 We are ready to reconvene here with the discussions about exposure. Okay. So this brings all in entries to the table. All of you all start. Yeah, I guess I'll just work it out. Since last meeting, there been conversations between Mike and myself about how to tackle the issue of exposure, considering there's not enough exposure paid out there to tackle with. So basically we have to develop some tackling dummies to help us wade through the problem and have to deal with a lot of data. So we spent the better part of a month and a half on, we have charts in the back of the book, the most important chart. The first thing we did was decide the issues that were out there in terms of what product or source may be there. It doesn't mean we're going to do an exposure calculation for each one, but we want to at least lay out in a systematic way what we call the product or sources, what the root of exposure, what the target populations would be. So in a sense this is the first order differentiation task. And what we did do is we labeled as our A category children, and you can see that everything from diet to the toys were involved. And that doesn't mean that there's data out there, but at least we wanted to define a dozen issues for that population. That's one of the populations of interest. The other one which we thought would be very important would be personal care products, which would be more generally used by the expecting mom's mother or a woman of child-bearing age. And that's C. And household B is something that would be the second order of concern for the children when they're being, after they're born, and the first few years of life, since they're going to be spending a lot of time being rug rats, near or close to the floor, and pretty associated with that. So I think we were pretty comprehensive. We probably missed a few things, and anything we missed, please let us know. We then acquired some data, and the data is sparse. Just to give you an idea of what we're looking for. We're looking for a particular compound, and so when we look into cosmetics, cosmetics is one thing of very important use for adults, mainly pregnant and expecting mothers, and infant use, clearly, at least once a day, twice a day, maybe, at least more. And see what we could find in terms of some concentrations, and we found concentrations in various products. Again, that doesn't mean necessarily that's exposure, but that means that there's a concentration there, and now we have to translate. We have to use various tools to create a license. If you go back to cosmetics, or children's use, you go back to the original pages we developed under the, let's say, shampoo. All right, and the shampoo, which is A7, A7. One of the things is inhalation in children, which would be an appropriate concern. Baby lotion, dermal, actually shampoo was dermal, and baby lotion was dermal, in relation to the hand formula. So we would have to come up with scenarios that take the concentration of the media, the frequency of the amount of use, the amount of absorption you might expect to determine what the potential exposure would be for the dermal pathway. We're looking to generate an internal exposure number, which is the first step to developing a daily intake, which would be the next step of looking to the acceptable daily intake value that's going to generate in the toxin with the hazard index. So our goal eventually is to come up with daily intakes from each route for typical scenarios and doing it for a daily value. For most of these, they're probably one time during the day when this occurs, or periodically as they shower the child's shampoos in the morning, shampoos in the night, or up to two exposures. Or if it's a single contact, then it'll be a one, basically a two contact for that day. And you can build that over time if you want, based upon what you'll be trying to do with the daily intake and relating that to the target and affecting the concern. There are some, the next few pages are more and more cosmetics and they change. The main thing is the compound of concern is changes from each page. Then we move down to indoor air and some dust, which is a summary dated and collected by a number of different studies that was done by Weshler and Nazarov in 2010. And you can see that there were studies that were done by at least four of the groups. And then you have the backup data from the other groups that are mentioned in Nazarov. And Weshler paper shows exactly where they got the data from. This indoor air and some dust will be used for two purposes, going back to the original material on the first page. The inhalation, direct inhalation exposure that can put indoors. And then there can be dermal uptake and ingestion of subtle dust. It could be resuspension of subtle dust. There are a whole host of different ways in which subtle dust can reemerge and become part of a child's body model. So we'll be doing calculations for a daily intake based upon that. It's sort of similar to what, you know, you have the ingestion of soil in the yard while we're now looking at the ingestion of dust. Based upon activity patterns, based upon the frequency of children touching surfaces and amounting to the absorber surfaces. These are all things that are considered in those calculations. There's a lot of estimation here, and to be quite frank, we look them up with numbers that have a lot of uncertainty. Whether or not there uncertainty is any more so than the data that you generated because you have a single spot sample that's not of the same person but you're taking multiple people over multiple periods of time. It remains to be seen how much uncertainty each one has, but that's what we're dealing with. We have the indoor air and subtle dust from different locations in different countries. Settle dust in bedrooms, which is another area. Each one of these subtlety studies have either different locations, different number of samples, or looking at different phthalates. And we have to come up with a decision as to which phthalates and what data is going to be used in the analysis. Then we have another set of GAO, which is basically the migration of areas to the PVC and DINP and DHP. We have migration rates for those from surfaces, and that would require us to do a calculation based upon children's malding of toys and the migration rate from the toys and coupling those two together. They come up with an intake value based upon the frequency, the duration of contact, and then the migratability of the material from the surfaces to the mouth, which is sent in the essence of YouTube and in a way that they don't, excuse me, lead to an ingestion exposure, and then a deli intake based upon that exposure. So, where are we? Well, we're still acquiring data. There are three types of data that are necessary. One is concentration data where they're available. Two is the activity pattern data from the EPA, the exposure handbook, which is being collected. And then there's going to be a discussion between Michael and myself and hopefully Matt Lauer as to which of the most important things to include, which are the most important things to do the most calculations for, and come up with an assessment for like a day in the life of a child or pregnant woman, which will basically show their deli intake from those particular sources and products and try to see if we can come up with enough data to establish a mean 90th percentile, a 75th percentile, 90th percentile exposure, which would have to do with things like the amount of contact that you have during a day, the number of times you use a product, all those variables. The concentrations we get from various sources will be part of that equation, but the biggest things are going to be the variability and activities and frequency and lifestyle that we're able to acquire. That's where we're going to go. Now, the plan is that Mike is involved right now with assembling a lot of data. We've been marching back and forth about what's available. In fact, he just sent me an email a little while ago, I guess, CC'd everybody, and he just gave me some more data on some of the, I think, the teeters this morning. Well, it's everything in here. Oh, there's more in there? But it has... There's some backup data of that. I didn't print it out because the book would be... Okay, fine. All right, so there's backup data for, but it's not new data, it's just backup. Yeah, because it looks like it was fairly long. We're acquiring data, and in the beginning of June, we're going to be sitting down and we're going to decide what would be the quantitative exposure system for them to do. I think the first week of June, we decided the first or second week. And we'll be down here, if we meet myself, if I allow, we're in minimums, and probably some people from my staff. And that's where we're going to go. It's work in progress. It's really slow. The reason why it's really slow is because the backup data that's available is difficult to ascertain, because you have so many chemicals, and there are just so much data available because there's not enough exposure studies inside. What would be the main outputs or goal that you would see in early June, mid-June, when you put... I mean, what are you... Why we should have DIs dealing with takes for key products, key populations, individual populations at risk? So it would be for specific products, and then you would sum across the products? We can do summations across the products for individual phthalates, or we can do total phthalates. We can do individual... We can do the calculations. You can summon. We can do the individuals, which I think we should do both. We've done them. It's just a matter of summation. And then what we have to do is we have to define a range of variability based upon certain parameters, like either the inhalation rate or activity patterns. The amount of activity that a burden a child has for a particular product. Some kids do almost nothing all day long, simply videos, or other kids would be mouthing all day long. So you have that range of variability that you have to account for in your estimation. The 90th percentile, which Holder talks about, would be that child who does almost like pike behavior. Whereas the person who sits there is a couch potato, their source of exposure may be, you know, the potato chips that we eat in 24 hours a day, so you have to take into account there may be two high-end exposures, high-end exposure from diet or high-end exposure from the activities surrounding the use of a product. It's a little complicated. It's a matrix that you're going to end up with, but there'll be daily intakes, which can then be used to compare to the numerators that are determined by Chris Holder. And that's the end, the essence of what we're doing. We all have to focus on how the DI's focus on the ADIs or the hazard indexes generated by your calculation. And the DI would be a mean or there'd be a range? We're going to go for a median, a three-point median, 75 percentile in the 1950s. Anything more than that is ludicrous, considering, I've said it today, that itself is going to have a lot of instability in it. Anything more than that, I think, with being a little bit naive. So when you do the summing, you could sum medians, you could sum 95 percentile of the century, comparable to the bio-monitoring. Yeah, that's exactly the way we approach it. Could you say a bit more about how you get the behavior patterns in there? Is that something you're just going to summarize? Well, we have the exposure factor handbook from the EPA, which basically gives you a range of behaviors. You know how many times it touches their mouth in an hour, and you do the median, the 75, the 90, probably gives you distribution function in the latest version. So therefore you can choose what you want to use as your values, because obviously that will have a big difference in the calculation in terms of the actual contact that you have and the dose that you're going to get. The same thing for inhalation. It should give you very sedentary. You're sitting in a house very sedentary. Their breathing rate's pretty low. They're watching TV. They're not going to breathe very much. But if you're like one of my two grandshelves who runs around like a maniac, you know, until they have to go to sleep, well then their inhalation rate's going to be a little higher. So you have to be able to establish that range and use it. The inhalation rate's a pretty big number in terms of changes. The frequency of contact by hand and mouth is a big number of changes, very large. The body weight is not that big of a change. There's a sedentary and a 23% change there. Those are the largest, and then there are going to be other factors in terms of how many times a day does a child use it today. You know, you'll have those kind of things that are judgmental. You'll have to get information like that. That'd be great. If you assume once a day, that's kind of, for some toys, it's okay. For other toys, it's not. It's once an hour. What about food? The food source is a tough one. Do we have any data on the food yet? Well, I haven't pulled it together yet. I mean, there are sources. There are papers. But the food's going to be a lot harder than most of the others. It's hard, too. We've done dietary studies and compared that to food basket surveys. Survey data. They're doing great at all. Food basket surveys are based on the national median, whereas the actual dietary behavior of the child is so much different. You know, you think they're going to be X, Y, or Z, and they don't. It's just an issue that's very different from the geographic one. But we'll do the best we can. Can I just ask how we're presumably going to have another approach based on bi-monitoring and then back-contracting the exposures using information about kinetics. Will