 So Alan Sholdinger Unfortunately couldn't make it so I was asked at the last minute to pinch it Many of you who have heard me speak know that I like to either quote classic rock or a movie in every talk And I didn't have time to put in a slide But it occurs to me that for those of you who know the princess bride Alan would be Wesley And I think that makes me Prince Humperdink the alternative to suicide and I think that makes Mark Princess Buttercup Yeah, that's me the other point I'll make is that yes I did intentionally with the same shirt as less and that's to highlight Rex's point about the difference between clinical and research Being very very hard to truly tell Let's see here got my pointer So two disclosures I am a paid consultant to medco health solutions on their laboratory and therapeutics committee I don't think that makes any difference here today But another important disclosure that I'll factor into what I'm gonna tell you Is that I'm one of three co-editors in chief and founders of a new resource called gene facts Gene facts is intended to be the resource that is more generally useful for clinicians at the point of care Not so detailed as the traditional genetics resources, but more useful for the typical general medical use resources The other point I realized that I meant to make at the top is that for anyone who desperately needs a bathroom or a cup of Coffee I'll tell you now the summary of what I'm gonna tell you as a clinician Tell me what I need to do make it quick and easy to find and preferably make it accurate and reliable That's it. If you need to go to the bathroom, you got my main point back to gene facts though and and the concept that less pointed out about the perspective versus a perspective one of the principles that we've been using in trying to build gene facts is rather than taking the The thought that we as experts know what everyone else needs to know We've tried to go out to anticipated users and ask them. Well, what do you want so that we can actually build something useful? And with that in mind what I tried to do is instead of stand up here and tell you what I think is a useful resource Because I am a geneticist and like most of the rest of you if not all of you we geneticists already know What's wrong with the existing databases? You don't need me to tell you that? I think what we need to talk about is what do non geneticist clinicians need to know So with that in mind, I figured okay, let's collect some data So I did a survey and I have to air my dirty laundry my IRB approval was last minute I'd like to get some data, so I'm not going to go to the IRB informed consent was Minimal I sent out an email and I said I'm giving a talk tomorrow. Can you answer a couple questions for me? So they knew what they were getting into they knew I was going to give a talk and present the data Subjects were by and large my colleagues at Johns Hopkins with a few exceptions So this is an academic primary care selection of clinicians made up in part of clinician educators who are actually taking care of patients Some of whom know some genetics most from don't consider themselves expert in genetics Some of whom are also clinician researchers seeing patients half a day a week But mostly doing research and you might think that researchers are going to think about things a little bit differently than clinicians But again, the line is blurred researchers do see patients and clinicians do occasionally dabble in research sometimes with dangerous consequences So the recruitment period that I listed here was seven hours yesterday as of half an hour ago though I got two more responses, so we'll stretch it out to about 16 to 18 hours of data collection So I asked two questions the first was and these are actually the questions that were in the original email to Alan So I'm just kind of picking up and trying to give his talk What resource or resources do or would you turn to if you want more information about the significance or management of a genetic variation? I'd rephrase the question a little bit But that's that's what I asked my survey respondents and the second one was well, what do you want? What would the ideal resource provide for you? So the next point is that just like I can dress up in a suit and make myself look presentable That adds some authority to what I'm going to tell you I'm going to give you data from a highly biased small Set of data that is completely unscientific, but I've dressed it up really nicely in a pretty table and That makes it look much more relevant and and while I'm intentionally trying to make you laugh I think it's an important point That's when grading the the value of the evidence and what to do if something is presented in a neat organized way And looks really slick it carries a lot more authority than a guy in a t-shirt and shorts saying yeah I think that would be a good idea So here's what I found There were 16 respondents as of last night now there are 18 and Not surprising to anyone who actually does clinical care the first and most commonly listed resource was up to date by Half of respondents and in fact there's one more who responded this morning that also listed up to date Equally distributed among the researchers and clinicians The next most common source that people went to Was a consultant pick up the phone walk down the hall and ask the expert and Elaine made the point of don't oversimplify But as a clinician the clinicians out there want it simplified They don't want it oversimplified, but dumb it down give it to me simple After that after up to date, which is a internet-based primary care subscription resource All right, so at Hopkins our institution buys it we have it I'm not sure that the family doc in Idaho has access to all the same resources that we in major academic institutions have after that an Expert after that is general internet resources Google and wikipedia are the most common places people go and frankly I sometimes go there myself not wiki But certainly Google after that you then get into true primary literature and Note that only two of the clinician educators mentioned going to PubMed or Google Scholar although one of the responses this morning also listed PubMed All the rest of the people who are going to primary literature and guidelines were people who are self-designated as researchers more than than than primary clinicians and Then the next group is specialty internet sites the NIH sites OMIM gene test gene reviews and my own database gene facts you'll see asterisks throughout here those are on the next slide I'll explain in a moment and there was one no answer But I'll add to that one of the responses this morning was a little bit more specific than no answer the answer from this other Respondent this morning was I don't think there's a resource out there that tells me what I need to know in a useful and accessible manner So the asterisk footnotes qualifiers, etc One