Doctors are apparently gods, get used to it

By Razib Khan | October 29, 2012 9:24 am

Here’s a caption from a Time article, What Your Doctor Isn’t Telling You About Your DNA:

Nice to know that two physicians in Philadelphia not only have medical degrees, but specialize in mind-reading the parents of this nation! Above the caption is a photo of  the two concerned and worried looking professionals in question. Let me quote the first two paragraphs of the article:

The test results were crystal clear, and still the doctors didn’t know what to do. A sick baby whose genome was analyzed at the Children’s Hospital of Philadelphia turned out to possess a genetic mutation that indicated dementia would likely take root around age 40. But that lab result was completely unrelated to the reason the baby’s DNA was being tested, leaving the doctors to debate: Should they share the bad news?

When it comes to scanning DNA or sequencing the genome — reading the entire genetic code — what to do with unanticipated results is one of the thorniest issues confronting the medical community. Many conflicted discussions followed the dementia discovery at the Children’s Hospital of Philadelphia (CHOP) before a decision was reached: the parents would not be told that this fatal memory-sapping disease likely lurks in their child’s future. Given the hopelessness of the situation, with no treatment and no cure, the doctors said forwarding such information along felt pointless. We came around to the realization that we could not divulge that information,” says Nancy Spinner, who directs the hospital laboratory that tested the infant. “One of the basic principles of medicine is to do no harm.”

The fourth in a five-part series exploring the promise and pitfalls of sequencing children’s genomes

Around the same time, Spinner’s lab also tested another child — an unusually short 2-year-old referred for kidney disease — and discovered the toddler had a gene linked to a rare form of colon cancer. In some cases, polyps arising from this kind of cancer have been known to develop as early as age 7. This time, the decision to inform the parents was easier: We feel good about that one,” says Spinner. “Proper screening can make a huge difference.”

 

Please. The high priests of medicine know better than you what you would like to know about your children! Think about this logically, and apply this rationale to non-genetic disease. Would doctors conclude that perhaps parents shouldn’t know that their child has a terminal disease which would cut short their life until the illness is closer to manifesting visibly, to save them worry? Perhaps doctors already do this? I have no idea now. My trust has just gone down with the statements above.

John Haws has post up titled Lying to patients about genetic tests is wrong. I agree with the sentiment. I understand that physicians are in a no-win situation. There are parents like me who want as much information as possible, and others who wish to not know information which might imply an unpleasant destiny for their beloved offspring. I say might because how exactly do the doctors above know that in 40 years this situation would be hopeless? Would medicine make no progress? More prosaically, if someone’s realistic lifespan is 40 years, they might wish to make appropriate preparations to live life to the fullest. Individuals with cystic fibrosis have had to make these calculations for decades, once medicine was such that it allowed them to attain adulthood, but middle age only rarely.

All this sort of story does is make me be convinced that what we need widespread personalized genomics so that we can analyze our own sequences with open source applications, and cut the physicians and institutional testing laboratories out of the equation. I also believe that this sort of fiat paternalism on the part of the medical community is frankly going to make enemies of exactly the sort of engaged high-information patients who can be their allies in staving off public hysteria about vaccination and the like. On the whole physicians do know better about illness than their patients. But they should leave the fine-grained ethics to others, because in that domain they’re capable of quite gross malpractice from where I stand. Be honest and do your best are coarse dumb rules which will serve us well in the future.

CATEGORIZED UNDER: Genetics, Genomics
MORE ABOUT: Medicine
  • Shayna

    It’s a hard one because like you said some parents really want to know, while others really don’t. I think you have to ask the parents the amount of info they want to know BEFORE you do the test. That is the only ethical solution. It’s hard because short of listing every possible thing you could find there will still be some gray areas, but you could at least say “Do you want to know if your child has a genetic marker for: a curable disease, a now uncurable disease, an increased incidence of a disease in the future, etc.” There has to be some way to eliminate the guesswork of the doctors. I know I sure as hell would want to know, and I don’t want a doctor deciding on my behalf that it would do more harm than good. Then again, what about telling the child? At what age do I as a parent ask my child if they want to know about their genetic results? It is their right after all.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    Then again, what about telling the child? At what age do I as a parent ask my child if they want to know about their genetic results? It is their right after all.

    - in the USA parents have wide discretion on what children should know

    - seems like it would dependent on how smart & knowledgeable your child is. my daughter will know quite a bit more about statistical genetics i’d hazard at 10 than most people will their whole life (i will introduce her to bayes’ rule ASAP)

  • Frank

    If the doctors were conflicted, how will the parents feel when they eventually find out? The ‘do no harm’ statement is clearly a rationalization.

