Panelist Comments: Question Three
DR. YOUNGMAN: Thank you very much.
I guess if we can switch now to question three, which has to do with how data are reported. Is reporting sufficient to be able to use data from the three arms effectively? How would you change the report so we could use the data more interchangeably? What should be retained and/or omitted? And if we can start with you, Lyle.
DR. VOGEL: I think we should go a little more basic before we talk about the reporting, specifically. I have concerns that the whole program needs to be improved for public health surveillance. You know, as we talked about before, I am not sure that the slaughter samples are really reflective of what resistance on the farm, and I am not sure if the retail samples are reflective of on-farm, or the slaughter.
So, consequently, I am not sure that the sampling from humans is reflective at all of what is going on on the farm. So with that caveat, and we have discussed some of the ways to change that to make the NARMS more effective for public health surveillance, focus just on the reporting.
And I think we all agree that the current reporting via the three separate arms is not effective. We need to have a consolidated report where all arms are reporting the same thing at the same time in sufficient detail that the readers can understand the correlations, if there are any.
We need a summary report that reports all three arms; the human, the retail, and the animal. And on the animal side, we need the detail down to the separate species of animals and the separate sources of the samples for those animals. And on the retail side, we also need a detail as to the species of animals rather than, for example, consolidating all Salmonella together from all sources.
I am very thankful and grateful for the information that we have been given about the proposed changes, or the contemplated changes in the reports. And I think that is going to be a big improvement as that goes forward. I think one of the things that was mentioned that is under consideration, but I guess not a determination yet, is an interactive database. I want to emphasize to please give great consideration to that. A database that is available to stakeholders to gather the information that they need.
The other improvement, and we need to compliment all arms, CDC, USDA, and FDA in their efforts to make the reports more timely with just a six-month leg from yearly reporting. That that be a great step forward.
What should be retained and/or omitted in future reports? Well, I guess, in an ideal world, we would retain everything we have got now and then add other things. But I guess the only caveat I would put on there about retaining is sometimes in the past in the narratives of the reports, the authors of the reports have started putting in interpretations in there, and evaluations of the data. And I would prefer that just the data be presented and let the reader make their own interpretations.
DR. YOUNGMAN: Thank you very much for your comments. Awa.
DR. AIDARA-KANE: Thank you. So I think that there is a need to harmonize between the three arms in terms of availability of the reports. And also, in terms of reporting system and reporting format.
At retail, I think that having the species specified can be useful. And also, I think that it is important to consider having an integrated surveillance report. And to publish a collaborative report and use it -- this can be used to advocate for prudent use of antimicrobials in food animals. I think that it is important to have an overall picture of what is going on at farm level in human and animals to advocate for this.
And maybe in order to have this done more easily, it could be interesting to consider having a NARMS coordinator nominated by the three agencies to work on this issue.
DR. YOUNGMAN: Thank you very much for your comments. Scott.
DR. MCEWEN: I liked some of the things that I heard about the improvements that are expected in the NARMS Report. I think now, we currently have reports on the parts, the components of NARMS, but I think it is bigger than just the sum of the parts. I think you should shoot for a much more integrated consolidated report that covers the various branches of NARMS.
So, some of the words that came to mind would be consolidate, integrate, synthesize. I think we do need to have some interpretive summaries there. I think the part of surveillance is analysis and interpretation of data. So I think in the past, we have kind of thrown out lots of data, but the missing bit, I think, have been attempts to analyze it, interpret it, in an appropriate manner and using appropriate techniques involving, again, more sophisticated statistical analysis and trend analysis where possible.
I think that there should be a report which is readable, more narrative. As many of these comments we have made to the CDC -- review last summer, I think they are appropriate here. It should be more readable and sort of -- I think it would have much more of an impact in delivering on the goals and objectives of NARMS and reaching out to the users of the information.
I think you should consider subjecting the reports to peer review. I know that would slow things down. We don’t want to do that, we want to maintain timeliness. But sort of a short turn-around review, I think, would be helpful. I think when we are all preparing our papers and so on, if you know it is going to be critiqued by someone else, that helps you sharpen the things you say.
I agree with the comments that the availability of a searchable database for stakeholders would be a major plus.
DR. YOUNGMAN: Thank you very much. Okay, Marissa.
DR. MILLER: In a large part, I agree, again, with all the other excellent comments. To have an integrative report, I think it is critical. I think it was CDC that stated that they have a six-month goal. A goal to have the data out within six months. I think, you know, that should be a goal for all the arms in this integrated report.
I think it would be helpful to have an executive summary that is released as soon as possible, that can stand alone. And, again, that is for policymakers and others that may want to have a sense of what the findings are.
I wonder if it might be helpful to go back and look at DANMAP and what Denmark is doing as far as their reporting. Because in my mind, they have the most functional system in terms of enlightening public health and policy activities. So there might be some lessons to learn from how they do their reporting and generating of data. I like the enhancements that Paula spoke to about the trend analysis and web-based queries. I think if you can move in that direction, that would certainly increase the usefulness of the data to both folks inside the government and outside.
DR. YOUNGMAN: Thank you very much. Sue.
DR. KOTARSKI: I guess I will just add to the comments that were already provided by the reviewers. That it is a great body of information to summarize, so the aspect of trying to provide one report for all the information at the same time would be an awesome goal. Everybody wants it, I guess, but that is something for the future. I think that would be a difficult one.
In any event, I also like the aspect of seeing the MIC distributions. Those summaries come across. I think that is very useful because as breakpoints change by NCCLS standards, or you are using different screening mechanisms to detect a change in susceptibility, those distributions are very helpful and they last the long-term. So I commend you for putting them in when you can.
I like the idea of seeing the patterns of resistance if you are seeing -- to have that listing. I think it is very useful for your audience as well.
More context regarding the data. For example, the sampling strategy. How the sampling strategy changed from one year to the next. The limitations and the biases in the data. That stated explicitly helps those that are trying to use the interactive database, if you will, for their own purposes.
At least you have there in writing what those biases are, how those biases have changed, and it allows you public statement as to the extent to which you can examine temporal trends. Because, heretofore, in many of the cases of the different aspects of the NARMS, the sampling strategy has changed as just as a process of developing the system.
So maybe in the future, it will be systematic for the long-term and that won’t be necessary. But at this point, that would be useful for the reader.
Another aspect that comes up is the updates. As best as we can QC all of our results, we do find errors. And when we find those errors, we provide those updates. It seems that the CDC updates it in terms of the latest report provides the most accurate set of data. For the USDA, the NARMS, the USDA reports there was a period in which the website said, we have changed some of the database, watch out for “X,Y,Z.”
Either one can be effective but, explicitly, I would like to see on the websites, and in the reports, how you update it so that the reader can use the most accurate information and knows where to find the latest and greatest versions. Because there will be errors, and there will continue to be errors, it is the nature of the beast.
And, I guess, I would add too, just to emphasize, that some information about the nature of the sampling would be very useful in the text, and the summaries as well.
DR. YOUNGMAN: Thank you very much. Sean, your comments.
DR. ALTEKRUSE: The three arms of NARMS, they meet on an ongoing basis to discuss the reporting process. And I think that is really helpful. If they can use that to identify some consensus issues that they want to address, it is really helpful.
In particular, I agree with what Lyle suggested, and that is that by serotype information is valuable. Because there are thousands of serotypes and they have unique niches and so, in particular, the serotypes that I would recommend emphasizing are those top ten serotypes in people.
They are, actually, the top 20 serotypes in people, but overlaid with the ones that are found in meat and poultry, you come up with a list of about 10. That is Typhimurium, Enteradidus, Newport, Heidleburg, St. Paul, Monovadao, Thompson, Agona, Hadar and Redding.
There was a chart in Elvira’s presentation that I thought was really helpful. It showed the human isolates and their MICs by serotype as sort of a referent group. And then it showed the meat from retail in that serotype. And then, again, the meat at slaughter for that serotype.
So if that could be repeated, then you would have serotype-specific information and you can sort of see if the human isolates look similar in terms of their resistance patterns to what is being found in each commodity group.
Also, one last thing is that I think in addition to collapsing the HACCP isolates by what is called slaughter class, it would be helpful to expand it out so that ground beef and young cattle and older cattle are shown separately. Because that data should exist. And they may have very different characteristics. The serotypes coming from those categories.
Similarly, ground chicken versus chicken carcases and ground turkey versus turkey carcases. There is some value in keeping those separate. Okay.