this be part of the report or will this be a second? I was doing that. I don't think we're doing, we're not doing pharmacokinetics, are we? I think it's just as sophisticated for the data that's available on you. Well, there's an angst data available about tonight. Yeah, you can actually, that's where you can come up with your jelly intake. It complements what? Chris and Holder. If you're looking to do the next step, which is basically go from what's in New York and back out to where it came from, that's a task that requires a million bucks. We're doing that now for the federal government, but it's very, very expensive. No, not when it came from, I understand that's impossible. No, it's not impossible, which is very expensive. I'm sorry, go ahead, Andreas. Andreas, if you have a look at our manuscript, you'll see that we calculate the daily intake for each individual. I know. I do realise that, and I do realise it is needed for this, but wouldn't it be better in a section about explorers because basically that's what it is? Well, that's interesting. Or is this the... would all of this sit better in a section on risk assessment? No, I don't think so. Because we're comparing the hazard which is derived from the ADIs versus the exposure which is derived from the ADIs and you're doing that comparability now. Is that what you're thinking? Yeah, I think we should keep it along the lines of the silver book or the red book of a section on exposure assessment. You are using one possible approach, the other possible approach is using urine data and then kinetics and backcalculate. It would be nice, I think, to have it side by side. And then, am I destroying now the whole hazard index channel? No, no, you are causing trouble at work. But it's good trouble. Because actually what you're suggesting is almost the same thing I was suggesting this morning except from a different point of view being able to link the exposure to the hazard index. You have to have a section in the hazard index that explains why it's being done to compare it to the external exposure and what we're on doing is I'm trying to provide this DI daily intake to help enforce that value. Don't try. Is it really causing causing a lot more work in my mind is just shunting what you've done, what you've put there over into a new section, isn't it? Yeah, I mean, so we have estimates, distributions from the biomonitoring what we would expect to be the daily intake and so it would be a nice table and figure sort of combination to have a distribution estimated by a monitoring versus your distribution point estimates to see how they compare. That's right, yeah. That could be a first. Well, and also I think it seems, maybe I'm jumping to a conclusion here that's not correct, but it seems to be an important feature of this is to be able to say what proportion of the exposure do we expect to come from various pieces, right? The best we can. The best you can. So, and is that going to be on a per chemical base? Could be as an aggregator for a chemical. The comparison will be based upon what you guys come up with. If you're going to do it on a per chemical, we'll do it that way if you're doing it on a total. We'll do it that way because what I don't want to do is I don't want to present something that's dichotomous from you because if we do that then we're almost lost because then people are picking us as far as they why do you do it this way for this and that way for that? It's not that, it's just the way we present the data for consistency and relevance. I think it's the most important. I think it's a little hard to say at this moment in time what the outcome might be but I think it would be very good to have this side comparison using this true quality of approaches. You know what it also does is that we're probably going to have very weak diving extremely. At least this will be able to we can compare the estimates that we made for the products to what the daily intake is the daily intake is that they calculate. We can say well the denominator here which would be the daily intake from the biomarker is X whereas what we calculate from cosmetics is Y and Y is towards magnitude less than X or vice versa then either is it the minimum part of it or it's a major part of the issue and that requires different risk management positions on that product. I think we can do that which would be kind of an interesting outcome. Are there other standard sort of quality steps in the exposure world where you say here's our calculation now here's an external way of looking at it to see if we're in the same ballpark we have. Do you know what I'm saying? Are there cross checks somehow that are typically done? The only cross check you really have is basically the comparability between studies and you're dealing with a human observation and comparability. With respect to the exposure factors handbook I believe it's the European version of it now isn't there? Or do you use EPA? Does REACH use EPA? I'll have to find out whether or not REACH uses EPA's exposure factor handbook and they have their own if you're doing their exposure calculations if that's the case it doesn't mean that it's going to be symmetrically consistent because European children probably don't have the same behavior patterns in American children so they're all different. That's right they're all very fixed. So in the exposure factor handbook I don't think I've ever seen that does it talk about it's on the internet just go over it and put it back in handbook 2009 it's about 700 pages maybe 700 pages more the key thing is that it has different categories. There's like body weight distribution function for children it has inflation rate it has a whole bunch of different variables that one can use in certain types of exposure sensors so we're going to do a select one that's most appropriate for this particular event. Okay just try and get my mind to get my mind around so you're talking about you're going to assume here's a sedentary child who sits around playing a lot on the floor is that the kind of level you're going to do and then change the scenario to be some money? We're going to use based upon what's an EPA handbook choose it what are the sedentary versus active child there's a whole bunch of things the time duration of active they have it down to some cases how long a child does certain things in certain activities so we select when we think of reasonable ones there's always the god damn fool in the hill that you can't deal with as a child there wasn't anything you'd say that still goes out to the level of tracks but those are not the ones we're going to choose all the band went into effect in 2009 are these data from before and since the band or are they from early? Yes and in fact most of them are in the last 10 years I think it's helpful in some cases to look at how they change over time because we know that manufacturers are reformulating their products so there's always, for example each year there are fewer balance and the same may be true for cosmetics and other kinds of products so it's important to keep that mind and I think normally you would give the greatest weight to the recent studies and also US studies but if all you have is older from Europe then that might be what you have to use I was thinking of an alternative to Europe we're looking to our friends to the north and it might have a very similar use profile and a similar set of toys and things that kids suck on but did the band apply there? Well they have their own in fact I think theirs is more recent and I think it was they just announced it I'm not sure what the effective date is but it's essentially band it's essentially the same band that is they have now or will have soon and the world well Europe and North America have kind of been through this more or less together so Europe band issued the first bands whereas in North America there were a lot of voluntary bands or voluntary reductions but I think we're all more or less on the same page right now but because of the band or at the time of the band some they should have gone down in fact they were going down even well the toys they were going down some of them either were substituted should have gone up exactly and we see that well I mean there's not that many studies well just talking about what's in toys in toys you've got well two things you could use a plastic on a PVC or you could use a substitute plasticizer and they've done a little bit of both that's for the titration experiment well that's for what's in the products right as for what's in people what's in people we don't know the substitutes at least for the most part I know what you're is T.B.H.A. and Haynes I don't know but I think in a few years these studies may be shifting to from ballads to the ballad substitutes to the biomonitoring studies well even in exposure studies it used to be that in exposure studies we used to measure PCBs now we measure PDVs instead we switched our chemical of a month above that and probably eventually when that's banned we'll switch to something else but then studies define it in the home which usually lies behind in two to four years depending upon the ubiquity of use and the amount of concern that is expressed in the community about whether or not this is putting a rampant increase in XYZ and ZZ well we're three years old we should begin to probably see some of the work but I don't know how well we can publish it I guess I'm looking for some prior validation and something is done about extending interim bans or new interim bans that we would have the confidence that the data from before reflects something happens if we can't tell that there's any change in the exposure from a ban then why would we ban something there's something wrong with the assumption that a ban would cause a decrease I think one of the things you have to look at there is basically the product data and the impact that the ban has led to reduction in product use of this material and in place of the other use that can lead to you to make some some kind of assumption make a calculation about the turnover of this product and we might be able to get some information on that to figure out how much there's out of distribution in this point of time which will lead to reduction in exposure even without any absence of data I mean there's the data are few and far between but there are, for example one or two studies in Germany showing that DEHP exposures from the biomonering have gone down over time but DINP is increasing because DINP has been slowly replacing DEHP and you know you could see on a very broad scale trends like that even those data are hard to come by if there's a movement away from those products that contain those chemicals of concern that probably means that whoever buys these products have moved into products about which we know less and we have a good handle on what the substitutes would be in those products that by surprise began to be purchased when they weren't before because the original ones are now well we know what's in toys now and we know what the substitutes are whether the tox data in some cases aren't as rich we were talking about not just whether the data exists but whether we have access to the original studies that was an issue for a couple of the chemicals I'm just wondering how risky our recommendations are because just based on thalates it might be clear what we want to recommend but the downstream consequences of that recommendation is something sure enough being asked to consider well in a way I mean that's why they put in the substitutes but whether the data are equivalent you know I mean that's the problem with comparing roots of exposure if you have really good data on one root and really crappy data on another that makes it hard to do the comparison I guess what I'm wondering is that the possibility is there that a decision, a recommendation that we would made if it's followed through would drive to a new set of exposures that we hadn't anticipated you have the substitutes that we know today but with this open the door for others, a new substitute that has come along and we have very little data on it so we jump from the tripod into the fire it wouldn't be the first time well we've already gone from DEHP to DIMP and then now we're into the other plasticizers I'm just trying to figure out how we the point is really well taken and how we frame it exposure assessment slash hazard assessment analysis and we're you're looking at chemicals that have an intimate brand some of them are going off the market we rely on older studies we rely on older data which would tell us basically free band and even there we'll be able to see and I will get an idea of what their relative proportions are on all the valleys versus their relative proportions that does snapshot in time with other materials that may be many or have inherent concentrations of valleys the question you're asking is a perspective question which most of us don't have an answer to because we don't have this rule what is it what has the band done to reduce the current exposure the interim band done and what has the interim band done to the increase in valley substitutes that could lead to different exposures that we've heard before have never measured and in the latter one for least if we choose an example like toys we probably won't have some product data that we can use in the analysis to say this could be an eventual possibility and in the relative scale 1 to 10 due to the levels we project based upon the substitute lead to a same level of concern we have for something that we have band internally or is it much lower and I think that's something we could probably do for other things depends on how much data is available but it is a conundrum like with some of the other things we were talking about we have to at least