    Be prepared for a new set of charlatans setting up shop for the great hordes of the uninformed. I heard it termed ‘genomic astrology’ and am quite sure someone is already preparing to capitalize.

  • Jay Fox

    The “Do no harm” requirement could be understood to mandate full disclosure, since withholding information could, by it’s omission, be the basis of future harm. Living a predicted short life to it’s fullest is just the tip of the iceberg.

    My experiences with medical professionals suggest that they are no different than the general population in that there is a range of competence within the profession. There are very good ones, and there are others at the opposite end of the spectrum. Interestingly, most consider themselves at the “good to very good” end, although statistically that cannot be true.

    A little humility is in order. Rarely have I heard the phrase “We don’t know . . .” And yet, that is exactly the case when it comes to genetics analysis. We think, but we do not know for sure about most of this stuff. The field is still young and there is much to be discovered. With full disclosure, a patient can search out and follow the path of their own choosing. With proper monitoring and documentation, we could, possibly, develop some information on things that either help or don’t. But if the patient does not know about a possible outcome, it is unlikely that any strategy might be undertaken to mitigate unpleasant outcomes. Or live life to the fullest now.

    Most of the genetic markers I’ve heard about merely suggest an increased chance of a specific outcome. This implies that a certain percentage of holders of these markers will not suffer the outcome described. We don’t know why. Perhaps some of these folks are doing something different. Hiding the genetic status when it is known does not advance potential knowledge to be gained by following those individuals.

    Disclosing genetic test results is no more harmful than a person knowing their own family health history.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    Disclosing genetic test results is no more harmful than a person knowing their own family health history.

    i think this is fundamentally true. that being said, i think a more apt analogy is getting a non-genetic test which suggests biomarkers strongly predictive of future pathology, but that pathology will only manifest after a precise period of delay. the reality is that until we know the details we don’t even know how confident these physician-gods are about their supposition. i assume there was a statistical geneticist on the committee, but who knows, these are much rarer than medical doctors in this country.

    in regards to medical doctors, on average they are far smarter than their typical patient. that’s certainly true, and there’s much a patient gains from consulting physicians. but the final decision does have to rest with the person who is the subject of illness, unless there’s a public health/contagion consideration. additionally, though physicians are very smart, they are nearly as bad as the public at doing quick & dirty probability, and have a very poor grasp of the calculus of genetic prediction, and the high uncertainty of any given value of probability. the reality is that it’s totally irresponsible to hold information back when the world is already filled with uncertainty and low signal.

    i’ve had the ‘live life to the fullest’ argument well before powerful genetic prediction techniques were around. should you know if your illness is terminal? i’ve had to deal with people who say what i think are pretty stupid and vacuous things like ‘you should always live life to the fullest!’ but that’s a choice, and you need to give individuals (or parents as proxies) the choice to choose. this is particular important in the case of children, whose quality of life in their 20s is going to be strongly impacted by whether they pursue higher education or a field (e.g., science, medicine) which will require lots of labor with minimal disposable income or time early on, with the benefits generally back-loaded. if you are going into decline by 40 it would be stupid i think to make such decisions.

    then again, the doctor-god-committee might periodically meet to review the information and release it to the mortals contingent upon whether it is not harmful to them (because the doctor-gods have so much time to keep track of the lives of their patients over several decades, and can make all these meetings to manage the lives of the mortals).

  • Dm

    It took you a few days after the Time dateline to publish this; I knew I will want to comment, but … my fury has subsided a bit over time.

    This unreported mutation which is predicted to cause dementia in midlife … how do the docs know they still there won’t be any intervention or prevention strategy, in the next 40 years? How do we know that the early awareness about the genetic condition won’t prompt this patient to seek the life-saving treatment as soon as the modern medicine comes up with a treatment … and as we are always told, early preventive action may make all the difference?

    That’s what’s boggling my mind, this doctors’ assumption that the medicine will make no progress, even decades later.

    Of course even in the absence of a life-saving therapy, there are still plenty of meaningful life choices to make with this sort of a knowledge. But the ethicists will kill you for taking this “life planning changes” into account when the patient is underage. They will tell you that a plurality of adults from Huntingtons’ Chorea families don’t want to know; and a large fraction of 23andMe’s users never unblank the Alzheimer risk result. The logic is, if there is such a strong chance that the child wouldn’t want to know after one grows up, then it is unethical to “make him or her know now”.

