection if they have a previous felony conviction.

Collection from Convicted/Adjudicated Felony Offenders: Who qualifies for DNA collection following conviction or adjudication?

a) Any person (adult or juvenile) who is newly convicted/adjudicated of a felony offense, or who is newly convicted/adjudicated of a misdemeanor or infraction offense but has a prior felony (California or equivalent out-of-state crime) of record; (b) Any person (adult or juvenile) currently in custody or on probation, parole, or any other supervised release after conviction for any felony offense committed prior to November 3, 2004; (c) Any person (adult or juvenile) currently on probation or any other supervised release for any offense with a prior felony (California or equivalent out-of-state crime) of record.


Collection from Arrestees: Which arrestees are subject to DNA collection?

Until January 1, 2009, only adults arrested on or after November 3, 2004 for murder, voluntary manslaughter, a felony PC 290 sex offense, or an attempt to commit one of those crimes are subject to DNA collection under Proposition 69. Beginning in 2009, adults arrested for any felony offense are subject to DNA collection. Note: Prior felony convictions do not trigger DNA collection from any arrestees. The arrestee provision is not retroactive.

California's DNA database will become much more effective as a gradually growing percentage of all criminals in California are placed in it. The British experience suggests that arrest rates for crimes will go up dramatically as the California database grows.

Expect to see fearful parents putting their babies and young children into DNA databases in order to find their children again should the children be kidnapped and raised as someone else's child. While that sort of kidnapping is a rare event the perception is widespread that the risks are much higher.

Also, we shouldn't be too surprised to see at least some countries start collecting DNA samples from visitors from other countries both for crime investigation and anti-terrorism efforts. The CIA and other intelligence agencies might even surreptitiously collect DNA samples along with fingerprints and other biometric data from suspected terrorists.

Update: I forgot to mention one really obvious way criminal DNA databases will be used once we know a lot more about the human genome: Criminals will be genetically classified by their propensity to commit various types of crimes. Suppose a murder crime scene has no DNA evidence. Well, the local police will be able to look at the DNA sequences of all criminals they have on file and see which ones have genetic sequences that most strongly correlate with the propensity to commit murder. Then likely murderers can be placed at the top of suspect lists and receive greater attention.

The ability to predict a criminal's likely development from, say, a drug dealer or car thief into a murderer or rapist will be considered in sentencing decisions. Perhaps not all jurisidictions will allow it. But I bet most will.

The identification of genes that contribute to criminal behavior will also ignite a debate about reproductive rights. Should couples who are, genetically speaking, "bad to the bone" be allowed to reproduce? I expect that some governments and populations will say "No".

By Randall Parker    2005 April 19 01:03 AM   Entry Permalink | Comments (19)
2004 July 31 Saturday
DNA Research Data Handling Insufficient To Protect Privacy?

Current DNA privacy guidelines may not be sufficient to protect genetic privacy.

STANFORD, Calif. – In their exuberance over cracking the genetic code, scientists have paid too little attention to privacy issues, say researchers at the Stanford University School of Medicine. Their findings, published in the July 9 issue of Science, suggest that traditional means of ensuring confidentiality do not apply to genetic data and that additional safeguards are needed to protect patients from potential abuses.

“I am surprised that no one has looked at this problem before and asked, ‘Can we really release genome-wide information about individuals to the public,’” said Zhen Lin, a genetics graduate student who led the study. “Nobody did a careful calculation to find whether ‘anonymous’ patients could be identified from this data.”

Supposedly stripping out non-DNA information is enough to protect a patient's privacy.

A 1996 federal law that governs medical privacy requires that research data be stripped of identifying information such as names, addresses and even the last three digits of a patient’s ZIP code before it can be shared. But the law is essentially silent on the issue of DNA, and most researchers have interpreted this to mean that sharing sequence data linked to information from a patient’s medical history is safe.

Sift through a big chunk of published DNA sequence from a published research paper and you can dig out enough points to compare to a person's DNA gotten some other way to get a match.