of the people who said Google actually shows up in there twice And this was a slightly more sophisticated answer so that person actually goes to Google But uses a fairly smart search so says NIH and the mutation that that person is looking for so That's at least trying to put some scientific value on what they're looking for the NIH answers One of them was that Google search that I just mentioned the other one was Some NIH polymorphism database which I assume the respondent meant DB snip But I don't know for sure and the respondent didn't know the name of the database But at least there was a concept among one and that was a clinician by the way that was over here not a researcher Someone at least was aware that the NIH has some sort of a database Hadn't actually gone there But thought that that that that might be a useful resource the people who said gene facts and gene reviews are in fact My other two co-editors in chief so another Another flaw in the study is that anonymity is clearly not preserved because if you know or can figure out who we are Then I've just called out the other two people among the people who listed omem So there was one more vote for omem. So now we have a total of four people who mentioned omem Three of those four said they would use it one specifically said I would definitely not go to omem Because it's way too dense and I can't get what I need of the people who mentioned omem One is a one of the co-editors in chief of gene facts and the one who said that omem would not be used In fact is the other co-editor in chief of gene facts And then the the notes of H or H H or H H H that was the number of times that someone commented I might use that resource, but it's really hard to find what I want within that resource So that's the results of my highly unscientific biased irrelevant study What does the literature say pretty much the same thing? Clinicians like to get their data from the internet. This is old data. So CD wrong was listed I'm not sure that's as often a place that people will turn anymore textbooks don't really exist anymore But maybe you'll look it up on your Kindle Consultants and specialists remain highly desired sources of information seminars meetings other things as well People in general especially highly active clinicians don't like to go to lectures because it takes up time They don't go to the journals in primary literature again because it takes up time and generally don't actually go to specific guidelines Although as we'll talk about later they do by some osmosis filter in some of the advice from their governing bodies or society statements Other resources that I use it didn't get called out in that database I certainly go to farm GKB a few years ago. I would tell people it's like omem It's a dense encyclopedia, but not so clinically useful That's gotten better in recent years and importantly the CPIC clinical pharmacogenetics implementation consortium Which Alan would have told you a lot more about where he here And other guidelines on that side are actually making it much more user-friendly for the clinician I often will go to the website of a specialty lab or call up the special lab So Elaine sorry the pressure you're feeling is true once you make a call We're going to assume that you've made the correct call So yeah, the clinical lab has to get it right and likewise we'll go to the researchers sometimes So the researchers have to get this right. I still love the flock heart tables And there's a website there for it for you. That's slowly collecting some some actual clinical variant data But even without that they're still useful lists of what genes are I'm sorry What drugs are metabolized through what P450 enzymes and then algorithms for decision support? Warfarin dosing org I think most of you're familiar with and hopefully there'll be more of those coming online as we get more clinically relevant data I should point out the results of the needs analysis study We did several years ago to show that we should build gene facts We didn't ask questions about can you look up a specific genetic variant? But we did write what we thought were 20 reasonably relevant questions in clinical genetics And we interrogated nine different databases to see if they could answer those questions completely and accurately And what we found was the the genetics resources, especially gene reviews were good Up to date was the best of the non-genetics resources and most of the others really weren't good enough and really Distressingly some of the resources had wrong answers my favorite which most of you are probably familiar with some of them explicitly stated That you can only inherit hereditary breast ovarian cancer BRCA mutations through the maternal lineage and you don't have to worry about the paternal side black and white clear statement of fact yikes Second question what would the ideal resource provide I got 12 responses to this characteristics of the resource first and foremost it has to be concise easy to use I've got to get in and get out fast and It also needs to be accurate free is desirable Educational was listed as something desirable someone just said something like up-to-date by which I presume they meant put the actual genetic variant data In there and make it so that I can find it and one person wanted links to other resources Which we saw highlighted in one of the other talks already as? far as the content give me management information give me the clinical significance tell me what it means that was four responses and I listed separately because I think it's important to note But that's really part of management one person actually specifically called out give me clinical data support a clinical decision support Right within the EHR so that's really part of management and implications But it's nice to see that at least one person is thinking about putting it right there in the medical record Actionability clinical utility is common information about testing tell me about the clinical validity Tell me who to test and when to test tell me about the test method tell me how to interpret the test Tell me the cost of the test one person called out Other things for the content tell me about the path of physiology the phenotype the prognosis the whole spectrum of clinical manifestations What does this thing actually mean tell me about the frequency of the variation and especially highlight the most common variants? Two people wanted to know about inheritance patterns or the frequency of a de novo versus inherited mutation I think it's important to point out that I sent out my my little survey and while I mentioned we're asking about genetic variants Most people probably just read what do you do when you have a genetics question? So some of these answers might not be unique to our question today Some of them might be tell me about a genetic disorder, but still people want to know about inheritance and inherited versus