explain ourselves thoroughly the responsibilities whatever makes the decision to see it will be on our responsibilities yeah true although any to the extent that you could make recommendations or otherwise enlighten that decision we could raise that would be helpful yes we could raise some yellow flags definitely who else comments your questions Bill anything from you no nothing from me working progress working progress working progress I fit in I'm down here in this section but I don't know how this happened I see my role in supplementing providing animal data animal evidence on mixtures of phthalates and other chemicals and phthalates and then moving over into the area hazard index etc etc let's look back at the assignments we may well have assignments assignments assignments assignments assignments are you're the beginning here in fact I know after these sent out there was some discussion about there were comments on some comments on the assignment and this captures all of those comments have you done it for individual and cumulative exposures so what do you want to enlighten us about well first of all the individual and cumulative exposures individual and cumulative that laps very much into the area we've just discussed so I don't think I will there other anti-antibes I will do, I can do and really what's my notes really from last meeting don't quite 100% agree with this outline as I said there's already a sub-chapter in I think yours or Phil's concerning mixture effects animal evidence I will build on that plus evidence concerning other chemicals quite a lot has happened recently there so then in the risk assessment arena of anti-antibes that's going to be complicated we're gathering data about exposures for pesticides we've just recently published about this ourselves this is going to be difficult but also in the light of the discussions which we have this morning this will be used as sort of additional evidence to inform us whether the assessments we make for phthalates after alter in one or the other direction and other than that I think I will be very comfortable joining and contributing to the hazard index bug about chrysanthemum does that make sense can I ask a question going back to the outline something that's signed environmental fate I don't think there's any environmental fate issues in this report we're hoping mainly about developing exposure estimations for humans whereas environmental fate sort of gets into realm of any sort of ecology why do we care about the fate of transport in this particular report I just don't think it's necessary to discuss it which is consistent with the calculation that would be in the human exposure chapter right I think that's what that was made to say well fate doesn't ecological definition I just wanted to bring that up so that later on we didn't bring it up with Andrea I want to make sure we're right with Andrea where it says individual and cumulative exposures I wonder if that's supposed to be risks just talking about the whole area of cumulative risk assessment the theories and so on yes the cumulative risk has been sort of co-opted by the hazard people for the last couple of years because when I I was at this EPA meeting in December and I was he was noticeable there was a lack of appreciation that you would have to go backwards to the source which is one of the goals of this group for CPSC is to figure out where risk management can deal with the sources of this material so I was struck by the fact that the cumulative risk assessment has really really going away from the whole risk paradigm and been more associated with a hazard paradigm which is precluding the ultimate goal of risk management which is one of the issues that we're dealing with here your observation is correct in the sense that the field is currently grappling or has been grappling with with the question do certain chemicals act together in a cumulative way and that's question number one and number two can this be actually predicted when you know the potency of the individual players and the mixture that's the that's essentially a hazard assessment what the field is grappling with right now is how can this knowledge mainly from experimental studies be utilized in risk assessment and hazard index approaches one way of doing this alternative to doing this as well but your observation is quite right but that's sort of at the moment the discussion is headed because at least I see where you're heading in so therefore how we frame is more based upon well looking at cumulative hazard as one component of the risk assessment one aspect which needs to underpin the use of the hazard index approach is to scan through the experimental evidence to see whether the data are actually in agreement with dose addition or concentration addition which is an application of the hazard index so this is in progress as well so I'm sort of going through the data to provide the basis for an application then of this in terms of hazard index approach and some of that is in the NRC some of that yes but that's happened since then it needs to be updated that NRC that NRC report really hurt the CPSC analysis because the fact that ignored the exposure issue to the point where it was only three pages an entire NRC document it was more of a not a cumulative risk it was a cumulative hazard assessment document and that when you go back and you read it it was working it's about what the exposure really was going to provide because the remit for the NRC panel explicitly excluded I understood that we were not allowed to do a risk assessment I understood that just that the way it was they would have gotten very nasty with us I know, I know, but if you read it it doesn't come through that that that weakness has to be felt eliminated somewhere else no no no it says very clearly at the beginning what the remit is I know what it says but I've interpreted it and I've seen it poorly interpreted well it hurts this analysis because a lot of data is not being collected right now which is very valuable and that's where it turned whereas in this analysis we have a chance to overcome that weakness by reaffirming the fact that with what we're doing here we're reconnecting to the external and the internal exposures which I think is essential for the purposes of what you're doing what it has to do with what we've done just an observation so first of all do you agree with how Andreas is going to work with you and what he's going to attribute so we gotta read can we also have any questions for Andreas well one question I don't know if it's for Andreas or someone else but actually it says individual and he looked at I guess we could change the name to risk are we meant to be looking at individual sort of risk evaluation of each chemical by itself even then we're going to put them together we're going to stop it and do one at a time first hmm I might just let him turn it when he's been producing that's going to come up tomorrow morning oh how are we actually going to start out some of those pieces to arrive at a risk but that's a very important question we have to answer I think it's sort of hard to whether or not we're going to look at total daily intake or look at individual daily intake let's look at that after we've had the full discussion on the hazard I agree that's why I put it tomorrow morning so anything else before we go to the hazard index Bill no if I understand the discussion I just occurred with respect to Andreas's contribution he's going to lay the foundation for doing cumulative risk assessment is that correct yeah the spade is at the ready that's what I'm going to do that's exactly what I meant yeah that's going to be clear yeah I think he's clear that was clear yeah okay then let's get started on the hazard index and expect that we're probably going to put temporarily in an over-solve just for a break and then continue but you have the rest of the afternoon and I think we've already had a lot of hints that this is going to require a fair amount of discussion so let's get started we're of course at the start we're going to jump to the microphone if you have an updated version please disregard the one that's dated 314 and refer to the one that's 329 um and Phil I hope that you got the one I emailed this morning if I did um I'm not sure the best way to tackle this I mean should we break it up into the general approach and then sort of what the analysis of the data do you want to do it that way um programming do you want to talk about the part so how do you I know it's complicated to um well I guess you know somebody said this morning about the hazard index definition shouldn't be in this chapter maybe that's true it needs to come earlier um but this is just you know the way we got started recognizing that we need estimates for daily intake we need estimates for reference doses we chose here to use reference doses related to anti-androgenicity um similar to the court and camp and fast approach we could certainly do it multiple ways I don't know that there's value in saying we have to do it only one way it seems to me thinking of it like a sensitivity analysis if we're coming up with about the same thing from a whole lot of different approaches it seems to me there's some strength in that it's not necessarily a statistical test where you know this is the way I'm going to do it um so that's where we are now I think we could come away from that and change those reference doses um so the point here for this approach is to use biomonitoring data to estimate daily intake and I guess in short we've done this in a couple of ways um based mainly on different values for the reference doses which I think we can hope will be um sort of supported in the earlier chapters the values that we use there um we've done the analysis from biomonitoring data in pregnant women from NHANES and children in NHANES which ends up being 6 to 18 years old and then we've got data from Shauna Swan in up to 3 years old I think a few weeks or up to 36, 37 months old um so I mean I guess so in short what we've done is we've taken the um 6 um diasters and we've added to that DIBP um so the ones that were considered in the Regulation and the DIBP um so that gave us the 7 and we considered them just as a set and then we also considered them with other um we had data for other 3 other antiandrogens um we considered the hazard index with without the others 3 and then with the 3 and then we went back into the Quarten Camp paper um and essentially got constants for an additional I think 7 and also uh antiandrogens to just sort of shift the distribution based on constants from the hazard index and we can go through that but in short that's what I mean we can go through the details of the daily intake estimates, do you want to do that? The basis of all our calculations is the calculation of the daily intake of each individual delate based on human biomonitoring data so this is the core step in this hazard index approach and we use for the daily intake calculations and of course there are a lot of assumptions to be made and we assume there is a steady state excretion of the metabolites we use and that's the nice thing a formula to calculate this daily intake which has been both published by the delate ester panel and um independent research from CDC and the environmental toxicology program of the NEH so I think we are in pretty solid ground using this daily intake calculation formula so that's the formula in itself then the parameters we use are the metabolite excretion fractions so this data is extracted from metabolism studies human metabolism studies some of these studies we know it's a weakness in this approach are based only on one or some individuals but we are confident that for many of the daily intake questions we soon will get more reliable data on more individuals so this is the one assumption we make the metabolite excretion fraction the second assumption we make is we have of course to extrapolate from the spot urine samples to a 24 hour exposure so this is the one we use to smooth creatinine approach and for the pregnant women we use the normalized creatinine excretion all this is described in the publications by me and both by the delate ester panel and the cornet al publication from 2000 in DHB so this is the basis we calculate the daily intake for each delate and each individual and then we sum up that we relate this data intake to the tolerable data intake and sum up the ratio as the acid index can you explain what you mean by summing from J to C on the very front page so C being the number of chemicals in the set so there's going to be an estimate for daily intake for each chemical the diastrosity there's going to be an estimate for the reference dose that we choose and so it's that I guess it's called a hazard quotient and then we sum that quotient across the different chemicals so that could range from one individual chemical to a composite of a bunch of them the DI is based upon the measured metabolite concentration in the in here metabolite or metabolites single or it's a summation depending upon what you're doing with total versus individual daily intake and we we have in table number one we have the metabolite excretion reference so let's say the dilate that represents 69% of the oral dose of dilutive dilate and using these excretion fractions you can extrapolate to the dose of the parent dilate maybe we should add references for each of those those are all the ones that are not set at the others are all on literature values so maybe we should put another example here is from a different study most of them yes but I think Mike we can say that we are aware of some other studies that might produce new excretion fractions I think some of the studies are initiated by even the European chemical industry or the American chemistry