“Traditionally people believe that if there is no identifier attached, then the sample is anonymous,” Lin said. “We found that’s really not true because the DNA code itself is an identifier.” To demonstrate this, the researchers looked at specific sites in DNA that commonly vary from person to person, accounting for many genetic differences. Each person has about 5 million such sites in their DNA. Using a statistical model, the researchers found that matching 100 of these sites would identify an individual to a high degree of certainty.

In theory, if a person collected a small amount of genetic information about a former research subject, he could match it to database material in the future to get personal medical information about the subject.

Very few people are research subjects on projects that publish a lot of DNA sequences. Most research projects that use DNA variations look at only a fairly small number of sections of DNA to find variations in particular genes thought to be involved in some disease or perhaps in producing differences in athletic performance or cognitive function. So this concern about published DNA sequences is not important (at least not yet) for most research trial subjects who get their DNA tested in some manner.

For research efforts that sequence larger chunks of DNA sequences per test subject more safeguards may be needed for handling the data. However, at this point the cost of personal DNA sequencing is so high that few people are getting so much of their total genome sequenced by researchers to allow for identification of each person. It is not enough simply to have knowledge of 100 out of the 5 million sites which vary between people. Those sites have to be scattered across enough different chromosomes to provide coverage of all (or nearly all) of the chromosomes. My guess is that most research projects that are sequencing large sections are doing so on a limited number of chromosomes and so are not providing the kind of data needed to enable unique identification of each research subject.

In the long run I believe genetic privacy is going to become impossible to protect. So I'm pretty lackadaiscal about this whole subject anyhow.

By Randall Parker    2004 July 31 10:35 PM   Entry Permalink | Comments (6)
2003 October 16 Thursday
US Senate DNA Privacy Vote Will Have Large Unintended Consequences

The US Senate boldly moves to try to prevent people from being judged by their DNA.

The Senate has moved unanimously to ensure that breaking down the human genetic code will bring health benefits without exposing people to job and health care discrimination. The Genetic Information Nondiscrimination Act that cleared the Senate Tuesday on a 95-0 vote would bar employers from using people's genetic information or family histories in hiring, firing or assigning workers. Insurance companies could not use genetic records to deny medical coverage or set premiums.

Any effort that seeks to prevent people from taking advantage of useful information will cause harmful consequences. If genetic information suggested that some workers were uniquely susceptible to exposure to some class of compounds which were completely innocuous to most people this proposed law will probably prevent companies from assigning workers based on that information. If particular genetic profiles turn out to make some people more likely to break down under physically dangerous stressful situations then, again, under this proposed law employers be banned from using that information.

One might argue that the legislators want to assure that each person is treated as an individual. But individual genetic profiles will provide a great deal of information about each person that will allow a far more detailed breakdown of just how much the members of any group of workers or job candidates differ from each other. But the problem is that using genetic profiles seems to many people to be too much like predestination and the use of genetic information threatens the belief that free will can allow people to mold and develop themselves in ways that transcend individual genetic inheritances. The ability to use DNA test results to more accurately predict future performance in jobs and school and the odds of getting illnesses threatens people's sense of self-control over their destiny and their belief in limitless personal potential. There is a resistance to the use of technological advances to automatically produce more accurate evaluations of the potential of individual humans.

Supporters claim they are voting for civil rights legislation.

The chairman of the health panel, Senator Judd Gregg, Republican of New Hampshire, described the bill as "civil rights legislation" for "a world where the secrets of human life have been plotted out and sheep have been cloned."

US Senators say they are promoting high moral principles.

"This is a moral responsibility and a practical necessity," said Senate Majority Leader Bill Frist (R-Tenn.). Senate Minority Leader Thomas A. Daschle (D-S.D.) said, "We can't afford to take one step forward in science but two steps backward in civil rights."