    With the clinical genetic testing in grownups, this specific ethical argument doesn’t hold water, of course. But still, the industry-wide approach is, to report what has been asked and nothing more, and to mask any incidental findings. We’ll see how well it holds in the future. A single well-publicized lawsuit can change it all…

  • AC

    I’m sympathetic to the sentiment, but this ignored the fact that most patients are dumb. No other high-IQ profession has to deal as much with the median American. God-complexes aside, doctors have a better visceral sense of the needs of the median American than you do.

    I predict that even as personal genomics becomes increasingly feasible, its use will be confined to the intellectual 1%. Most smart people have no idea what normal people are like.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    God-complexes aside, doctors have a better visceral sense of the needs of the median American than you do.

    this is an interesting statement to parse. visceral implies to me that they might just amplify the dumbness of their patients! which strikes me as true. and i agree most patients are dumb. e.g., the tendency of doctors to over-prescribe antibiotics against their better judgment to get dumb patients off their back. but the reality is that most of the work of modern medicine isn’t in the skill of physicians, but the medications they prescribe.

    I predict that even as personal genomics becomes increasingly feasible, its use will be confined to the intellectual 1%

    what’s your identity? email me privately (contactgnxp -at- gmail -dot- com) and out yourself. i want to make a substantial monetary bet that’s public with you on the value of this quantity (i’m willing to hide your identity unless you renege on the bet). i assume you’re not bullshitting here, because i expect people who leave comments here to actually mean what they say when they say it precisely.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    #6, i was on diaper duty this weekend. lots of stuff i don’t blog now due to the caprice of cloth & disposables.

  • Neil

    That the doctors are making things up on the hoof is all rather depressing, given that there is quite a bit of research in this area currently – see e.g.:

    http://www.ncbi.nlm.nih.gov/pubmed/22436882
    Wolf, SM et al (2012)
    Managing incidental findings and research results in genomic research involving biobanks and archived data sets.
    Genet Med. 2012 Apr;14(4):361-84. doi: 10.1038/gim.2012.23.

    “We suggest that findings that are analytically valid, reveal an established and substantial risk of a serious health condition, and are clinically actionable should generally be offered to consenting contributors.”

    So – even ignoring genetic enthusiasts – as per @1, those performing the testing should anticipate that they may make a finding that veers towards the clinically actionable, and ask their patients what they’d want to know.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    #10, not sure this is “on the hoof.” but even if it was deliberate and planned this is not something which is currently amenable to technocratic calculation. IOW, the extra specialized knowledge of the doctors in this case doesn’t lend to them genuine understanding of the variety of issues here (e.g., most doctors are fuzzy on statistical genetics for example, so i’m very wary of the confidence of the initial result unless they had one on staff).

  • Dm

    findings that are analytically valid, reveal an established and substantial risk of a serious health condition, and are clinically actionable should generally be offered to consenting contributors

    Thanks #10! With the current near-stranglehold of bioethicists over genetic testing, this nice-sounding verbiage is actually hiding obfuscation on a grand scale.

    Consenting: in clinical genetic testing, the patients have to sign long, narrowly specific consent forms, with complex technical test specifications (what parts of what genes are going to be tested, and what kinds of results are going to be reported)

    established and substantial risk is meant to exclude most variation from reporting, because it can be argued that the proof of risk hasn’t been established thoroughly enough, or the magnitude of risk hasn’t gone up far enough. But truth be said, today’s patients are most universally irked by incidental findings of variants of unknown or weak significance.

    clinically actionable is to exclude variants which do not, at present, have treatment or prevention.

    So Drs. Spinner and Kranz are right on target: their patient didn’t broadly consent, and the mutation isn’t clinically actionable, hence no disclosure.

    BTW their finding might not have a formal analytical validity either. It generally requires CLIA certification and proficiency testing of a very specific process, and the incidental findings may or may not come from the part of the workflow which has been certified (it depends on whether their protocols and profficiency tests were geared for a whole exome or for a candidate gene set within the exome)

    BTW also #6 I’m so glad to hear that you’re spending more of your time with the little girl … not just with the Internet ;) … and I hope that the diaper duty isn’t narrowly interpreted as “just the technical things with the diapers”, but also as time for emotional bonding and skill development … hey speaking of the skills, the girls can, in most cases, get potty trained before 1 ;)

  • Stephen

    @10 – More about “incidental” findings in genetic tests can be found by Googling the recently coined term “incidentalome.” It seems crucial to distinguish two kinds of incidental finding: (1) Results of obviously true and strongly predictable effect; (2) Results where the clinical effect appears drastic in principle, but it’s a “one-off” finding where the true effect is uncertain or unknown.

  • http://dispatchesfromturtleisland.blogspot.com ohwilleke

    “Given the hopelessness of the situation, with no treatment and no cure, the doctors said forwarding such information along felt pointless.”