consulate but there, last year we've been to Berlin we have several presentations on this topic I think they are European I have a question about the excretion factors when you have multiple metabolites and multiple excretion factors do some of them you don't average so let's take let's take the HP as an example we have four metabolites and these four metabolites taken together represent 62% of the dose and we add up these metabolites on a molar basis and then multiply with or divide by 0.62 ok so this hatch index is it going to be a single number or is it going to have it's a single number the neat thing about this I think is you end up with a single number per subject per subject so there's always that question I think in terms of these hazard indices about what should be in the numerator you know are you going to have the median for each one or what values do you choose for daily intake the neat thing about this is each subject is their own sort of mixture so from that we would never if you have a thousand subjects you have a thousand per subject yes so you can get to look at a distribution of hazard indices to see and so a lot of this report it's like being some histograms and only the estimates for the daily intakes which are back in the appendix in the very back but even histograms for the hazard index in the various cases you know with a reference line of one or zero on the long term scale so the biggest I guess the biggest source of uncertainty is the one for infants if you're having infants except for a small number you want to study from the outside right so we've got okay so if you want to switch now to just look at the data part of it I'm just looking at we were talking about children versus women and for the infant database you have 6 to 18 versus adults but you have a big dark hole except for that one small study so you need to data at this point just trying to get an idea so we got data from Shauna Swan and that's we have I can't remember the number is it 130 130 and that's the pregnant women it's like it's coming it's 291 children ranging from months old to up to 37 months old and the thing that we didn't know about them that we made an assumption about was the what the smooth creatinine excretion value was because that study that came from was actually for children 3 to 18 so we set that value based on that the Riemer study but we didn't assume a continual slow down we set it at a constant so that's an assumption we made children's creatinine and diapers and that's what so the one thing that I still need to work on is also additional to the infant data that Shauna Swan provided for us were pre and post natal estimates or data from pregnant pregnant women and I haven't had a chance to do anything with that yet so that will be added to this I think it will be interesting to see sort of even correlations are bivariate plots of the estimates pre versus post natal per chemical basis but just in terms of quality checks Holger and I spent 5 hours yesterday going through this in detail and we actually confirmed some calculations where he had some data he had his calculations done and then I used the algorithm or the program that's used to analyze these data to also calculate the hazard index and or the daily intake or whatever and we were able to show agreement in the two different calculations so I think that was a nice all of the stuff that's moving around here I think that was a nice quality control step so should we go through the data or I'm not sure what is that ok um well ok so if we just kind of glance at the tables um table 2 is a table that maybe could be um linked to what's in the previous chapters so this is where we have uncertainly factors and reference doses given on a per chemical basis um so the case 1 is largely from the courting camp in Faust paper the case 2 we constructed based on I guess testimony or results from Earl Gray when he visited us at our second meeting um we already discussed yesterday that we might implement the case 3 with the EPA reference doses although they are not all fixed on the same endpoint but we can add this case and we can add a fourth case like using the European TDIs so we can play around this we can add cases with different factors and reference doses and I think the value of that is I think largely what we're going to see is it doesn't matter that much what the assumptions are we get very similar distributions it's not like going as an order of magnitude off of the other um which I think makes me think there's no point in worrying too much about small details can you give a mathematical interpretation of this my my first idea to that why that is is that the variation in your daily intake is much bigger than the variation in the denominators don't collect I think that's probably a reasonable assumption okay so like we say this table 2 is what this report is based on um we're certainly you know as we develop additional results and the other factors we can consider other cases so that is something that could be added um so we're ready just to go through the data then if we look at um I have a question that can piece one to question the fetal testosterone synthesis why is there an uncertainty factor uncertainty factor for the two lower loopholes and 500 for the higher loopholes um that was described in the courting camp paper and I think there are Andreas you maybe even have that in mind I can answer that yes we made uh this is of course 2 degree arbitrary I'm free to admit that one can now go into fine detail discussing this and the other number but in principle what we did there we made an adjustment for the different number of animals used and so if you like it's a judgment on the quality of the of the underlying studies and we've highlighted in this paper that say it again that um actually none of these studies were designed by the authors for for point of departure I guess I have to do a policy of the hormone it has to do with the design of the study yeah too few animals but still you can derive some sense of potency from these data that's not the problem we just to finish the argument the concern is that due to the small number of animals in these studies the estimates for point of departure be that no elves or benchmark doses are a little dodgy really normally you want more just as we go further I mean take a peek at the reference doses in case one in case two you know you're going to authority compare the two yeah so there's some some differences there that you um you know and so think about that when you see the distributions as we go through this did you expect all these chemicals to be bad tightly bad what in terms of a reference it shows a lot of variability I would say it's more based on study design the reliability of the study really because um I think that's the beauty of case one compared to case two in case two the study design used for all these relates here was the same but it wasn't designed to to define NOIL so the case number two is very similar of same study design for all the relates but no investigation of the low doses and the right case number one is different studies with different study designs just another example in Europe for example the reference dose for dilute is 10 not 100 all depending on the studies that I use but this is the if you like strength of this has an index approach because I mean we can now begin to argue about finding those examples it doesn't matter at all when the numerator varies by say three or four orders of magnitude it really does not matter there's quite an interesting insight already and I think that histograms are going to point to that so so to look at some results then so the first section here is pregnant women so how detailed we want to go into these examples but I guess one important thing of the 130 women who were included in the sub-sample that identified themselves as um at least one of the 12 phthalate monoesters were detected in all of them and 7 were detected in 98% 9 were detected in 90% lowest detection rate was MMP at 19% so I mean it's a lot of phthalates in these pregnant pregnant women um okay just following down there so that the appendix figure A1 has the distributions of the estimated daily intakes for these pregnant women um and the distributions are on the law 10 scale the units are micrograms per kilogram per day and on the law 10 scale largely you know we're seeing things that are relatively symmetric with the exception of maybe I guess DNLP that's a little funky yeah but that's still the detection rate yeah and some skewed values up in the upper tail say DEHP and maybe DINP um but these I think will be distributions and we could put these in another place in the report but these may be the kinds of things that would be interesting to have side by side with some exposure estimates in general estimates if you could get to the same units for kilogram per day okay so it's just a draft so um okay again just looking through the report so um if you think about each person as being their own sort of mixture of exposure right and so there's a different sort of mixing proportion per subject so one thing that we thought about was to actually look at what that distribution sort of on average is so if you look at table three um this is again the same pregnant women and it tells you it shows you um what the average proportion is for each chemical and then what the standard deviation and the minimum and the maximum so what we see is that you know the average DEHP percentage is 82% and that ranges from 49% up to you know over 99% the others are lower in terms of the percent within you know a subject on average um and what you see in that table are are the um the means the standard deviations minimum and maximum distribution for case one so like if you can flip over two pages then you can compare it to the distribution in case number two so if we use the reference doses from case one we can see that DEHP is the major player here if you flip over to case number two which is table four you see that it's a more balanced picture of the so it's still DEHP 53% but the other delays are of importance also so one thing that we wrote about here was just to sort of say so on the maximum side you know what were the maximum values and in this case we see I think all but um you know it was over 50% maximum there was at least one person in the data that had a more than half of each of these their particular exposures so I think that just goes to the address the issue of you know the mixtures on an individual basis are all very different um as you see from these from these date tables so now I guess if you go back to um table figure two for the need of the matter here um figure two so these are um histograms in figure two a you could go yeah so figure two a there that's the histogram for the hazard index for these 130 uh pregnant women um and you see the summary statistics there the mean the median so you know the if you can't read that the median is uh about 0.09 uh the mean is is higher because it's skewed it's about 0.37 but there is somebody who has a value of 10 1% sorry can I read this now so more than 5% of the of the subjects have values above one it's hard to see this is one do we know what the mode is uh I can I can cap it would you like to see the mode what look at me um if you and Mike sorry to keep but if you switch over to the just to the side of that figure b is the same data except now with on the log 10 scale oh yeah no no I can see the value of that so now you have but it's a log yeah you log it you log it so you see that tail the upper side there um something that I haven't done yet that we could do if you guys are interested um for the infant data I actually tried to model the hazard index as a function of code areas which we can do now that we've got this on a per subject basis um so we were able to show that the hazard index was higher not here we'll get to in girls then boys um significantly higher so if we have some things you want to kind of look at I think it's right for looking at things so in terms of covariance um okay so now back to the if you go back to figure c sorry Mike um so now the point is this is now the hazard index for just the seven valleys um in figure c what we did is we said well now let's look at let's add three additional antiandrogens um it turns out that these additional so this was biomonitoring data from um this phenyl A propyl, paraben and um doodle, paraben is that right and so it does shift up a bit but not that much it's out in the decimal places so that so these additional um chemicals as the way we calculated in terms of their um the fraction and stuff didn't change the distribution that much um but it just shifted up some um figure d is just the histogram for that on the log scale but then if you go down the page so now this is the point where we get to say now what about these other antiandrogens so we went back to the courton camp and Faust um paper and there were seven other antiandrogens that were considered there now we don't have biomonitoring data I don't think for those so what we said is we said well let's just add a constant so we took the hazard quotients um for those other seven chemicals and let me see if I can then closalin for chlorase I'm not sure persimidine venerate penetrothium ppde and bde99 so we took those uh they were considered in the in the courton camp and Faust paper with um using median estimates for um exposure and then high estimates for exposure so the figure d here sorry figure e here is the hazard index distribution where we added the constant from those additional seven chemicals their sum of their constant um it's a constant of their hazard quotient um and again so we shift up a little bit more with those sums that's at the median intakes and the figure just below that figure g is