But how would this bill work in practice? Some day personal DNA sequencing will become very cheap and most people in developed countries will be able to afford to have their complete genome sequenced. Once they have that information they will be able to find out their risk profile for a large variety of diseases. This will create a real serious problem for medical insurance companies: those people who are most at risk for getting sick will sign up for more medical coverage and those at lower risk for serious illness will sign up for less coverage on average. This will cause the insurance companies to collect less in revenues while paying more out in benefits. They will respond by raising rates on everyone who seeks insurance. This will cause those at lowest risk to further reduce their coverage. If individuals know their risks and the insurance companies do not then people will be able to game the system. The result will be fewer customers for medical insurance and those who do seek it will be at greater risk of getting sick.

By Randall Parker    2003 October 16 01:19 AM   Entry Permalink | Comments (6)
2003 May 14 Wednesday
Will The Death Of Genetic Privacy Hasten The End Of Freedom?

Writing in The New Atlantis Christine Rosen has written a very useful survey of some of the current and potential future uses of individual human genetic information entitled Liberty, Privacy, and DNA Databases.

And yet, step-by-step and often for defensible reasons, we are paving the way for the universal, compulsory, DNA sampling of citizens. These are not simply the musings of science fiction; they are the logical conclusion of the technological infrastructure of DNA identification—such as Britain’s Biobank—that we are eagerly building. In the beginning, the reasons for such databases will be familiar, modern, liberal, and compelling: to cure disease, to catch criminals, to ensure that children have a healthy beginning to their lives. But the end in sight is a drastically different society and way of life. We may come to know too much about ourselves to truly live in freedom; and our public and private institutions may know so much about us that equal treatment and personal liberty may become impossible.

My biggest problem with her essay comes at the end (partially excerpted above) where she argues that the loss of genetic privacy will contribute to a loss of liberty. Certainly the more we know about each other's abilities, weaknesses, and predispositions the more we will judge each other differently. But we already do that using countless other means of knowing about each other. Harvard graduates are routinely assumed to be much brighter and more together than high school drop-outs. Someone driving an expensive car will be treated with more respect by police when pulled over than someone driving a junker. Good-looking people are treated better than less attractive people. People routinely hire other people based on advice they hear thru grapevines of social and business connections. Countless other examples can be cited.

Christine Rosen does not flesh our the motivations for her conclusion in sufficient detail for the reader to know why she views the loss of genetic privacy as a threat to liberty. Possibly she'll do so in a future essay (hint, hint). However, her mention of equal treatment in conjunction with personal liberty provides a clue. Is meaning here equal treatment before the law? Or is it equal treatment in the workplace? Is being treated equally the same as being treated fairly? Do people have to know less about us in order to treat us fairly?

Let us look at some of the ways that genetic information can now or will in the future be able to be used:

  • To know that a person is who they say they are.
  • To know who is a person's genetic parent or child.
  • To identify who was in a particular place carrying out a particular criminal act.
  • To idenfity who was in a partcular place carrying out some non-crimnal act that is of interest (e.g. who left dirty dishes in the sink).
  • To more accurately predict a person's risk for various diseases.
  • To more accurately predict a person's intelligence.
  • To more accurately predict a person's probability of engaging in various criminal behaviors.
  • To more accurately predict a person's desires.
  • To more accurately predict a person's level of motivation or some personality characteristic.

How helpful will genetic sequence information be for the purpose of prediction of characteristics or behaviors of a person will depend on just how big an influence genetics has on various aspects of personality and intelligence. Certainly physical environmental factors, chance accidents during embryonic development, and experiences all have impacts that limit the value genetic sequence information by itself. In order for genetic sequence information to have a significant impact on how people treat each other the sequence information has to have some predictive value or it will have more in common with horoscope information as a means to judge people.

Suppose genetic information does have some predictive value for each of the above listed purposes. Is it wrong for people to use it to help make guesses about the potentials and likely behaviors of others? If so, why? As a practical matter we need to judge each other and do so daily. If we can do it more accurately will that make us treat each other more or less fairly overall? Keep in mind that our judgements are frequently wrong and sometimes unfair. But they are also frequently necessary.

Put yourself in the position of a job interviewer. You have one position and twenty applicants. Suppose the availability of DNA sequencing information would cause you to choose a different person to hire. Is it wrong to use that information? After all, the person who you would not hire if you did not have the DNA sequencing information would not have been hired because of some other reasons that the DNA sequencing information effectively weighs against. Is it fair for you to have access to that other information but not the DNA sequencing information? Judgements are always less than perfectly accurate. DNA sequencing information may make judgements more accurate on average. Is this a bad thing? If so, why?