    This, and a lot of other situations in which doctors commonly withhold or distort information that they communicate to clients (such as survival prognosis for critically ill patients) are intensely aggravating to people like me who are lawyers in the business of helping families make economic plans for their futures.

    While the doctors may not be able to do anything useful, for lawyers helping families plan their estates and for financial planners, knowledge like this is golden. It is the difference between allowing a family to make adequate financial provision for a child (or having a family set up a trust with terms sensitive to the risks associated with a beneficiary with early onset dementia) and a family that through ignorance sets up their child to be destitute in the years that the child is in need.

    Just because a condition isn’t clinically actionable doesn’t mean that there aren’t non-medical steps that have value that can be taken.

  • Justin Loe

    #14. Good point. Many people neglect to update their wills or insurance policies and it is important for them to know if they are predisposed to a fatal illness in the near future, even if it is not medically treatable. Of course in a young child the issue of estate planning is not relevant.

    To what extent is medical paternalism justifiable in the case of genetics? Not very justifiable, in my opinion. In my view, the more information that one has about oneself (within limits) the better off you are.

  • JeffH

    Here is a somewhat related paper: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2778270/. They found that there was no increase in stress for people who found out that they had the Alzheimer Disease susceptible variant. Obviously these are vastly different scenarios, adult vs. child, ApoE versus early onset mutations. It’s off-putting that the doctors made an unilateral decision and that the consent forms were not forward looking.

    @12 Can you back up this statement: “But truth be said, today’s patients are most universally irked by incidental findings of variants of unknown or weak significance”

  • JohnT

    As a physician, it is often difficult to determine how much information a patient needs to make a reasonable informed decision about their health care. Every patient and situation is different. Genetic testing is not unlike reading an Xray or interpreting lab results. Each test provides varying degrees of information. If a test indicates a significant disease with certainty, the doctor must inform the patient and discuss the treatment. Most test are merely pieces of the puzzle. A good doctor helps determine which pieces are important and uses that information to recommend treatments. If a particular test or information is of unknown or unproven value, what then? How do you recommend treatment with equivocal information? Should I just give all test results to patients and have them “research” their own treatments on the internet? It is not straight forward.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    Should I just give all test results to patients and have them “research” their own treatments on the internet? It is not straight forward.

    there are two issues. first

    1) yes, give them all the information.

    2) the majority of your patients will then ask you & your collaborators (e.g., genetic counselor) for advice in how to interpret the data. a minority will interpret it for themselves, because they know better than you about statistics if not the particulars of the disease, and know how to use the requisite analytic tools (which are in many cases superior to physicians, and can be competitive with genetic counselors from talking to genetic counselors).

    in any case, you said above: “If a test indicates a significant disease with certainty, the doctor must inform the patient and discuss the treatment.” the text above indicates significant disease with certainty. should they then not inform the parents?

    it may be difficult in specific cases how to deal with information (e.g., non-paternity cases). but the broad outlines seem rather straightforward in our information rich but analysis poor world. the value that you as a licensed professional who has a monopoly on particular services thanks to government fiat is not on the information, it is on the analysis. so provide it.

  • Dm

    #16 I know, I know, Razib bans people for admissions like this, but I’m still going to say the truth: I’m speaking fom my professional experience, and I’m not going to look for publications corroborating my knowledge :P

    I underscored that I’m talking about today’s practice. Today’s incidentals in clinical genetics have little to do with whole exomes or large gene sets. They mostly come from tests for specific (familial or founder) mutations, when the patients and their providers sort of expect a very clear answer: positive or negative for a specific mutation.

    When they are reported, in addition, that there was some variant of unclear significance, they are, like, why did you even look there? Have you told us what are you going to do? Sure. But nobody wants to RTFM ahead of time. The patterns repeats over and over again. It’s worse than just a mix of surprise and indignation, because there is always a risk that the doctor will misread the suddenly-complicated result as positive when the familial mutation has not been found, or as negative when it has been. One must understand that some doctors don’t read the reports before interpreting them (a few don’t even open the envelops), while others read the stuff without getting it, and mix-up genes and acronyms, and interpret the results in uniquely creative ways. So aiding the docs’ confusion is a really bad idea.

    There are far too few Genetic Counselors out there, and too much low-quality variant research to make DYI interpretation easy. Polyphens of this world often call things damaging when they aren’t; association studies love to report flukes, any flukes; controls of many case-control reports are vastly underpowered; and functional studies following in vitro mutagenesis suffer from additional spurious mutations in the constructs, or from lack of relevant assays. It’s actually complicated, doh. Have you seen your own Exome results already?