if we go up with the um the high intake estimates for exposure so I believe in this case if I remember right case one about fourteen percent of the hundred and thirty women have values in this case I think above one um so that's that um I guess this is all summarized if you want to look at the summary tables uh table five so we did this both for case one and case two I don't know if you can remember the distributions enough these are just to compare the summary table if you go down um there's a pretty large table five that's got the um there that's it so case one and case two for the pregnant women and so for each of the cases it's the seven thousandths align the seven thousandths with the three anti-imagines from biomonitoring data and then that with the median uh intake estimates from the house paper and that with the uh high uh intake estimates with the seven other anti-imagines so you know the so here's the point where we can compare the distributions just in terms of point estimates for the median 75th and 95th percentiles um you know the medians are are relatively similar if you just look case one versus case two in corresponding pieces the 75th percentiles you know I mean I think they're pretty comfortable based on all of the moving parts here um the 99th percentiles are a little larger for case one but still I think pretty comfortable why do you think the 99 99th percentiles are uh higher in case one than in case two I think it's the DEHP it's the one that has some really high intake estimates and it's the one that's given the most weight in the case one but as you can see with the infants in case one and two look at the 99th percentile here it's case two with 0.75 which is higher than case one so it depends on the distribution of the of the metabolites of the parents 0.05 probably one in case two when you're moving it this on a percentage basis and asking why at the medium if you look at 7 tallys versus 10 and 10 at the antigens there's a triplet between those two from quite 1 to 0.3 and if you look at the 99th percentile and the 95th percentile they're almost the same there's a triplet in difference there's very little difference why do you have a larger difference at 90, 50 and 99th percentile because it's the it's the it's the well I remember the the third and the fourth row there that's just an added constant so you're looking at a ratio of things that you're just adding a constant so the constant is more in the middle than it is at the upper tail the tallys are dominating in the upper tail and you're not necessarily in the dominant part you know what I mean difference versus a ratio yeah ideally it would be nice to by monitoring data all of it but then you conclude that the other antigens don't add much on the factor relative to the median relative to the tail relative to what? they have an impact on the median they do have an impact on the median but not so much on the higher external tail because as you say there there's an exposure to DHB predominant the tail in the other direction would be minimal difference as well below well so the third and fourth row of this we're adding a constant but it's not the 5th percentile of a constant for a single 25th percentile population that are exposed to the tallys in three antigens because you start with a small number compared to the small number the added effect if I remember right it's like 0.8 and 0.56 or something like that the other antigens would have more of an effect than the low level exposures low percentiles than they do in the high that's right so if there's a place here where the antigens are having a noticeable difference it's in the low percentile population that's right but I would say even in the middle of the population even in the middle of the distribution like you pointed out it's in a 0.13 to 0.72 correct and I'm just looking because they're a part of the population towards that impact is even greater it's not at the high exposure not at the high percentile it's at the low percentile at the medium and the medium represents the effect of this to the greatest percentage of the 131 but there will be a couple for which the a couple of the 130 for which the effect of the anti-androgens is greater than what's for all the rest but it may fit right the high levels, the other anti-androgens and low levels are now like this it's still the effect the assumption is that I'm just learning slowly the dose of the anti-androgens at the 5th percentile is the same as at the 95th percentile the construction here is that that's right because we don't have biomonaging data and that's not true for the three anti-androgens this phenylate purple and butyl paren butyl this is actually based on biomonaging data so that shift there is because we've added additional chemicals interesting simply because frequency distribution and intake these data are not available for pesticides we have to be aware that we didn't add up individual 95th percentile so it's individual urinary data that was extrapolated so we might have an individual with high DEA tree but very low value of the late exposure so this is not a worst case construction it's a picture of the exposure of the distribution we could model a worst case in saying that let's assume that somebody had exposures all in 95th percentile so this is not a worst case but this might be more representative than constructing something like that the biological effect of these three anti-androgens or alternative may detectable an absence of phthalates would there be a biological effect of these anti-androgens in the absence of phthalates I don't know that you could just subtract off the first row from the others and that would be the effect of the added part why do you think the median is not part of the I understand just meaning an assumption Mr. Burns is asking the effect of the others that's under the phthalates I'm saying the factor I mean this is loose but the factor that's due to the phthalates is in the first row of each block why is it just the first column the first row in each block there is the seventh phthalates by themselves and then like that so I think by any answer to the question is the S so we're composing the effect of the phthalates of the anti-androgens we're not down there in an area of anti-androgens where by themselves they have no effect what I'm trying to figure out is which should we worry about more the anti-androgens or the phthalates well I think we're in a data poor case here for most of the anti-androgens except for the phthalates we've got more data for three of them there's probably 52 phthalates I agree with Chris the problem with the other anti-androgens is the data quality is poorer the data quality is good for phthalates so I think it's 50-50 but because of the uncertainty about that should we actually be doing this step I understand the point do we just keep the first row versus column well we're returning to this morning's discussion I think I think it is instructive to see what happens if you factor the other one's in why do we have to be so important because we're we're going to come we're going to come which also can lead us into vast array of questions and confusion because we don't have the strong these questions can be answered I'm not sure about that I think the point to take home here is it isn't per se clear or a theory clear that the other anti-androgens might have an impact on the hazard index distribution of phthalates do we care for the purpose of this exercise do we care I think in the context of what happened before with the NRC report it would be we could be accused of being negligent of not taking this into consideration what then in ultimately we do with it with this kind of supplementary information if you like is open to debate it may or may not have an impact on our recommendation I think it would be that's my opinion politically unwise to totally ignore this considering what the NRC panellists said considering what's out there and debated in the peer-reviewed literature but do we have to go in a manner of detail why do we have to go medium intake high intake we're becoming much more uncertain as we go down the list it's not much more uncertain we can do this because of data of that you should be, do you normally have people with more data of that when you have more uncertain data it becomes uncertain I think from the court in camp and Faust paper there are estimates that were studied examples of what the daily intake would be and what the high intake would be and there were considerations of what the records those would be so all that we've done here is just said we don't have biomonogy data on those other seven but they certainly are very likely to be anti-androgens so we don't, it's not just that you would start to say for example in case two the 75th percentile is 0.18 of the seven phthalates and considering the other anti-androgens the value ends up being at the high intake so that may be the most conservative but it's 0.96 so now all of a sudden you're very close to the value of one so you I think what it means is you don't start at zero you start with other things that could potentially add important important risk I think you make a valid point we should first need to discussion only focus on the phthalates maybe we can extract just the phthalates in the first table and then discussing the phthalates what's the problem is with the different cases in the different populations and then we move the interpretation with adding other anti-androgens to another table that would be good because I would separate the argument rather than mixing them up exactly which is what I'm fearful happens here which is a fine wonderful presentation again it's presentation because you muddle up the questions that we're supposed to ask you don't have to put too much explanation into it or if you separate it out you have a clear explanation and then you say to be consistent with the NRC report we went and did the next step alright and the next step shows this let's be in simple words in other words we can say for the pregnant woman or pregnant woman using the case 1 more than 5% are above the or have a hazard index higher than 1 if we use case 2 it is probably only 3% that are above the hazard index of 1 if we look at the infants for both cases we are close at 95% higher below the hazard index of 1 and looking at the children we are almost at 95% higher above the hazard index of 1 using case 1 and just below using case 2 yeah that's fine I would share a concern that is different from that I think what you just described is an improvement in how to present this but I still have another concern and that is just mentioning this phenolate this phenolate has a lot of baggage and it has a lot of people who are activists in making sure that its concern is maximized and in another group of activists are certain to make its potential effect minimized and the only concern I have is that they will find either one of those extremes will find a way to use our report to their advantage and sight of the satellite question will be lost so I don't know how you can do this but it is a logical question to ask so I'm in favor of that I just don't want this report to fall into that trap so all I suggest is caution in not giving a lever here that somebody else can use to their advantage that we need to attend and that distracts from usefulness of our report but if we separate the satellite question as a main question and then we say for example we say due diligence with respect to the NRC report we need an initial cut on other materials that I think in some ways will deter the argument that you're predicting is becoming a focus because we're not going to focus on them maybe we put this other table in a discussion we can make a statement we're not trying to make a statement on other anti-androgens we're trying to make a statement about the relative risk of satellites and consider the exposure to other things that act by the same mode of action yeah I agree with that yeah I feel comfortable the databases for the skin of A with respect to anti-androgens in fact profile is very very poor anyway so do we want to just get rid of this family is that the same this time in the environment probably but okay well we can discuss that if we want to get rid of it later I'm not suggesting that you do I just suggest to be careful how we play it open I mean just one addition this blend which I really fully support also be justified in the context of the silver book in the sense of you the silver book and the silver book certain scenarios are suggested where the effect of a certain chemical is not only judged in isolation but in the context of background exposure to similarly acting chemicals and we can that's one way of stop okay feeding in the other chemicals so the first table could actually just be comparison of the two cases for the seven phthalates alone for the different data sets four cases using other yeah we have other cases the regulatory reference sources and the European TDIs but I think if you just kind of compare the values for the phthalates alone in the two cases I mean the meat of the distribution are very similar the tails may be somewhat off but the tails are so much above one I don't know how much above one you have to be reporting well it also impacts directly on no sorry cancellation well cancellation so I mean I guess you know we can look at the same kind of thing for the children but from this table I could just go down the smidge on that table you know I think we end up with a similar kind of conclusion so these are children from NHANES not the infants the infants are the swan data the children of the NHANES I think age six to eighteen which can use this so it's still on that same table but I think the conclusion is the same the distributions are very similar