Must the death of privacy translate into the death of freedom? Science fiction writer David Brin argues in his non-fiction book The Transparent Society: Will Technology Force Us to Choose Between Privacy and Freedom? that we really don't have a choice about the future loss of privacy. Technological advances will make it too easy for individuals and governments to do various forms of survelliance. But if we allow not just governments but also individuals to use the survellance technologies then we do not have to wind up living under an authoritarian or totalitarian state.

There are technological trends in biotechnology that will make personal DNA privacy impossible to maintain as well. The development of nanopore DNA sequencers combined with microfluidics will eventually allow the construction of very cheap, small, portable, and easy to operate DNA sequencing devices. Such devices will be impossible to effectively ban. They will be able to be made to look like other devices and will be easily smuggled. Also, surreptitious acquisition of cell samples will be too easy to be prevented. For example, a woman will be able to get a cell sample from a date by just giving him a full mouth kiss at the end of the evening and then spitting her mouth's saliva into a cup when she closes her front door. Or she can do it by running her hand thru the guy's hair to get some loose hairs that have a bit of cells at their ends. Or if a person has dandruff on their jacket and leaves the jacket on their chair to temporarily leave the room of a business meeting then someone else will be able to quickly get a bit of the dandruff off the jacket.

I'd really like to see more specific explanations of how increased information availability will make a society less just and less free overall. We face a future where the quantity of available information of all kinds looks set to increase enormously. If there will be harmful effects then the mechanisms by which these harms will be done ought to be articulated in greater detail.

By Randall Parker    2003 May 14 06:42 PM   Entry Permalink | Comments (1)
2002 October 07 Monday
Genetic Privacy And Medical Insurance

Scientists are gradually identifying the many genetic variations that affect health risks. As the number of known variations increases and as the cost of genetic testing falls an increasing number of people will decide to undergo genetic testing to learn more about their health risks. It seems likely that genetic variations will be found that influence the risks for almost all diseases.

When individuals can learn much more about their personal genetic risks for a variety of illnesses they will in many cases respond by making different choices for medical insurance. Many of those faced with greater genetic risks will respond by trying to buy more medical insurance and those faced with fewer risks tend to buy less medical insurance. This will cause a large change in the medical insurance industry regardless of what regulatory changes are adopted in response. These changes will happen in any country that has private medical insurance. While the life insurance industry will also face changes the scope of this essay will restricted to medical insurance.

Logically speaking there are 4 main scenarios for knowledge about genetically derived health risks as they affect medical insurance:

• First, the individual and the insurance company both know very little about the individual's DNA sequence. This describes most of history up until this point.

• Second, only the insurance company knows the health risk information from an individual's DNA.

• Third, only individuals know their own genetic health risks.

• Fourth, both the individual and the insurance company know the individual's genetic health risks.

The first possibility describes where things have stood for most of the history of medical insurance up till now. However, since there are already a number of testable genetic health risk factors this has already begun to change in the direction of the other 3 choices.

The advantage of the first possibility is it literally makes medical insurance possible. The medical insurance industry and applicants have nearly equal ignorance of the health risks of medical insurance applicants. Everyone has a motive to buy medical insurance because no one knows whether they might fall ill. This provides the revenue to pay for those who do develop an illness.

The second logical possibility is unlikely to happen much in practice. Once personal DNA sequencing becomes affordable its likely that individuals will choose to have themselves tested. However, it is conceivable that some people won't want to know what their DNA sequence is or what risks they face from their DNA sequence. At the same time an insurance company, if allowed by regulatory agencies, might in the future insist on a tissue sample for DNA testing as part of an application for medical insurance. That way the medical insurance company can create a more accurate picture of the applicant's health risks. This possibility puts an individual who has an excellent low risk genetic profile at a disadvantage vis a vis an insurance company because the company can charge him a premium that is average for all people even though he's at below average risk and he will not know that in theory he ought to qualify for a lower rate.