  • Dan

    How much expertise was required for the doctors to know that the patient had the genetic marker for the disease? The most important point made is that the patient can get their DNA sequence and run the various algorithms that find those markers. How much medical expertise is really involved? Are we not talking information? The diagnosis of a genetic marker is the output of software, not a biochemical test right?

    Maybe I misunderstand but I think it could come down to having an application that can analyze your DNA (a data file on your computer) and you could select which markers to look for. If you don’t want to know about marker xyz fine, don’t. Where is the Doctor needed again?

    Yes, people will misinterpret the results and often don’t understand x% likelihood by age n. But the process of finding out could be entirely in the hands of the individual. Or am I missing something?

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    I’m speaking fom my professional experience, and I’m not going to look for publications corroborating my knowledge

    the prior here is that if you are outside of your professional expertise it is important to ask these things, because in those cases people rely on their peer groups, and don’t have good understanding of the primary literature. i know who you are, and can vouch that in this domain your own ‘personal communication’ carries weight.

  • JohnT

    It is not that easy. Every single test suggests several different “possible” conclusions. Ever read the fine print on any medicine. If I discuss every single “possible” problem in the far distant future, I would see one patient a day. The physician has to determine which test are relevant and diagnostic. That takes much more than a knowledge of statistics. That requires experience and an understanding of the limitations of medical treatment and patients expectations. A large percentage of patients don’t understand the nuances of diagnosis and treatment. I know it sounds crass, but if you “confuse” them with the science involved with the decision making process, they will go across the street to the “Best doctor in the world” who will tell them what they need to do and leave the science out. A physician needs wide latitude to care for patients. Everyone is different and very rarely is anything cut and dry. If I was seeing you in the office I would hand you the report from the lab and discuss it to the best of my ability. That discussion would probably involve me saying “I have no idea if this is significant and I doubt anyone knows”. Most people WANT the GOD doctor. They want to be told what to do. They do not want statistics and science. That is not good or bad, it is just human nature. Try and write a government regulation that covers that relationship!

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    #22, i’m confused. are you a doctor or a lawyer? you’re just repeating what i basically said. :-)

    1) if they want you to be god, you do the best divine impersonation you can, assuming that that comports with your morals (otherwise, do the best demi-god impersonation!)

    2) if they don’t view you as god (i used to help people in chemistry who became M.D.s, so i have a more prosaic view of doctors, as do many of us with advanced backgrounds in biological sciences), you give them all the information for them to interpret

    the key here is that step one gives the patient the choice to follow the path. medical atheism or medical theism. there shall be no compulsion in faith!

    if you knew what i meant by statistics obviously you wouldn’t respond ‘If I discuss every single “possible” problem in the far distant future.’ no shit. statistics is in some ways just a formalization of normal analytic process. if you see a big odds ratio with a high probability of developing a disease from a predictive algorithm, you report it. if you have 1 out of 8 great-grandparents who died of heart disease vs. 5 out of 8 great-grandparents, you make a judgement about the weight of this information. probably the most legit way to go with stupid people is to discuss the large risk factors there are in the report, if any, and just give them a reference to the huge list of tiny ones and tell them to ignore it.

    in the specific case here they are omitting a big-ass risk factor which has a big-ass effect on someone’s life. i personally think that that’s egregious, and not really a super difficult issue to address morally (legally it is different because of the clusterfuck of consent guidelines).

  • JohnT

    I’m a physician. I would have told the family the DNA findings. If that got me sued, so be it. The overarching issue in this case and all of medicine is the issue of the validity of testing and decision making concerning treatment. No test or treatment is perfect. Most people overestimate the ability of physicians to diagnose and treat disease. We spend a ridiculous amount of money on questionable diagnostic test and treatments. As DNA testing becomes more powerful, the problem will intensify. Who makes the decisions about testing and treatment? Does everyone have a RIGHT to DNA testing and are we prepared to pay for the results. What if there was a treatment for this child that cost a million dollars a year and has a %20 chance of preventing the dementia? Should all children be tested?

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    Should all children be tested?

    they will. or at least ~90% (some people will refuse for various reasons). the price point is going to be too low.

    What if there was a treatment for this child that cost a million dollars a year and has a %20 chance of preventing the dementia?

    that is a separate question. the issue of widespread testing is moot. it will happen, and it will be cheap enough that people will be able to pay out of pocket. there is even a possibility that people could do it at home, without a lab involved (see oxford nanopore’s marketing; though i’m skeptical of the DIY aspect).

    so the issue will be what do we do when people come back with particular results munged through analytic software? i think we’ll have to have a serious discussion about cost vs. benefit, which we’ve avoided in this country, but is already relatively understood and accepted in europe, from what i gather.

    my answer here is simple: the information should be given to the patients, without any qualification. whether treatment will occur is a totally different question. we shouldn’t dodge the second conundrum by choking off the first step. not only that, we can’t. you’re just putting off the inevitable.

    the question of treatment i view as consequentialist. the question of right to one’s own information is deontological.