across the two cases we're looking at the seven phthalates which again I think is an important point that we may jump around a lot in terms of choosing what kind of references to look at but the meat of the distribution here doesn't change that much with the different cases and I think that's an important take on message can I add one thing to children and girls I mean if we include girls and you made a comment about that has it indexed distribution for girls slightly different than for boys there's a bit of a problem there because this is based on endpoints relevant only to boys so maybe we want to omit the girls this is very strange hearing me say this because I this is the first time in my life I say this but do you see my point I do the thing is do we have the right reference doses for children and infants our reference doses largely were based on reproductive or developmental things like fetal development right so once you have a one to three year old child are those the right reference doses no one knows no one knows but I wouldn't worry about this too much at this stage because you have shown that small variations in estimating reference doses don't really matter much because the numerator varies so much the numerator being deli attack and in the in the absence of any other data what else can we go by I think that is a it could be justified so those are things that we could discuss either the alternative would be to omit this for children totally which would be a shame I think I'm actually more interested in the infants than I am in the children really like that yeah or in infants so I think we agree that data from a pregnant woman is the most interesting one because it covers the window successfully and looking at the numbers comparing the infants and the children the numbers support each other so I think we can be there and there is no need to feel a lot more sorry so back to Andreas's point though, get me rid of that but we want to work together so we just don't work do you want to stratify well so let's look if you look further back I don't think I did a table I just did a little description so I'm jumping around quite a bit but the very last section of the paper I don't know the page number but it says analysis of swan and al data what I still want to add here is the pre and postnatal pregnant women so I haven't done that yet from the swan data table 8 of this it's the mono esters that were measured by swan and al and just to keep me straight I also put in the in Hanes variable name just to make sure I have the right things linked to the right things so we end up with 6 different phthalates instead of 7 here so those are the variables measured by the swan data and then if you go to figure 6 just below that so this is the age distribution of the children of the infants so you can see there's a 7 about the 1 year age about the 1 over 1 up in the 2s and 3s below that is a table that gives the percent above the limited detection for each of the phthalates does that mean anything does that mean anything above the limited detection but these mono esters were detected in most of these babies the fact that it was detected meaning that it's largely above the detection can be expected 20 that number doesn't it's not a table relative to other things in number 4 I think it's interesting to point out that these are chemicals that appear to gather they're not chemicals that generally are one at a time that can be easily stayed without putting a table because when you get a table it's misleading we would stay in an attack without using it yeah it is misleading let's just keep talking about that I don't I'm just saying it's misleading the table doesn't say that the kids had some measurable level above the limited detection because when it was measured it was between 1 and 2 fold what the limited detection was the percentage the percentage of the children that had a value detected above the limited detection it doesn't say what it is it just says that there's no meaning to that number that's the problem we can misread it percent of the children above the limited detection we just like this we're going to be all just like this and without some relevant it makes no sense it did probably a lot better gig in that study to quote for meaningful data so then in connection also with the without making my go all the way around maybe wants to but the distributions the babies are in the back too the daily intake estimates just this at the very back how many is the next change it has no index the daily intake estimates are in figures through A3 so again you get the sense of the distribution of I think that's the children go down lower so these are the estimated daily intake values six different diesters for the infants on the log scale and what we see again is that the DEHP I believe is the one with the highest mean and median values these are pretty low so and then if you go so I might all the way around so then back to figure 7 so I did a little analysis where I model the long of the hazard index as a function of baby's gender maternal age baby's race and study center these were across three or four different centers and girls were significantly higher there was an increase in the hazard index by baby age group so the 12 to 24 month old babies were significantly higher than those that were less than 12 P less than P.01 the babies who were 24 to 37 months old were significantly higher than the baseline in less than 12 months P less than .01 but there was not a difference in the age I and detected across study centers mother age or baby race so what you see so Mike's got it up in figure C there those are box plots for the different genders this is again on the log scale and the negative values so zero is actually one zero is one pretty what I expected the meat of the distribution the tail and then you see over in figure D is how it shifts up as the children or as the infants you know go from less than a year to two to three years old figure D so I believe this is so this is case one I'm not sure that results stay the same in case two that baby age group had a p-value 0.139 in case two so that age related result may be dependent on the DHP or something that is the highest percentage of the index in case one so I guess you know the thing is I think there's a lot of opportunity to look at this in different ways I think we can study things across different reference doses if we want to do some more modeling I'm happy to do that if there are particular questions we want to ask about the data like that is there a case look in looking at how this changes with different years and age data we have problems comparing them directly do we so I changing for the ages look at 2005-06 in subsequent surveys they've also measured all these times yeah so I said I would do that I would prefer not to go back and do everything again in another dataset if we had like certain distributions or whatever I could do that that was all like to do this sort of a quality you know like a validation step and I maybe would propose that we think about that for the pregnant women maybe focus on that just to see if we could reproduce a similar kind of distribution in another dataset it goes back to what Bernie said about the impact of the year 2006 date for the for the ban so it would be interesting if the percentages shifted but does that as a distribution look like for more recent years at those tables that had the percentages you know the average percentage I didn't do it for the infants I have a table 7 oh right I actually did compare I think this was to some of the intake estimates that you had in the Court of Camp and Fouse paper and that's not percentages that was actually just based on the distributions were pretty similar except for one of them and I can't remember which one it was do you remember that actually had a different profile in Europe somehow so the point was that you can pick up sort of different patterns of usage so it would be interesting to see if we did that and I guess in 2007 or 90 the ban was the ban was in 2009 was it yeah it was February on nine was the effective date and that only affects a small number of products and in a lot of the you know by them the phthalates I mean the use of the phthalates in those products has been going down over a 10 year period anyway so there wasn't a abrupt change in the crew well and I don't think the data available we could go to 2007 Andreas I think you will have the publication in mind we did in Germany with my medical specimen where you could see a face in and face out of some phthalates the problem with enhance is that over the years you had an increasing number of metabolites and it's very difficult to compare the years if you would want to observe trends it would be sufficient just to look at the metabolite level and then the do is that the explosion maybe five years ago to the HBSP in three times higher so you wouldn't have to have this complicated approach to have the knowledge that you observe some trend fair enough struggling with how to link the data from the infants more closely to I guess the overall concept I just have to think about it it sort of just sticks out there right now it doesn't fold in well because in the other stuff you're dealing with large distributions of data and you're dealing with quasi statistically distributed study and I have a 200 children you're trying to separate it down between males females I'm not sure that's it's worth it to do some modeling I don't know if it's modeling it's just I don't call it modeling, I don't call it data analysis we can debate that another day but the point is that basically you're interpreting the data and it stands out in the zone which is fine but I'm trying to fold it back to the rest of the analysis are you doing this as a case study that if we had more data of this type this is the kind of analysis that can be done or you're trying to make a record or you're trying to make a statement about children's hazard effects the infant's hazard effects with respect to values and studies and such to do that it's a good case study but I'm not sure where it fits how we can direct this report because how are we going to take that go backwards and say does it answer any of the questions we have because most of the kids are very low unless we're going to go and take it back to the exposure assessment we're going to do and see in fact we can relate the levels it has been that we've thoroughly been taken to calculate for us and related back to what we've seen some of the toy products and the consumer products that have been used by children at that age group well I think things in very well exist but I'm directly to the charge to mean it if you look at the the index distributions have very similar to the adults and in any case we are charged with coming up with a rigorous and disinterested analysis of the situation and children are just part of it it's directly written down in the study is what I look at a very solid study actually could be an application or for the whole concept of doing this has been that and then we're linking it to an exposure assessment based upon the children's related project products that we expect them to be receiving those first years of life we're using I think it's considered the last meeting that we were particularly interested in because it's the data it's the mother child pairs with the energy and distance measured so we agreed at that time that it would be a very interesting population to investigate so we I have no problem with that we all agree so that's the reason why I'm trying to think about is how to use it and how to use this analysis most effectively because in doing the exposure assessment generating scenarios okay now if we have this group of 233 children 3334 children is it wise to then in addition to the general analysis we want to do focus specifically on this age group then carve out the scenario for the most important in fact if we can delimit the analysis in some ways in fact if there's any plausible relationship between the deli intake we calculate and the deli intake we estimate for that subject well the other thing that we can do that we haven't done which I thought you were getting to is with the mother child pair we can actually see the women who have high exposures do their children also have high exposures I mean that would be interesting and that's work that I haven't completed so we do have the mothers oh that would be terrific we did that but even so that would really demonstrate clarity for this case done because then we could it's interesting because people are assuming that the youngest children have higher exposures and your exposure goes down as you get older and I don't know if this is statistically significant but that age trend is the opposite of that it's also the opposite of mouthing activity so it would be I probably you can't you can compare these data to the in-hates data but you can compare to the mothers that would be but you see that was my concern we're trying to go back to the in-hates data from this data right six years blinding two adults from the in-hates so we can actually use this as a defined case study for hazard index with having the mother's data you can answer the one question about the issue of the mother child relationship with the hazard index and then we can do the exposure assessment for that age group specifically and see in fact what would the influence would be a projected use of toys from shampoos in that age group and the values in the hazard index to calculate see I was worried about the fact that nothing didn't matter just trying to figure out how to make it plain correctly and see what we try to achieve that makes more sense yeah it makes more sense and does it make sense to you guys? yeah that's why we watch to have