new mutation and Two people asked for evidence-based information only two of the now 18 respondents One of them was a clinician educator one a clinician researcher, so only two out of 18 What the literature says is clinicians wanted to be accurate they wanted to be accessible they wanted to be clinically relevant and they won't spend more than two minutes looking for answers and Other papers have gone on to say that essentially what clinicians are looking for is a trusted Authoritative research that will go out and digest all the primary data and present it to the clinician in a way that they can Just use from there. They generally don't want to go look at the evidence themselves So is evidence-based important and here I'm reporting not only what I got from my colleagues on my own opinions But but this comes in part from email communication with Alan So this was part of what he was going to tell you also There's a lot of variation in terms of how much evidence users actually will demand Some want very high standards will only implement some action if there's a good randomized clinical trial or better Others have somewhat less rigorous standards and will use Variants when there's a preponderance of evidence that suggests that they're valuable valuable And many people will follow their specialty society guidelines and recommendations So when a society comes out and makes a statement that actually carries weight, that's good or bad So for example outside the realm of genetics orthopedic societies will tell Practitioners that a patient with an artificial joint needs to take antibiotics before they get dental worked on and it fetches disease Society say nonsense the bugs that cause joint infections in artificial joints are skin organisms not oral organisms So this is stupid. Don't give antibiotics before dental procedure So do you want to follow the advice of your infectious disease expert or your orthopedics expert? Well, what's the evidence say? I'm not sure All right, that's all I got Do you get the sense talking to your colleagues that your clinician colleagues view genetic information any differently than other information? Or do you do you get the sense that you know just like any other test any other Factor considering clinical medicine. They just want to be So as mark pointed out we've all in this room done a great job of building up genetic exceptionalism Much to our own chagrin most geneticists just want to treat it I'm sorry most clinicians just want to treat it like any other information with the caveat that they're aware That there's some bugaboo about it and we need to be careful about something and generally what that typically translates into is Gosh, that's complicated. I better just refer to geneticist and not touch it Which is also not so helpful because there aren't enough of us geneticists out there Yeah, so what are your non-genetics clinical colleagues make of the classification probably causal? What do they make of probably causal? I equate that to When we would look for pulmonary embolism before we had good CTs and we'd send our patients down to radiology for nuclear medicine scan and The report would come back moderate probability. Please correlate with clinical suspicion Which basically meant yeah, we don't know might be might not be and it was completely useless So you put everybody on heparin and you waited to see if they really had a problem or not And we probably caused a lot of bleeding Problems in people who didn't really need to be heparinized because they didn't have a pulmonary embolism So no, it's not so helpful to say uncertain significance But it's better than saying a benign variant is pathogenic and it's clearly better than saying a pathogenic variant is Benign so while it's not great to say we have no clue what this means. I think it's better than lying It's probably about as useful as the Entry after earth and the hitchhiker's guide mostly harmless That that was the update. So the encyclopedia originally said harmless and the update was to add mostly. Yeah Terry Minoglio from from genome it's interesting that the list of information that people wanted about Variant didn't include cost of the test or where you can get it Practice over at the Navy and and boy trying to figure out how to even order these things. This is really a challenge. It is So one person did ask for cost Nobody specifically asked where to go to get a test But there was when and who to test and test methods So I think people danced around it remember This was yesterday and I said spend no more than 30 to 60 seconds answering this question So again, it's not a very scientific study, and I'm not criticizing your respondents are you Yeah, just asking you for your opinion Do you find that your folks at Hopkins being a tertiary care referral center? Yeah, that where is not a problem that they just do it at Hopkins or or is this really a challenge? So I guess a little bit more about the sample size or sample population. I surveyed general internal medicine folks I didn't survey specialists Specialists get a relationship with one or two labs. They know where to get their test. They do it It's auto magic Generalists don't have a relationship, and I think that is another barrier that even if you have an idea of what test You want to get it's so hard to figure out where to get it and fill out all the paperwork that it probably doesn't happen And in fact when I'm seeing patients in my primary care office without a genetic counselor I have a much higher threshold to order a test than when I'm seeing someone in genetics clinic with a counselor Because I got to fill out all the paperwork and figure out where I'm going to send it off So even someone who's informed it's a barrier Just to add on to that I Think this is it this is a big issue, but as we develop these resources I think there's a couple of other barriers one relating to cost where with the genetic testing registry that's coming online that there was a lot of pushback from clinical laboratories about actually submitting their test cost data to that resource not all of them were really Thrilled about the idea that that would be out there the other thing that Eileen didn't talk about But I think is is really critically important in a discussion like this is that some people treat their Test specific information as proprietary and use that as a competitive advantage Which I think in the long run has been to be extraordinarily harmful To this type of effort. So I think those are things that we as we think about what is going to be in the resource How do we incent people to to really do this? And if I could just add to that, you know, I think this whole discussion highlights the need for us to be thinking about the day when Either the whole sequence is there or a variant table is there because it completely makes all those issues go away Great. So if we move on to our last speaker now before the general discussion