  • JohnT

    I agree. The “blank check” of American medicine is about to come to an end. Cost vs. benefit is not gonna sit well with many people. It is inevitable.

  • Dm

    #25 the issue of widespread testing is moot. it will happen, and it will be cheap

    Cheap is cheap. The “it” which *will* happen probably won’t make a difference for great many people (for the ones who really need the test results because something is dangerously amiss in their genes). Between structural variation and repeats and pseudogenes and noncoding stuff, these cheapo sequencing results will have too many holes, and too much important variation missed. And since most clinically significant variants are rare and not extremely penetrant, and since most clinically important conditions are common, the public-domain genotype-phenotype data will remain both woefully insufficient, and contaminated by garbage.

    So the era of cheap sequencing may become the era of exacerbated data inequality. More sensitive variant detection will remain costly, and more accurate data interpretation, costlier still. The regular entertainment-value cheap genome crowd will have to stop at the common-SNP level of complexity, just a baby step beyond today’s 23andMe. To get the real deal, it won’t be cheap, or DYI’able.

  • chad

    Imagine being the parents of that first child, reading the article, and realizing that it was YOUR child they were talking about….

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    And since most clinically significant variants are rare and not extremely penetrant, and since most clinically important conditions are common, the public-domain genotype-phenotype data will remain both woefully insufficient, and contaminated by garbage.

    what i’m thinking might be low hanging fruit are are 100X coverage of genomes, and comparisons within the pedigree and some friends. how far into the future do you think this is for a <$1,000 price point? <$100? basically fishing for variants unique or near-unique which might be relevant.

    what date are you imagining? i think 2020 might be a conservative date for the combination of commodity sequencing and computational power and critical mass in the phenotype databases.

  • Dm

    #29, the biggest challenge may be biological rather than technological, basically the mankind isn’t into another bottleneck yet (despite dramatically dropping fertility rate). The populations of the world have been growing exponentially very recently, and for a long time. So recent, and therefore rare or private, loss-of-function variants account for a very large fraction of our overall mutation rate. Longer genes in particular are saturated by recently emerged mutations, and making sense of their effects requires “phenotype databases” the size of nations.

    The second largest issue is with the definition of “sequenced genome”/ “100X coverage”. 100X isn’t 100% – the outcome depends on the uniformity of coverage, on quality of sequences and systematic error rates, on the length of reads, etc. I strongly suspect that the affordable 100X-for-the-masses product which wins the hearts of recreational genome enthusiasts – probably even before 2020 – will deliver a genome full of holes big and small due to regional biases, repeats, and error-prone sequence contexts. It will probably fail whenever pseudogenes are around, and with most if not all copy number / structural variation.

    We can use a parable of youtube vs. big screen industry, twitter vs. essayism … of course the cheaper / lower quality forms are hugely empowering, and their impact is enormous. But will the “youtubes of genomics” doom the “Hollywood studios of genomics”? I really doubt.

    For one thing, it’s about health, and we tend to value good, expensive advice in medicine. You only have to have your genome analyzed once, and it makes fairly good sense to do it well; I can see how the patients, the doctors, and the insurers may agree on this one thing.

    And then, in a positive feedback sort of an outcome, medical-grade providers may continue improving their datasets, while the public-domain data will keep facing not just a mere “lack of critical mass of data”, but also a huge burden of data errors.

  • Neil

    #30.

    For one thing, it’s about health, and we tend to value good, expensive advice in medicine. You only have to have your genome analyzed once, and it makes fairly good sense to do it well; I can see how the patients, the doctors, and the insurers may agree on this one thing.

    And then, in a positive feedback sort of an outcome, medical-grade providers may continue improving their datasets, while the public-domain data will keep facing not just a mere “lack of critical mass of data”, but also a huge burden of data errors.

    And outside the US? Under a “socialized healthcare system” – such as we enjoy in the UK – large-scale sequencing for clinical purposes is being done as research projects. See e.g. http://www.uk10k.org/ and http://www.ddduk.org/

    So – are you going to discount this data because it doesn’t come from a medical-grade provider?