this population yeah I like that we suggest that we take ten minutes and then come back refresh for our discussion after that Bill? yeah it's time to take a ten minute break go to bed I need a party break go to bed hello we're back we're going to start with a question from Russ and then we're going to proceed to another discussion that I'll introduce to you the discussion from Russ's question okay Russ? sure so I had a question I guess for Chris Holder and the committee in terms of looking at the seven phthalates if we were to also consider an additional approach where it was done for instance just for the three phthalates on the interim ban and see what the distribution of the HII would be because you could work under the assumption that if the ban for the other three is permanent at least in children's toys what would be the contribution to the HII of the other three and it's just a thought it's not something that we need to do I just wanted to bring it forward I mentioned it to Bernie briefly it could be something we reject and don't do but I thought just to think about it I think we've seen from the previous discussions that the picture is already quite complicated so it should be really over complicated we have individual daily intake data we have it here we agreed that we might superimpose on the daily intake data the different TDIs and reference doses we have so we would see for each dilate individually how much of the TDI or reference dose is exceeded or how close the individuals are for the individual phthalates yes the reason the question arose is if you look at the table and you see the seven phthalates you don't really get a sense is DEHP contributing 70% to the HI and what about the other ones but it sounds like all of you said there will be information for potentially each one individually to look at how it contributes and then of course someone can do with that what they want I was just going to say that's what those tables of average percentages tell me so that's so that's like in case one it ends up that the DEHP is the large contributor to the hazard index like 80% on average is due to DEHP but if we change the reference doses down to smaller values for some of the other chemicals as in case two then it's an even spread but it's the other chemicals contribute more so these are the distributions like the figures so if you look at the corresponding table there on your left hand so those are tables that talk about so if you think about for each subject there is a hazard index so the question is what's the percentage of that hazard index per subject due to DEHP for each of the seven and so everybody is different because their exposure is different so what that table is giving you averages them across subjects so I think what you can interpret from that is what the major contributors are I agree if you look at table seven for instance you could all park what it would be for instance these two or three right but the contribution is smaller for some other chemicals but again that distribution that distribution is going to depend on what the reference doses are so we see more variation I think in those tables then we do about the distribution of the index itself I think do you know what I mean so in case two you know the reference doses are smaller than they are or lower than they are in case one for many of the chemicals handful of the chemicals and so if you compare table compare table three and table four table four, table three which is from case one says that the average percentage for DEHP is 82% but for in case two DEHP's percentage is 53% and DBP is 19% DBZP is 13% on average well I mean we could write about it more but I'm sure comments will need all we can expand this changes quite a bit in the case one what that's telling you is on average what the contribution of each chemical is to the hazard index the distribution of the hazard index okay Russ then what do we do with the rest of the afternoon there's a question that some of us discussed over the break and part of it followed a concern that I expressed that I was still having a difficulty going from hazard identification to risk assessment and Andrea clarified for me that we've been talking about risk assessment for the last two hours which I underappreciated but now I understand that the term hazard hazard index is really a measure of risk and we've been using it as you have been using the information of this in the context of biomonitoring data and the discussion with all about the exposure data will allow us to do a parallel risk assessment on a different set of numbers a different type of exposure now we can compare the two and decide how we're going to work between these two so if I understand that part of it correctly what I would recommend for the rest of this afternoon I may be the only one who didn't understand that well but I think it would be good to help go through that with everybody so that we're all on the same page as it relates information and the tool and how we're going to match that up with the other exposure data and risk assessment to set the stage for a more in-depth discussion tomorrow about risk assessment because even when we have that information in mind from these two methods of providing some quantitative evaluation of relative risks and this can be populations by phthalates plus something else but a number of the factors that we've been talking about from the beginning we as a group have to go up to some criteria for how are we going to decide what is sufficient risk to support a recommendation and I don't think we've talked about that yet but I think it's time that we do that tomorrow because by the next meeting we're going to be needy than that so I this is a topic that sometimes takes groups to get their hands around and I'm not saying that we're going to be one of those but I think we need to start and so that's what I propose tomorrow is that we have more discussion about how we're really going to array this family of risks that we're going to have in front of us the decision that this is suggestive of supporting a band this is supporting the lifting a band this is supporting a brand new band this is supportive of doing nothing because we have to translate that array of risks into those categories and we can do that tomorrow so I'm curious would you lead us in the discussion piece of the kind that we had during the break to open the discussion for the setting stage because really it was successful yeah in the break we did this with the 8th whiteboard piece of pen oh that's a flip chart should we do that or we can use the board we can use the board to flip chart we'll do I think can I just get the pen I've got some here so for a single chemical risk assessment the simplest form is done by dividing daily intake by a reference dose that can be acceptable their intake tolerable whatever and the demand is that this quotient should be smaller or equal to 1 and if that's fulfilled the attitude of risk successes is okay to answer if this quotient exceeds 1 then usually a discussion begins with risk managers about risk management exercises so that's the single chemical okay and the the hazard index is probably really a misnomer in a way because all it is all it is is really making the demand that this simple equation here for single chemicals is also fulfilled for mixture so we are now moving from so imagine this so in the language of specialist mixture toxicology this is called a hazard quotient this is not very good language but that's what it's called so the demand is now that the hazard index is the sum of hazard quotient and the demand is that this sum is still smaller than or equal to 1 so really we would have this here's the case for three chemicals this can this could be done for as many chemicals as you like but the key point is that the sum of these terms here in analogy to here is still required to be smaller than or equal to 1 so we've seen from St. Horger's analysis that under certain circumstances and for a certain number of people that demand is not fulfilled anymore it depends, we've seen the numbers so the other thing to note is that this equation here may wonder why it comes from is nothing but an application of the concept of dose addition of that's an assessment concept in mixture toxicology and it is only justified in cases where chemicals work together along with the principles of dose addition and so what we will have to do in the hazard assessment part of this report is to see where that actually is backed up by scientific evidence but really that's it so really we have been as Bernie said at risk for the last two hours one other thing how then Paul's work would link in here we've seen that Horger and Chris have derived these daily intakes through biomonitoring and back calculation taking into account the toxicogenetics and what Paul will do is a calculation of this more according to first principles so let's call it Paul first principles yeah so what this will enable us to compare the two cases so the estimates for the intakes derived by toxicogenetics and biomonitoring and derived by let's call it external markers external markers yeah so that's where this is headed I think right that's so the other thing besides dose addition that we're assuming there is that we have co-occurrence to do the combination which I think is clear from the biomonitoring data but maybe we need to look at some correlations across daily intakes or something just to show that the question now is what do we do when we find in these exercises that the number one is exceeded for a certain fraction of the population but that's right okay this exercise answers directly to a couple of items on in the charge in the law section 108 but it goes back to the point I made earlier my concern is that if we make the A recommendation for or against continuation of a ban dependent on risk assessment of this kind we will then open the door and invite a substitution process so for example if the answer is that for DHP these quotients will approach one or exceed one and on that basis we say a ban 48 I'll make just an example that the ban should be continued that means that DHP will then be substituted by another ballet which then means that the same game begins again and then so another ballet it could in principle be a dangerous one which is currently below one very far below one according to this risk assessment all of a sudden it's used more as a consequence exposure goes up and then after a time this quotient approaches one or may exceed one even so what I'm trying to say here is that in my opinion any recommendation or decision about ban or otherwise should be based on hazard not risk assessment of this style however we have to carry out a risk assessment of this style in order to answer various items in the charge to us I agree with everything that came through your last statement then why do we bother to do anything because if it's based upon a hazard assessment then we can just ban everything now the question is is that what happens if let's say D1 which is or D2, D3 or something that's in the purview of CPSC are de minimis and the only thing that comes up large is something we have no control over alright so there lies a conundrum for us and it can't just be based on hazard because this stuff is out there in the world it could be food, it could be in all kinds of other other sources or locations so I'm not sure I agree with everything up to the last statement I fully agree that I think we have to discuss this a little bit more before but that's why I said it we need to discuss that because there are a lot of permutations which may in fact lead us down another road well it hinges around the term how we interpret in number one and two examining all of the potential health effects that's open to debate I think but it's some people might interpret it such that this may include risk assessment as well as hazard assessment I don't want to come back to your point as well if we knew the amount of hazard of lethallis and substitutes and we had to handle on that we discussed this so we really made them from greatest type of hazard to minimal hazard we would in order to know what to recommend we still need to know the exposures to say that with that degree with that type of hazard and this kind of exposure we would make a recommendation to restrict exposure in some way and or a number wouldn't we well here number number three says we're we have to examine the likely levels of children, pregnant women and others exposure to thousands we have to look at so once you have the hazard characterization and you have the exposure then the only thing that remains is to do the risk assessment I think taking the biomonitoring approach and the exposure assessment approach and using it with respect to the RFP gives you a chance to do the local risk assessment and do it in a way that you're comfortable and come up with a credible result rather than waiting one versus the other we can discuss and analyze the information to be able to determine whether or not what was it number three we were talking about the thing about the minimizing the exposure can be best handled in one way or another you know it's spanning it or not I'm not even going to have to worry about it or it's so de minimis that there are other areas that people are better well spent on it I would argue that the phrase here consider potential health effects means risk assessment I agree totally agree because what else would it be it couldn't just be hazardous it's just toxicological yeah so we have to do that but these numbers here number one through to number eight do not help us wish that would have been the case but the way this is phrased doesn't help us to define criteria for saying this button should continue the or not or otherwise that is the problem so we have to make them up ourselves I'm sorry about this but this is the way it is it's been done many times before by other groups but that is what I see our task for tomorrow to come as close to that as we can I agree totally agree see if we're back on the same page early? I have a question that you may answer tomorrow morning but I it may be at 5.30 my time tomorrow morning so I don't know what that will be either waiting or able to understand the time of the day but I have a question about the interpretation of the hazard index we now have done it and we have information and if I read the similar book if the hazard index is one or above there's it's likely that there's some risk but there's not much risk either in terms of what's the difference between a hazard index of 2 and 4 is it twice as much risk? no if the book says no alright then how are we going to take the information that we generated to make decisions about whether the exposures to valley individually or in combination depending on which kind of analysis you do is sufficient to warrant one actual or another I'm not clear how we do that Phil this would be the same as for any single chemical so if you find out the quotient of the intake of a new reference dose exceeds one then with single chemical everyone knows what to do so it's just if it's a bad one it's a bad actor yeah okay and it could be a bad actor either because of exposure or hazard it doesn't necessarily one or the other it's the combination of the two that can lead to that level of concern so we don't have to take into account for example on the 129 pregnant women if there are only you know 2% of those women that exceed has an index of one that doesn't make any difference as opposed to 50% well this now is a question of this is a question of the law I'm not familiar enough with US law but in Europe there is an entire element that you want to protect 100% of the population well in the US it's not to increase excess risk by one in a million and I forgot what the hazard index is okay well there's some interpretation it's a hazard index over one but the question is do you it's not a bright light the question is do you are you concerned if the hazard index is a little bit over is it the mean or exposure or the 95th percentile there's no hard and fast answers that's where it gets squirrelly how many are you dealing with the 90th percentile if it gets a little bit hard is it more difficult with the hazard index well the silver book the new philosophy there says bright red lines have to be regarded with caution in the context of the decisions that are to be made well that's what we're discussing tomorrow morning I guess whether we want to hang any decision on this we have to carry out my interpretation of the charge here would be that we would have to carry out a risk assessment and this hazard index calculation would be in response to this but it could just sit there we could say yeah we've done it to the best of our abilities and to the best of what science currently can offer but it's a second question as to whether we want to make decisions on balance dependent on hazard index exceeding one or not to otherwise and I've said what my opinion is and that's what we need to discuss yeah yeah yep you decided on a specific example of a hazard index of two versus four is the answer the same for two versus twenty the nominal exceeding of one does not give you any handle it's not really a risk optimization it doesn't give you a handle of making statements as to the likelihood of her effects occurring otherwise so then are you distinguished between point nine and two you can't that is why it is important to take into account other factors that you can't quantify as we've discussed today say other background exposures to similarly acting chemicals etc etc that can then inform and illuminate your risk management decisions always customer assessment outcome you may think though look at it as the uncertainty in the reference dose itself the reference dose is you have say no effect level or some 5% a low effect level 5% or something in animals you divide by factor 10 in cases where you're more sensitive another factor of 10 because one individual may be more susceptible than another or sensitive than another so if you have a hazard index of 10 you've lost one of those uncertainty factors essentially so that might be one way to look at it but it's not a risk per se it's not really a risk in the sense that it's not a probability like people do for cancer risk assessments right that puts more partial address to come up with another chapter to talk about uncertainties because if you were at a hazard index of 0.9 and there was the underlying information from the toxicology was a very robust strong study about which we had a very high level of certainty that would sway you differently that was based on a study that was marginal and repeated the study may end up with a quite different answer so you have the latitude to take down but you have that anyway yes it has nothing to do with the question you have to ask when you stylize as a criteria by which you raise your level of concern you have to have a benchmark or a series of benchmarks or saying well it's an issue it's an issue I have to be concerned about it's an issue if I don't pay attention to I should hang up my hat and go home and do some garden because I don't understand the process and I don't understand the problem anymore as we go through this process there may be let's say there are 10 things that create some uncertainty I think we need to articulate what those are and how we are about the uncertainty and what do we do with it well if we take just take the the hazard index that Chris and others have produced so far we can't make a statement right you have exactly what I would anticipate a few people that have a hazard index above one maybe it's two to five percent of the people it would be logical and you have a small percentage of between you know 15 percent you know about 15 percent or 85 percent of the number of one because then the rest are pretty low the question that comes when you get up here and start summing all these different hazards together where does it all fit does the hazard associated with food ingestion in different populations for these people to 95 percentile far away any of the risk that would be associated with consumer products that have valence in it like hairspray or shampoo that's where it becomes really naughty because some of these issues may be background as was just stated and many of these people are in the 95 percentile just have values rather than being exposed to shampoo or teething or teething that's valence and the question becomes really one of management of what that risk means in terms of the questions that we have to ask ourselves in the next few months other thoughts to set the stage for tomorrow Bill yeah anything dad no nothing dad um give me a comment by burning away money look you're a brave man Bill make sure you have coffee I know is there anything that we is there a I'd be first to admit I have never structured or managed a discussion with the answer to this question so from those of you who have been in those discussions more than some of the rest of us how would you recommend we should start tomorrow so that we don't just get married in this long period of time I think we should begin to discuss criteria that would lead us to say ban this, don't ban that I agree with that totally agree with that because that leads to our charge correct and it actually makes our job simpler than we do that go back to what's the tab number 10 and go to the second page of that and this is why it's a number 7 what are the options for recommendations to chat it's the second page of the document that I taught in the process for chat determinations it's just a two page or a two page yeah so I did this to skip ahead to the end and try to identify what are the options that you really have for making some recommendations and what would be to restrict the exposure to a specific catalog that's not a way to ban it so that it's not in removal or it's not banned but we decided that they would recommend that it be restricted in some way and that could be a ban or a ban but secondly we can also recommend based on this overall analysis to remove the ban on a chemical disorder ban or recommend a temporary restrict exposure if it isn't already restricted recommend to remove the temporary restriction if there is one we can recommend to collect more data on the exposure of toxicity or more of action we can recommend to retain the bans that are currently in based on sex countless but we can take no action so those are those are the options as I saw them so whatever decisions we whatever recommendations we want to make have to be addressed to one of those so you're going back to where you started you said to try to identify that threshold for information about which we would recommend that something be banned that would be one threshold we're looking for not a threshold in other sense it's not a bright line it's just a distinction between are sufficient to recommend this is what we would ban something like that with some level of uncertainty greater than what we had to ban would be to temporarily ban so the temporarily means that we're not sure and it might be resolved with collection of other data I would think that a temporary ban would be linked to the recommendation for data not just that we couldn't decide temporary and we can recommend that other data be collected but then we need I think we would like to say why to my hand rather than just say we need more data on this we need to say if we have this is what we would like to do and if we decide to take no action then those are the chemicals that are called below these thresholds for either ban or error ban and below that other below that below those two other decision points we have to be able to justify to ourselves why we support taking no action that engages the criteria that we had for a temporary ban something is different between us two we have to be able to articulate what that difference is does that follow based on what you said we start to talk about something good questions I think the questions that you posed under items 5 and 6 of the process for chat determinations some of them I think are very germane in the discussion that needs to occur tomorrow okay we will go back to numbers 5 and 6 what are the risks for humans and points of concern vulnerable populations what are the sources of exposure that are of concern and I almost put down there to evaluate the results of the hazard identification the hazard identification will do more than that that's already a risk determination yeah it's a risk it's a risk yeah and then in 6 in what circumstances are levels of human exposure high enough to be of health concerns again what are the sources what are the chemicals what are the populations so did the answers to those questions help define what we would ban and what we wouldn't I think they would in the sense that of course you wouldn't ban what about anything for example no human exposure that would be silly so exposure assessment informs informs this decision making process very much and likewise I would argue that you wouldn't want to ban anything or any chemical where there isn't some degree of concern about health effects in the general population so that can be informed by some form of risk assessment or definitely hazard assessment but okay I deliberately said informs decisions but that's not helpful as such because still we would need some hard and good criteria on which to base these recommendations for banning something more or otherwise at number 6 I would change at least one of those phrases to risk to talk about high exposure, hazards the word risk is the challenge should be to determine a combination that leads to risks well as our consensus our levels of risk and human exposure human exposure doesn't have to lead to a risk of concern I think we get to pull up on the path and it loses its credibility when we're talking about risk assessment and the base lines of the equation I guess you said something that I want to follow up on is this kind of a suppose there is no more human exposure but if we have a chemical that presumably would be suitable as a substitute it would be a replacement for one that is being used if there's an action taken to get rid of it or to minimize its use and there's another valid out here that we know is five, six times more potent than the ones that we're worried about today but there isn't any known exposure it could be geared up for manufacturing fairly easily and in a year's time we'd have a high level of exposure to that one and we were silent on it yeah you're right that this would be so if we know of some kind of of that kind I think we have an obligation to make a statement and even though we don't know what the risk is because we don't have exposure to it based on the toxicity at all or some other factor we would be concerned about this one being put into commerce so basically in a situation where there is no exposure our contenders would have to be raised and are concerned that we're reunited if in fact they want to put into the project that's where there will be exposure because it isn't in commerce there's no demand but who would like to precaution against it being brought into commerce it's risky to do that because we were charged and we don't know what else is out there if we fall into that category one has come to our attention and we don't know how many others are out there but perhaps one is an example is enough okay maybe that's the base word yeah 830 tomorrow anybody object if we adjourn right now we'll call you when the pocket is ready to learn very good have a good dinner bye bye