Transcript for Dr Raj Bhula, APVMA

International developments in dietary exposure assessment—pesticides and revision of the IESTI formula

This presentation was delivered at the APVMA’s science feature session on 15 October 2015. The full video is available on our YouTube channel.

I'm going to talk about the ISD methodology and I'll explain what that is in a minute. Janis has very beautifully taken you through the health‑based guidance values. She has already given you an explanation about the acute or the short‑term exposure assessments that we do. I'll just launch straight into the short‑term exposure and explain what's going on internationally, particularly coming from Europe and FSA in terms of the changes to the risk assessment methodology.

First of all I will talk briefly about the exposure methodology, what's involved, what are the parameters that are included in the way that we do our short‑term risk assessments. I will also take you through the equations that we currently use and have been using for about 15 years now, and what's next in terms of some of these proposed revisions, and what we should consider as the Australian regulator together with Food Standards in terms of where this might head internationally.

Janis has already taken you through the acute reference dose and the short‑term exposure and why we do it. Basically what we're looking for is that hit, that single hit or dose of the pesticide residue that somebody could be exposed to as a result of the residues in the food that they consume over that 24‑hour period or within a single meal. We do this in two ways. We look at the exposure to children because as you know, children will eat a lot of particular types of food. They may not have the broad range of diet that adults do, but on a body weight basis, the dose that they are exposed to from pesticide residues in particular foods is going to be higher than the dose that any of you are exposed to consuming that same amount of food, so we do the risk assessment in two ways.

In terms of the methodology, the IESTI is the International Estimate of Short‑Term Intake and at a national level we do what's called the NESTI, the National Estimate of Short‑Term Intake. I think somewhere in the audience we actually have Steve Crossley who put together the IESTI. No? He's gone. He is responsible for developing that acronym so you can blame him later if you can find him. With the NESTI's genesis already explained we look at our own food consumption data for various groups. She has taken you through the food surveys and all of the information and assumptions that go behind providing those figures to us when we do our risk assessments.

The other thing I will talk to you about is about unit weight. A unit being a particular commodity and what that means in terms of the risk assessment that we do.

Here is just more a diagrammatic representation of what Janis has already talked about. We use the health‑based guidance values that come out of the toxicology assessment. That's the acute reference dose. At the APVMA we look at the residue concentration. In Jason's team the residue assessors will look through the residue field trials that are provided as part of a data package from a chemical manufacturer and they will have a look at what the decline values are over time and the residue concentrations in different commodities.

Food consumption information comes from FSANZ as Janis has already explained and for the short-term exposure assessment we use that 97.5th percentile consumption so it's a high‑end consumption. Then all together these components form the dietary exposure and as I said to you earlier, we can look at it in different ways depending on the types of food consumption data that FSANZ is also able to give us and break up in terms of sub‑populations. We look at children particularly because that is a very special and a sub‑population that we do need to be protective of, but we also look at the general population. Depending on the endpoint in the acute reference dose, we can break that up into different sub‑populations if that's what we want to do.


When you spray an orchard, when the sprayer goes through you can imagine that the apples on the outside of the tree are going to receive a higher amount of spray than the apples on the inside that are protected in here. Each single unit, each apple on that tree is going to receive a different amount of spray, which means that each apple can have a different amount of residue. How do we factor that in to the risk assessment when we're setting the MRLs? This I'll talk to you about in a minute.

The exposure methodology has been around for a little while. For short‑term intake it was developed between 1997 and 1999, and this is once JMPR started to routinely establish acute reference doses. It was internationally adopted through Codex and that's when Australia picked it up, around 2000. At the time, the Office of Chemical Safety which was in TGA, the APVMA and Food Standards sat down, looked at the international methodology in terms of best practice and how we would adopt that in Australia. That was the time at which we then started to make routine decisions on acute dietary risk assessments using the acute reference doses that had been set by the OCS.

The equations

There are three different cases or equations that apply and that depends on the type of the commodity and it also accounts for the way that commodities are sampled when you do your field trial testing, the way that the samples are collected in the field for residue testing. That's basically how the cases are put together. When you think about apples, our sampling guidelines say that you should collect a sample which contains about 12 fruit or 2kg, and you take the higher of that. For each commodity that we want to set an MRL for, we have a sampling guideline or sampling protocol. What that takes into account is only a composite sample which is a sample of a number of different units. So how do we go about determining what is the residue in a particular apple if our sample is taken as a composite for setting MRLs? That's a trick that you need to think about.

The different input parameters are there. As Janis explained before, with our short‑term intake assessment we used the highest residue or we use the supervised mean residue, supervised trial median residue to look at process commodities.  

These are the input parameters. We have:

LP           the large portion which again we derive from the 97.5th percentile consumption data

bw           the body weight in terms of the body weight of the individual that ate that food or that commodity

HR           the highest residue within this field trial data that we've received

HR‑P      and, if there's a processing study or the food is eaten in a processed form, it's the highest residue in the processed form

u             the unit weight of that edible portion or that particular fruit or vegetable, and

v              the variability factor, and I'll talk to you about the variability factor in a minute.

Unit weight

Those of you that are avid gardeners, you will know that you can grow a myriad of different types of tomatoes in your garden. If you think about the large beefsteak tomatoes here, you might not eat a whole tomato or your kids may not eat a whole tomato, but these ones here, the little cherry tomatoes, you might eat a whole punnet of those in one go. There's different consumption depending on the size of the unit that you're looking at. As agriculture develops and there are more and more varieties that are being grown out there by producers, the unit weight can change for individual commodities. Even though we are looking to set an MRL on a tomato, we need to go back and have a look at, what was the variety of the tomato used in that residue trial and is it representative of all the other different forms of tomato that you can grow from a consumer exposure perspective?

My favourite, eggplant. There are a myriad of different sizes of eggplant as well. Again, it's the same issue. People may not consume all the different types of eggplant but for different subpopulations, and you think about Thai curries that have that pea eggplant in them, they may have a higher consumption of that form of eggplant rather than eating the large ones which may be more from a Mediterranean diet. That sort of thing has to be factored into our exposure determination.

Case 1

This is the simplest. This is where what somebody might eat is the same as the residue trial sampling in the field. That's the equation there that takes into account the large portion and the highest residue and body weight. It's very simple.

Here's an example of a Case 1 commodity, berry fruit. For example, a high consumption person might sit and eat 250, 500 gram of fruit in one go or over the course of a day, that's the same sample that we take for residue testing and the sample taken from the field.

Case 2a

This is where the size of the commodity is such that one person may eat more than one of these within a day. Now, if you think about summer and mangoes, I myself might eat three mangoes in the course of one day. So with one unit is going to be smaller than the large portion that somebody might eat and that's the formula that applies to that. This is where the variability factor comes into play. If you think about the apple picture that I showed you earlier, each apple might have a different residue in it. To account for that a variability factor of three was set based on hundreds and hundreds of samples being taken in the field and analysed in a composite fashion as compared to the residue in a single unit and that's the variability factor that is now routinely used in our assessments.

This equation here also accounts for the unit weight, so as I said to you earlier, it takes into account the size of the unit as well as the residue in that unit. There's the example of Case 2a apples. Somebody might eat more than a unit, they might eat two or three apples within one go, that's what that equation accounts for.

Case 2b

Case 2b is a situation where the size of the commodity is too large for somebody to eat in one go. Here's a really good example. I don't think anyone is going to sit there and eat one cabbage in one go or one cabbage through the course of the day. You might feel quite ill if you did. That equation there doesn't have the variability factor because you're not going to eat more than one unit within the day.

Case 3

Case 3 is one where we talk about bulked and blended commodities and this typically applies to think like milk, cereal grains, pulses and oilseeds which tend to be processed before they're actually consumed. Again, cereal grains and oilseeds. it's a bulk commodity which means that when the commodity reaches the point of sale it's not necessarily going to come from one farm or one place. Particularly with cereal grains, if they go through a bulk handling facility there will be different parcels of grain going through the bulk handler before it then ends up at the flour miller or at the bakery to be turned into bread. With the residue data that we receive, we do get processing information where you would look at the transformation of the residue from the whole commodity through to flour, through to bread. That helps us refine our exposure assessment.

The proposed revisions

What's going on internationally? There was an international workshop held in Geneva last month and I was lucky enough to attend that workshop as an invited person. There is a report up on the EFSA website if you're interested in looking at what the proposed revisions are. The outcomes of that meeting are yet to be published and that won't be done until FAO and both EFSA have endorsed that meeting report. They are anticipating that it should come out by the end of this month.