  • Dm

    http://www.uk10k.org/ and http://www.ddduk.org/
    and inside the US, projects such as just recently completed NHLBI GO. Sure, it’s all (baby) steps in a right direction. With a few hundred patients per condition, it can achieve interesting results for common penetrant variants, especially for rare conditions. But it is, at most, a tip of the iceberg. Most disease-causing variants are too rare / too moderaly enriched to be caught this way. Even whole disease-associated genes will most likely fall through the cracks. And conversely, underpowered study sizes tend to produce some false positives.

    Most of the public sequencing projects of this kind are created to continue funding for the large sequencing centers, the vestiges of the Human Genome Project, which are now too big to be abandoned. But there isn’t enough money in the public sector to power the studies better. Of course, perhaps there will be more public sequencing as the costs fall. In any case the data generated by these projects aren’t public domain (the genotypes, sometimes, are; the phenotypes are almost always completely blanked out).

    The Cool Britannians have some nerve BTW, to claim that their measly 6000 exomes, spread between many common conditions, will “elucidate singleton mutations”. Wish them luck!

  • http://whitecoatunderground.com PalMD

    The ethics of the situation pretty much require informing the parents.

    This does not, imo, imply that we are in some way ready for “personal genomic empowerment” any more than we are ready for patients to order their own CT scans. It should, ideally, be collaborative. I’ve had patients ask me for genetic tests I felt were inappropriate (IOW, low PPV, low NPV) but have ordered them with that caveat (a very strong caveat). Others might argue against that and I’d support it. Not all tools are or should be widely available.

  • Neil

    @32 – In any case the data generated by these projects aren’t public domain (the genotypes, sometimes, are; the phenotypes are almost always completely blanked out).

    You’re complaining about being given only half a free dinner. But the government is trying to give you the other half (cf the NIH policy on tax-funded research):

    http://www.rcuk.ac.uk/media/news/2012news/Pages/120716.aspx

    RCUK announces new Open Access policy

    The new policy, which will apply to all qualifying publications being submitted for publication from 1 April 2013, states that peer reviewed research papers which result from research that is wholly or partially funded by the Research Councils [government-funded research]:

    * must be published in journals which are compliant with Research Council policy on Open Access, and;
    * must include details of the funding that supported the research, and a statement on how the underlying research materials such as data, samples or models can be accessed.

    Don’t see the bit that says “except phenotypes”.

  • http://dispatchesfromturtleisland.blogspot.com ohwilleke

    #15 “Of course in a young child the issue of estate planning is not relevant.”

    Not so. It is young children where the issue of estate planning is most relevant. The issue is what parents need to do in their estate plans to appropriately plan for their children. This is not an exact science. A lawyer, like me, sits down with a client and tries to figure out what problems the children are going to have later in life and set up safety nets in the form of guardians, trusts, etc. to provide for each child on an individualized basis later in life.

    In the absence of knowledge, an estate plan is based on some default assumptions about how children usually turn out and what works for average kids. But, the more you know about the problems a child will face, the more you can do to adapt the plan.

    For example, if you know that a child is likely to need government assistance in obtaining necessary medical care, you want to be sure to set up an estate plan that won’t disqualify that child from receiving that kind of care. It could be something as simple as using a 529 plan to save for education expenses for the child, rather than a custodial account in the child’s name which could be disqualifying, or prepaid college tuition units which the child will never be able to use.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    This does not, imo, imply that we are in some way ready for “personal genomic empowerment” any more than we are ready for patients to order their own CT scans. It should, ideally, be collaborative. I’ve had patients ask me for genetic tests I felt were inappropriate (IOW, low PPV, low NPV) but have ordered them with that caveat (a very strong caveat). Others might argue against that and I’d support it. Not all tools are or should be widely available.

    you’re collapsing different issues. i think that if the first order risk to obtaining information if non-existent, then people should be able to get that information without any approval needed from doctors or professionals. CT scans require some radiation, so a gatekeeper is reasonable (hypochondriacs might get themselves scanned over and over). OTOH, something like sequencing/typing just requires a tissue sample. with the ‘quantified self’ movement this is going to get much more advanced, and i can’t see the medical profession doing anything about it barring banning importation of monitoring devices in an unlikely authoritarian fashion.

    the point at which professional guidance is necessary and perhaps obligatory is when people want to ACT on that information. there are strong public health reasons to not let hypochondriacs dose themselves with antibiotics, or demand any surgery under heaven. but information in and of itself does no bodily harm.

    a confusing point right now is that to a great extent consumers are not paying for their medical procedures. it seems acceptable that if you want insurance or a national health service of some sort to pay for a test you’d have to get that signed off by a doctor. but in the near feature a wide range of tests may be easy enough to use that you can engage in self-administering, and cheap enough that you can pay full freight without making recourse to insurance or nationalized health care.

    if the concern is that a proliferation of information will result in a demand for more medical care, the issue should be addressed at the point where people demand care which they can’t afford to pay for, and which is not necessary. keeping people in the dark from information that may be trivial is probably futile.