The eventual debate will be between the third and fourth possibilities. Should insurance companies be allowed to know the genetically caused health risks of insurance applicants? That question looks set to become the cause of a large political battle in many nations. As genetic testing becomes cheaper and as it advances to cover a larger number of genetic locations most people will have their DNA tested in order to learn about their genetically caused health risks. So the most important question in the debate about medical insurance is going to boil down to the question of whether medical insurance providers will have the right to find out the medically important genetic variations of each applicant as part of the application for medical insurance. The insurance companies will want this information to decide whether to grant coverage and if so what premiums to charge and terms to offer.

When Genetic Testing Becomes Cheap Can The Insurance System Survive?

What happens when many genetic risk factors become known and genetic testing becomes cheap and widely available? This will change the decisions that individuals make about their medical insurance. The insurance industry will be greatly affected by this change regardless of whether it is allowed access to the genetic test results of individual medical insurance applicants. Lets look at the two mostly likely scenarios.

If only insurance buyers know their genetic risks the rates will rise for all.

If a government passes legislation that denies insurance companies access to the results of genetic disease susceptibility tests then individuals will still get tested and find out their genetic risks. People at greater risk will tend to buy more insurance and people at lesser risk will tend to buy less insurance. So the people who will cost the insurance companies to the most will rush to get lots of coverage and therefore will cost the medical insurance companies more than they do now. At the same time the people who face the least risk will spend less on medical insurance. So medical insurance companies will get less money from a group that currently pays in more than they draw out in benefits. The result is that with rates kept the same the revenue for the insurance companies will drop while their costs will rise.

Employer provided health plans will suffer similar problems. Someone who knows they have a greater genetic health risk will choose jobs that have fancier medical insurance plans. People at low risk will be more willing to be self-employed and buy their own medical insurance with more limited plans with higher deductibles and lower premiums. Such low risk people will also be more willing to take jobs that come with less or no medical insurance.

The general trend therefore will be that high risk individuals will tend to buy more insurance while the low risk individuals will tend to buy less. If current premium levels were kept the same then the total amount of revenue flowing to the insurance companies would decrease while their outlays would increase (since high risk people would tend to have more insurance than is they do now on average). The insurance companies would have to raise rates across the board in order to survive. Under this scenario some customers will be priced out of the market.

If both sides know the genetic risks the rates will rise for some and fall for others.

If the insurance companies are allowed to know genetic profile of each insurance applicant and are allowed to set rates based on genetic risk profiles then people with greater risk of illness (especially for illnesses that are expensive to treat) will have to pay higher medical insurance premiums. Under this scenario the lower risk people will pay less for the same amount of coverage than they would under the scenario where the medical insurance companies are allowed to know the genetic risks of applicants. More low risk people would buy coverage under this scenario because their premiums will be more affordable.

Of course, the rates for really high risk people will be too high for many of them to afford. In some cases the insurance companies will decide they will not even offer coverage.

How much will insurance rates change?

An insurance system only works because people don't know how much risk they face. If risks can be accurately calculated for groups but not for individuals then insurance works. Lots of people pay in since all payers think they are at risk. To the extent that any payers can learn they are at less risk they become less willing to pay in. If they can know that they are at no risk at all for a particular threat they would have no need to buy any insurance to deal with that threat.

Just how much insurance rates (and even the availability of insurance) will change depends on how much health risks can be calculated from genetic information. It also depends on how much the knowledge of the health risks can be used to reduce those risks.

Conditional Medical Insurance

Just because some people will turn out to have high risks for certain illnesses will not necessarily be a reason for an insurance company to deny them coverage. In some cases there will be ways to manage the increased risks. For high risk depending on the nature of the risk there are alternatives that insurance companies may offer:

  • Some high risk people will be offered accident coverage only.
  • Other high risk people will be offered medical insurance that has exemptions where their highest risk illnesses are not covered. This will be similar to pre-existing illness exemptions on some policies available today.
  • Required periodic testing for specific disorders as a condition for issuance of a policy for high risk individuals for disorders. There may be an increased premium to cover the testing and to treat the disease should it arise.
  • Required preventative treatments as a condition for issuance and continuation of a policy. It is highly likely that, for instance, a variety of compounds will be found that reduce the risk of some types of cancers (eg by up-regulating liver detoxification enzymes) and some of those compounds may work especially well in people who are at heightened risk for the disorders. Gene therapies will be developed that can reduce risks as well. So someone with, for instance, a higher risk for colon cancer might be required to undergo colon gene therapy to make the colon cells less likely to become cancerous.
  • Required dietary and other lifestyle changes. The problem with this type of requirement is that it can be hard to enforce. So unless an insurance company has a way to monitor compliance this approach is not likely to be used. Still, in some cases blood tests will be capable of detecting noncompliance with medical insurance terms. One might be required to get periodic blood or other testing to detect whether a particular type of food has been consumed.
Anyone have a crystal ball?

It seems difficult to predict exactly how medical insurance will change. It depends on how much genetic variations contribute to health risks, how quickly treatments come along that can cancel out the harmful effects of high risk variations, how expensive those treatments will be, how rapidly the genetic tests become available, and how various factions mobilize to fight over this issue in each political jurisdiction.

There is a bright side: knowledge of how genetic variations contribute to disease incidence will greatly accelerate the rate of advance of biomedical science and technology. Much more effective treatments will be developed for all illnesses. Gene therapy will eventually allow parents to change their progeny's genes so that future generations will have fewer genetically caused health risks.

By Randall Parker    2002 October 07 10:57 AM   Entry Permalink | Comments (2)
2002 September 20 Friday
Genetic privacy: can it be protected?

Governments can pass laws to protect genetic privacy. Certainly in the short term those laws will have considerable effect. But in the long run will the enforcement of genetic privacy laws turn out to be an exercise in futility? The answer to that question hinges more on the cost, availability, and the nature of future DNA sequencing devices than it does on laws that governments may enact. So lets examine some likely stages in the advance of DNA sequencing technology and how each stage will most likely impact the ability to protect genetic privacy.

If DNA sequencing per sample is cheap but machines are expensive and complicated.

If DNA sequencing machines are expensive and yet can process high volumes of samples then the cost per sample can still be low. Under that scenario the machines will be owned by a smaller number labs that can justify the expense of owning one because they will handle high volumes. This is a scenario under which government could try to come up with regulations to make it unlikely that someone other than the person getting tested would be the person who submits a sample. Under these circumstances regulations would probably manage to make the vast majority of all DNA samples legitimately submitted.

Of course there'd be ways of cheating. If someone has a buddy who works in a lab and who is willing to make some under-the-table money that person might be able to get a sample run thru. Though it might be possible to detect that sort of abuse to the sequencing machine since the machine would be tied into a secure computer that would track every DNA sequence result that gets generated and there'd be an audit trail where someone might detect that more sample runs were done than were requested by specific known customers.

Another way to cheat would be to send a sample to a country that has lax genetic privacy laws (the laws might even exist but just not be enforced well) and have the sample run on a sequencing machine in that country. This requires some travel or shipment of the sample. But the sample would be easy to hide owing to its small size. There could also be secret labs hidden in the countries that have strict genetic privacy laws.

Scenarios involving expensive regulated machines and illicit sequencing.

Marsha Goodbar is seriously looking for Mr. Right. She's had some bad experiences with previous relationships. There's a personality type in guys she's been attracted to in the past where they turn out to be far more narcissistic than they initially seem. In the year 2010 that personality type was identified as far more common in men who have a particular combination of Single Nucleotide Polymorphisms (SNPs) and Simple Tandem Repeat (STRs) sequence patterns. Marsha's girlfriend Julie Bond has a brother James who works as a diplomatic attache in a country that has rather lax genetic privacy laws. Julie goes to visit him periodically. On a previous trip down to see James Julie brought along some personal effects left in her bathroom by Derrick, the last guy Julie was disastrously involved with. James was able to get some good skin samples off Derrick's hairbrush, had them sequenced in a local lab. He was able to show Julie that Derrick had the genetic variations typical of self-absorbed manipulative narcissists. Julie told Marsha about this and offered to take samples from Marsha's romantic prospects to get sequenced if Marsha could get DNA samples from the guys she was currently dating. Marsha had just read a July 2011 Cosmo article about genetically based male personality types to avoid and so she decided to accept Julie's offer. Each time she went for dinner or drinks with a guy Marsha managed to slip the guy's empty drinking glass into a plastic pouch in her purse when he wasn't looking. She also gave each guy a real sucking french kiss at the end of their dates and immediately turned around, went inside alone, and soon as the door shut she rushed into the bathroom to spit into sealable plastic bags and then placed each bag in the freezer. This gave Marsha a couple of decent sources of cells from each guy for Julie to take on her next visit to see James. Marsha's own DNA would also show up in the spit but Julie said that with separate samples from Marsha the machine can figure out which DNA is hers and just tell her the additional sequences.

Johnny Law has a hunch that Ralph Ruffian may be the serial killer he is searching for. Its just a hunch. He sees a way to test that hunch. The same DNA pattern has been found at a few murder scenes. Johnny hasn't been able to get enough evidence to get a court order to compel Ruffian to submit to DNA testing. Johnny thinks he should try to confirm his hunch using a surreptitiously acquired DNA sample - even if the result can't be used as evidence in court he or even allow him to tell fellow law officers. It still seems worth it though. He needs to confirm that he really should focus all his efforts on Ruffian. After all, the murderer will probably murder again and Johnny's own rule-breaking seems justifiable to him if it helps him prevent additional murders. Well, down in the Metropolis DNA Testing Lab Cathy Compliant has a big crush on Johnny and Johnny knows it. So Johnny decides that to save the lives of potential future victims of the serial killer he's going to secretly get a sample of Ruffian's DNA by scraping the dirt and skin off of Ruffian's motorcycle handgrips while Ruffian is in his favorite biker bar. Plus, he's going to take scrapings and hair samples from inside Ruffian's helmet. He's also going to scrape the surface of Ruffian's front door knob when Ruffian is not at home. He's hoping that between all these samples some will be good enough for DNA sequencing. Then when Cathy's alone on the night shift he'll ask her to run the sample with no questions asked and without logging the results. Cathy's an ace with the computer and knows how to jigger the DNA analyser database to delete the log entry for a test run.

If DNA sequencing machines are easy to use, cheap, miniature hand-held.

Lets play a mental exercise: Imagine there is a gadget available for purchase for, say, $3000 that can sequence entire human genomes and fairly quickly. Suppose you could use it to get insight into people you deal with in business or in your personal life. Would you be curious enough to spend the $3000 to get a gadget that would let you find out the genetic personality profile or genetic influences on the health risks of others? Would it be more worth your while if you were dating? Or involved in negotiations in high stakes business deals?

If enough people some day start answering yes to those questions then wide availability of cheap small sequencing machines will basically make it impossible to protect genetic privacy for most people. People who are strongly inclined to respect all laws wouldn't cheat. But if the sequencing machines become small and cheap it will be easy for an individual to get a skin sample or other tissue sample of someone else and then sequence it without anyone else having any reason to suspect that it was done.

Its not just the cost of an individual DNA sequence analysis that matters. If a large mass production DNA sequencing machine could do a complete human DNA sequencing for less than $1 but some small portable DNA sequencer could do a sequencing for an average cost of $20 per sample it would be the latter machine that would constitute the bigger threat to human genetic privacy. Why? Its easier for a single person to act without needing to find one of the fairly small number of people who will have operational access to an expensive machine and convince one of those people cooperate (especially since failed attempts to secure cooperation would be legally risky).

A fairly low cost, easily concealable, and easy-to-use DNA sequencing device would be nearly impossible to regulate. It could be smuggled across borders and hid easily in homes. The ability to sequence someone else's DNA would not require that you cooperate with someone who works in a DNA sequencing lab. Only the person who decides to sequence the DNA of another person would have to know that the sequencing was done. The biggest challenge would be to find a way to surreptitiously acquire the DNA sample. So the advent of miniature DNA sequencing devices will do the most to erode the ability to protect genetic privacy.

Let us suppose then that miniaturized gene sequencers will be available at some point in the future. The only action that a government could take to protect genetic privacy would be to ban their sale and make possession and use a crime. The result of course would be that sequencers would still be available on the black market - albeit at a higher price and with legal risks for anyone who has one and gets found out. Can governments prevent people from getting things that they really want? At best governments can considerably raise the cost of purchase and to make people hesitate for fear of being prosecuted for possession of contraband.

Will democratically elected governments even choose to ban DNA sequencers in the first place? The answer to that question depends in part on whether more people want to be able to sequence others or to protect their own genetic privacy. In the war between the sexes each side will want to be able to sequence the other side while protecting their own sequence information (though even there it is easy to imagine a brother wanting to sequence his sister's boyfriend - each side in the war between the sexes has allies on the other side after all). It is difficult to say at this point how the politics of this will work out in practice. Still, lets explore some scenarios of people who will choose to sequence others regardless of the legality of doing so.

Scenarios involving cheap easy-to-use sequencing machines and illicit sequencing.

Joe Normal has a stepsister who grew up suffering from anxiety problems. Joe's seen how hard this has made her life and he's determined that the mother of his children should have no genetic tendencies toward anxiety. Joe read in a 2016 survey article of genetic influences on personality that way back in 2002 some scientists at the National Institute for Mental Health confirmed that a short variant of gene SLC6A4 (which codes for a serotonin neurotransmitter transporter protein) is far more prevalent in people who are prone to anxiety. Joe's worried that his girlfriend Annie becomes anxious in situations that don't bother Joe at all. Now, maybe that's just because it's natural for a normal sane person to not like going 80 mph on a motorcycle while it's raining. But Joe wants to make sure. Joe decides that he should use the dandruff off of one of Annie's jackets to test her DNA to see which variant of SLC6A4 she has. Joe doesn't want his future kids to get the anxiety causing version of that gene. In Joe's country personal DNA sequencers are not legally available. But Annie has been pleading with Joe to take her on a winter vacation to a tropical country. Joe looks up information about the country that Annie wants to visit and finds it has loose DNA privacy laws and that personal DNA sequencers are legally sold in consumer electronics shops. Joe tells Annie that she talked him into the trip and he buys the airplane tickets. She's so excited.

Susie Single really wants to have children. But she's anxious about being abandoned after the first kid is born. She wants a guy who is unlikely to divorce her. Susie has read about androgen receptor genetic variations that contribute to the likelihood that men will not stick around to raise their kids. In Susie's country personal DNA sequencers are easy to buy on the black market if you know the right people. She's willing to pay triple what they go for in countries where the devices are legal. She's dating a few guys who are all sending signals that they are interested in her. Susie is good at cutting hair and starts offering her boyfriends free haircuts. Every time she uses fresh combs and clean scissors and puts plastic down to catch all the hair. After the hair cut is done she collects up all the hair and takes the scalp flakes and later puts it all into the sequencer. For the most attractive ones who turn down free haircuts she lets them sleep with her and gets samples that way. For guys at the office she scrapes the handles of their coffee cups (some guys being pigs they rarely wash the handles) and the surfaces of their desks.

Any Future For Genetic Privacy?

It is likely we will first reach a stage where sequencing services are cheap but the sequencing devices are expensive. This will create some opportunities for illicit sequencing for a variety of purposes. But at a later stage DNA sequencing devices will likely become cheaper and eventually easy to operate by anyone. In this later stage the prospects of maintaining genetic privacy become highly doubtful.

Science fiction writer David Brin foresees technological advances causing the death of privacy. Genetic privacy appears to be as vulnerable as other forms of privacy to advances in technology. Therefore I conclude that genetic privacy will not be protectable in the long run.

By Randall Parker    2002 September 20 01:09 AM   Entry Permalink | Comments (4)
Site Traffic Info