What was the purpose of having the workshop? Europe had been criticised for a long time in not using the same variability factor that just about every other regulator and country around the world was using. They were more conservative in applying a variability factor of seven to some of the Case 2a or all of the Case 2a commodities which meant that their exposure assessments were out of line with everybody else's. There was a direction given through the European Commission that they should be looking to harmonise the methodology internationally. It was an opportunity to come back and have a look at the equations and see what's changed back when they were established back in 1999 to now, 2015.

The only real change that there has been is the introduction for those of you that are familiar with the OECD MRL calculator. The introduction of the OECD calculator which basically gives you a single output and where we're now involved in global joint reviews and we do use a lot of international data, we will use the calculator which gives you a consistent output no matter where you are doing your assessment, be that here or Canada or the US, providing there is a comparable GAP association with that particular commodity and chemical.

The main reason for the workshop was to have a look at replacing the high residue and the super trial median by the MRL. Now you can imagine that is going to make things quite conservative in some of the cases that we've already looked at but the approach is more to look at it in terms of risk communication so that a regulator or a risk manager could put their hand on their heart and say, "I know that MRL is safe because we've done the exposure assessment using that value rather than a higher residue."

The other outcome of the workshop was to remove the unit parameter in the Case 2a equation. Now, what that does is it is going to make Case 2a very, very conservative indeed because you're now not taking into account the size of the unit and the portion that the individual consumer is eating.

The other outcome was that Case 1 and Case 3 should be the same. That is maybe what is not going to make a large deal of difference but it will make Case 3 conservative in some instances where you don't have a good processing data package associated with your residue trial data.

These are what the revised equations or the proposed revised equations look like. Case 1 and 3 have now become the same and the MRL has been factored in where there was a high residue or an STMR. For Case 3, which applies to processed commodities, this is going to make this very conservative indeed. As Janis said before, we already have quite a high degree of conservatism in the current equations because it must be conservative in terms of consumer safety. But we already have that there so what is this now doing or what are the proposed revisions going to be doing in terms of the methodology we are currently using? The changes in terms of Case 2a and 2b you can see there. The unit weight information has been taken out. The question that we have to ask ourselves as regulators and also industry is, is this going to make our short‑term exposure assessment much more conservative and, if it is, how much more conservative?

What's next?

Once the report is published there is going to be a discussion at the Codex Committee on Pesticide Residues next year. This is an opportunity for national governments as well as chemical industry to then go and provide their opinions and their input in terms of whether the revision should be heading in the direction that FSA is suggesting. What FSANZ and ourselves will be doing is a head‑to‑head comparison of the current methodology against the proposed equations to examine the resulting exposure estimates and the differences.

That's with the intention of preparing a draft paper to take to the next Codex committee meeting. You can imagine if you try to harmonise this across different countries and you're using the GEMS/Food information which Janis talked about earlier, things are going to become very conservative which then means that if you're looking at MRLs for trade, this could get quite difficult in terms of the number of uses and commodities that will then pass that test to then be able to establish an MRL for trade purposes. From that perspective we're looking for industry assistance. Any of you that are chemical manufacturers that either have an application in the system with a residue component or are thinking about putting in a proposal and you would be happy for us to run our calculations head‑to‑head using your chemical compound, talk to myself or Jason during the day and we will be here tomorrow as well, because we would like to provide a very robust analysis for the committee to consider in terms of what the changes are with the methodology and whether it is more conservative, and if it is more conservative, does it necessarily have to be that way?

Janis already spoke to you about our dietary intake data. We are very, very lucky here in Australia that we do have a very good set of dietary intake values. We are also very lucky in that we do have good unit weight data and we are looking to update our unit weight information that we currently use in our risk assessments so it is more contemporary in terms of the sizes and the varieties of fruit and vegetables that are grown here.

My last message is this could be quite difficult in terms of trade if the methodology is adopted internationally but we need to do an analysis to make a determination of how conservative these estimates are going and why they are being driven in this direction by the European Union.

Thank you.

Errors and omissions excepted; check against delivery. 

To protect your privacy, please do not include contact information in your feedback. If you would like a response, please contact us.