  • Scott

    My wife is a doctor and information like this is always shared. This was the mistake of individuals. Do not attribute fault to all doctors.

  • Justin Loe

    35. ohwilleke

    I stand corrected. I was thinking only in terms of a child’s assets not the parents. Clearly, if the parents’ have a shortened life span as revealed by a genetic test, the estate planning must take into account the future well-being of the child.

  • https://twitter.com/CountryKidsDoc Matthew J. Weidman, M.D.

    In recent years we have learned that universal screening is very often a bad idea. Routine PSA testing – more harm than good. Routine mammography – more harm than good. How do we counsel patients when we very often have no clue what these genetic markers may mean for an individual patient? If a patient has a mutation that supposedly puts them at future risk for some life-altering condition, do they then undergo additional testing, with added expense, the potential for further false positives or false negatives, and possibly harm or even death?

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    #39, i addressed this issue earlier. the problem here is that choking off information flows that lead to these adverse consequences is futile and short-sighted. we need to fundamentally address issues of scarcity and the dangers of false positives head on. granted, in the short term it makes the paternalistic project more difficult. easier to hide confusing results than explain why test X as a follow of result Y may cause more harm than good (e.g., the CT fiasco).

  • Dm

    Just a little food for thought BTW, re: incidentals, and dangers of false positives (you beat me to the CT parable ;) but I got examples from the actual genome sequencing world)

    NHLBI GO project just added indel calls to their Exome Variant Server today (kudos!). They called frameshifts here and there in the really important genes which had nothing to do with the conditions they studied. It isn’t without an occasional false positive. For example, about 5% of their subjects got an rs80357868 call (a highly penetrant frameshift in a tumor supressor gene).

    23andMe added indels to their Exome Pilot reports a few days ago, too. They don’t have any incidentals since theirs isn’t a specific disease study. So they actually promptly reported a frameshift in a gene responsible for proper cardiac contraction … to the majority of their study subjects.

  • Jeffie Freedom

    As presumptuous (if not foolish) as it was for those doctors to assume that this dementia will not be cured or treatable in 40 years, it’s probably quite irrelevant since surely everyone’s genome will be available in an open access formate by then with automatic flags for anything interesting.

    It would be unfortunate though, if during the period before that happens a treatment were to be developed and this guy loses some brain function anyway because he hasn’t thought to get tested for something that’s rare.

    A responsible protocol would at least file this information in some system that would automatically bring it to someone’s attention should a treatment become available.

    A straight forward solution to their “dilemma” is simply asking patients or parents ahead of time, whether they want to know about anything extra that happens to get noticed.

  • Dm

    I hope that it is appropriate to cite from upcoming publications in this blog w/o detailed comments (please correct me if I’m wrong). Following are a few relevant exerpts from AMP’s report “Opportunities and Challenges Associated with Clinical Diagnostic Genome Sequencing”

    The personalized nature of testing at the genome level
    includes a subjective assessment of the overall value of
    the generated information. Apart from personal utility,
    which could be defined as the perceived usefulness for
    an individual given his or her interests (whether these are
    medical or not), there is the issue of clinical utility. Clinical
    utility is a measure of the net health benefits, reflected by
    the balance of benefit versus harm.

    the development of evidence-based, validated scientific
    standards to evaluate the clinical utility of genomic
    results in different populations with accurate genotypebased
    risk estimation is a considerable challenge. Furthermore,
    there is a possibility of potential overinterpretation of
    results based on a limited understanding of contextual information
    that could lead to unnecessary medical action
    and cause unwarranted psychological distress. Careful
    selection of patients for genome sequencing and concomitant
    genetic counseling are crucial

    In diagnosing a rare disease through genome sequencing,
    the presence of incidental findings becomes a more significant
    topic. These issues remain under discussion, but
    their solutions will no doubt involve an emphasis on counseling
    and education before testing is performed, informed
    consent with a clear explanation of the current
    limits of testing and interpretation, maintenance of privacy
    and confidentiality, and sensitivity to culture within
    families, their heritage, and their communities.

    there is considerable uncertainty about the regulatory pathway
    for NGS testing and resolution could take years. Thus,
    laboratories will, for the foreseeable future, continue to
    rely on regulations under the Clinical Laboratory Improvement
    Amendment.
    ….
    Notably, although the costs of performing the assays
    have decreased, the fundamental process of applying
    clinical expertise to the interpretation of the identified
    variants remains largely unaltered and remains central to
    our contributions to patient care.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »