Public Health Implications of Chronic Periodontal Infections in Adults

Question and Answer Session: Day 1, April 8, 2003

Moderator: Gary Rozier, DDS, MPH

Question
Marjorie Jeffcoat – How do you look at the possibility of a positive outcome? Obviously, we want to design every experiment to disprove our hypothesis….We heard a lot of what couldn’t be done, but how would you do it?

Answer
Philippe Hujoel – The science of randomized clinical trials has developed tremendously over the last 20 years. Several important principles must be adhered to in order to avoid junk science. One important principle is intent to treat. Randomization is an incredibly fragile process and can be manipulated extremely easily, so the integrity of the randomization process is another issue that I consider extremely important. Independent data and safety and monitoring boards are also important in making sure that one gets good clinical outcome data. The issue of surrogate endpoints versus true outcomes is another issue. Many lessons have been learned during the last 40 years of doing clinical trials that need to be incorporated into clinical trial design, analysis, and interpretation of results. If those are not present, you can find results such as if you give prenatal fluoride to mothers, you’ll reduce the incidence of preterm low birth-weight by 90%. Why are results like that present? Because a lot of important principles are not adhered to. So, there’s a checklist on the Web of clinical trial principles that should be adhered to. The JAMA (Journal of the American Medical Association) Web site will give you a list of elements of clinical trial design that need to be present.

M. Jeffcoat – Yes, most of us are familiar with those.

P. Hujoel – But it’s amazing how principles that appear to be very simple are actually not adhered to. The one I would like to stress is intent to treat. It’s fascinating to me how an article can say this was an intent-to-treat analysis, but if you look at the numbers you get the opposite conclusions if you really do an intent-to-treat analysis. One has to adhere to the principles and provide evidence that those principles were adhered to.

M. Jeffcoat – I don’t think anyone would argue with you about that. And those are review issues. If one method of therapy doesn’t work, that does not disprove or prove the entire hypothesis. Some drugs are going to be effective, and some aren’t.

P. Hujoel – I agree entirely with that point. The point that I’m very scared about is, for instance, with that particular study, if you are a practicing periodontist and you read that article, you will come away with the impression that the next woman who has periodontitis and who’s pregnant, I will prescribe metronidazole, because in the abstract it says if I do prescribe it, I lower the risk for pre-term birth by 80%–90%.

M. Jeffcoat – I’m not saying the result is incorrect in their study. I am saying one has to take the preponderance of the evidence and I agree with you on the metronidazole issue, but it doesn’t make the hypothesis wrong.

P. Hujoel – Absolutely not. When you do a multi-center clinical trial and you do find evidence, that’s how science advances. I’m worried that clinical trials are being published that seem to make very significant claims like about the low birth-weight issue itself—which has been troubling the scientific community for 40 to 50 years and for which no solutions have been found that are effective. Suddenly, clinical trials are being published in the periodontal literature as if we can eliminate 90% by doing periodontal scaling. Those things are worrisome from the perspective that if we look at evidence from the medical literature, the treatments being claimed to be effective actually are harmful. I really am concerned about people initiating randomized clinical trials and using metronidazole, given that evidence.

M. Jeffcoat – I’m not going to disagree with you on that. But you don’t throw the baby out with the bathwater. The hypothesis is still there to be tested.

P. Hujoel – Periodontal research does not happen in a vacuum. There are cardiologists making the point that, in giving antibiotics for chlamydia infections, there have been four or five trials and they’ve all turned out negative. Those trials and results have direct implications for periodontal research. If antibiotics do not have an impact on that type of infection, we should incorporate that evidence into our decision making. If metronidazole does not have an effect for low birth-weight in medical trials and actually shows adverse effects in that there is significant harm associated with it, we should be extremely careful before we do it and extremely careful before we publish such results. Like I say, if you start calculating the public health implications of giving advice that is potentially harmful, it’s worrisome.

Question
Deborah Greenspan – I have two questions. The first concerns the issue of papers that are published in reputable journals that we all cite and the concerns that were raised. I don’t want to discuss the paper about low birth-weight but, rather, to take it out more generally. The concern is about papers reporting clinical trials and what types of statistical analyses have been chosen for the data, yet when reviewed by a biostatistician or epidemiologist, may generate tremendous concerns. What can be done with some of these good journals when you’re asked to review a paper and perhaps you don’t have that expertise but want it applied? You may even suggest to the editor that it goes out to a biostatistician, but you as a reviewer never know if that’s ever been done.

My second question concerns the use of C-reactive protein diagnostic test. I’d like to know what the panel thinks about the use of C-reactive protein test in papers that have been published up to now, and comparing those with papers that may be published in the next few years.

Robert Genco – Maybe I can give some insight in terms of what we do in the Journal of Periodontology. If you are a reviewer and nonstatistician and would like to see the other reviews, you can ask to see them and we would send them to you. We always use a statistician except maybe on a single case report. We have a bevy of statisticians and Philippe (Hujoel) is one of them.

Brian Burt – Faith is coming to community dentistry. It’s not at all uncommon for reviewers to say that they don’t feel capable of dealing with these particular methods. They may suggest names; otherwise, it goes to a statistician who is familiar with these pieces of work. It’s also our journal’s policy when the final decision is made on the paper to send copies of all reviews to all the reviewers so they can see how their reviews stacked up against the others.

B. Burt – It varies. The shortest review I’ve received was three lines. I’ve also had eight pages. I’ve actually had reviews longer than the paper that was reviewed. I’m not sure what that tells us, but it’s interesting.

Question
Gary Rozier – Would a panel member like to respond to the second question about the C-reactive protein diagnostic test?

Answer
Steve Offenbacher – We have a couple of papers from the ARIC (Atherosclerosis Risk in Communities) dataset that are coming out on the C-reactive protein, and the assays we are using are of high sensitivity and quite comparable to the values that you get with the new high-sensitivity nephelometry tests.

Question
G. Rozier – Drs. Jim Beck and Scott Tomar mentioned the Surveillance Work Group yesterday. I know there’s interest in having biological markers for surveillance. As a follow-up to the question about the C-reactive protein, my understanding is that we have no diagnostic test at this point that would be appropriate for a surveillance system?

James Beck – We haven’t addressed that issue. We haven’t talked at all about diagnostic tests. Of course, whenever you get into that issue, the immediate concern is that it costs money. And when you’re talking about large numbers of subjects, then budgets apply. You’re talking about money. And if you ask us at this point, “What diagnostic test would you use to indicate whether periodontitis was present?” I think there may be some candidates for a test, but there would be a lot of discussion about which would be the appropriate one.

Question
Richard Valachovic – I’m the executive director of the American Dental Education Association, which represents all the dental schools in the United States and Canada, the allied programs, hospitals, research institutions, and their faculties and students. As we look at the public health implications of the issues we’ve been talking about, an important factor in a public health response is the knowledge, skills and beliefs of the providers who will interact with patients. If we think about the students and residents and others who are finishing their programs this year, they’ll probably be practicing until 2050, so what they understand and appreciate now probably becomes very critical to how we respond over the next half century. What should the curriculum of a dental school in the United States and Canada and throughout the world be doing to address these issues right now to prepare for the future?

Answer
P. Hujoel – I think it’s important that these dental students be provided with the tools to assess the literature for their roles. A course in evidence-based dentistry, epidemiology, and statistics, which form part of the medical curriculum, should be part of the dental curriculum, too. Dental students who graduated 10 years ago didn't have information on tooth-whitening procedures. If they were provided with the tools to assess the literature of tooth whitening—to know what is good evidence and what is bad evidence, what should I believe or not believe—they would be continuous learners and continuously evaluating the new evidence that comes out. That would be the most important thing to contribute to a dental education, because in all likelihood the dental profession will change tremendously over the next 20 or 30 years, and if we don’t give them those tools to be critical consumers of the scientific literature, it will be very hard for them.

R. Genco – I’d like to support what Philippe said. The knowledge that we have right now is going to change dramatically over the next 10 to 20 years. We can’t teach students what they have to know when they’re practicing 20 years from now. So the tools, the ability to evaluate the information out there, to be lifelong learners and to think critically, are paramount. Often we get hung up on the mechanical aspects of dentistry and the widgets, and we don’t step back and spend enough time on those thought processes and giving the students the tools to invoke those thought processes.

John Gunsolley – I’d like to reiterate that. But also I think there’s something very important lacking from many dental school curricula and that is the importance of community and public health dentistry. Most of dentistry taught nowadays in dental schools is directed to individual patient care, and that’s appropriate. But I think that community dentistry and public health dentistry in many dental schools is trivialized. You can’t flunk dental school because you did poorly in community health, in most schools. And I think that’s a key issue.

B. Burt – Another group that needs to be educated is physicians and nurses. My experience in trying to communicate with some people in positions of authority in those environments with respect to oral hygiene in hospital wards, for example, doesn’t often penetrate well. I think it’s because these people have not been educated to any extent in oral health. So it’s another venue where we can make inroads in educating people who can make important decisions.

G. Rozier – Another important target group are students of schools of public health. I’m not sure how I can address how the curriculum can be changed at this point, but certainly it’s an important target group as well.

Question
Dolores Malvitz – I’m not an epidemiologist and don’t pretend to be. But back in a former life, when I was teaching research methods to students who included graduate students in periodontology, we talked about stages of epidemiologic evidence, association studies, case-control studies, hypothesis formation, and clinical trials. I’d be interested in hearing what case-control research question we should be studying next if we’re to fill the gap between association evidence and clinical trial evidence that’s down the road. What case-control studies have we done and what do we need to do?

R. Genco – I put a paper together with those levels of hierarchy of evidence and I also teach to dental students. I’ve looked at it with respect to the associations and I think for most associations, we’re beyond case-control. We are probably at the level of randomized control trials for many of these associations, so there’s power in the case-control. In heart disease, that’s how it started, in 1988 and 1989 with Matilla, et al.— 50 cases, 50 controls showing an association. Of course the problem there is that they’re not population-based; you can’t adjust properly for smoking and other factors. You can attempt it, but it’s not too powerful. Then we did population-based cross-sectional studies and you heard those results, then the population-based longitudinal studies and the mechanism studies. In the heart and diabetes areas, I think we’re at the level of randomized control trials to nail down the critical public health question. So what? If you do modulate periodontal disease, will it have any effect on systemic diseases?

P. Hujoel – I would quickly like to say that in my opinion we are not at that stage. At this moment, there is no convincing epidemiological evidence that among “never-smokers,” consistent associations exist across different populations. I think a randomized clinical trial is a huge investment and at this moment we see associations between periodontal disease and low birth-weight, chronic obstructive pulmonary disease, lung cancer, stroke, and diabetes, all diseases. If we start doing randomized clinical trials for every disease that we see an association with, we’ll be bankrupt. I think we need to look at all the systemic diseases that have been identified and start looking very seriously at which of these pop up consistently from different investigators in different populations. At this stage, it is critical that the dental community agrees that a dental component is attached to ongoing cardiovascular disease and stroke studies. This will give us important information that provides direction. To jump into clinical trials without having convincing evidence is a very high-risk gamble.

Question
George Taylor – Thanks to all the presenters for excellent lectures today. One question that came out of the discussion was the notion of infection and inflammation and perhaps management of inflammation as a target. Does anyone have any thoughts about recent evidence that’s emerging with regard to sub-microbial doses of tetracyclines and the inflammatory responses? There are papers that are looking at the effects of glycemic control and perhaps complications of diabetes and other systemic diseases.

Answer
Unknown – One thing will be clear with the sub-microbial doses: If a randomized clinical trial is an issue, this will not be a trial of a hundred people if you want to look at outcomes such as stroke or coronary heart disease. This will be thousands and thousands of people. It’s not unusual to have a cardiovascular disease trial that is between 20,000 and 50,000 people. So to jump into this trial, once again, we have to make sure there is a lot of evidence that substantiates such trials. The important thing is to see that it’s worth it to make a $100 million investment in a clinical trial on this topic.

Frank Scannapieco – I think Maria Ryan is presenting evidence that Periostat has effects on a variety of markers, including C-reactive protein. So it might have an anti-inflammatory effect. Whether or not it will be useful in treating atherosclerosis or its effects, I’m not sure, although they’re probably trying to investigate that aspect as well.

Question
Kaumudi Joshipura – I wanted to speak to some concerns that Dr. Gunsolley and Dr. Hujoel raised. Do different periodontal measures used explain the inconsistencies in findings of association, or is it control for smoking and several other important confounders, such as socioeconomic status or healthy behavior, which is difficult to measure? You can’t control it if you can’t measure it. From our own data, I can give you an update on some recent work in progress. You mentioned that self-reported measures may be a reason why we are not seeing an association with heart disease. Our measures have been validated and shown to be almost compatible with the bone loss measures in terms of predicting the disease but mainly we now have radiographic data for the coronary heart disease cases and controls. And we’re not seeing any associations no matter how we break it down. We’ve looked at several cutoffs using bone loss measures, and we still don’t see an association with coronary heart disease. On the contrary, we have recently published data on ischemic stroke and peripheral artery disease. And we see a significant association there. We have consistently found associations with tooth loss for coronary heart disease, for ischemic stroke and for peripheral artery disease. It’s a little harder to interpret; it’s partly periodontal disease, but there are other things going on there. In the tooth loss data for ischemic stroke, we do see associations in “never-smokers?” For most of other periodontal associations, we don’t see them among “never-smokers.”

Answer
R. Genco – The bottom line is that the data are what the data are. Certainly a yes/no is not as exquisite as a more continuous measure. But, in that case, if you can’t see it in your data and it’s contrary to one or two other studies, that strengthens the other side of the argument that there’s not a relationship.

J. Gunsolley – I got into my time machine and went ahead 20 years (you’ve all aged a bit) and heard the same discussion. We’re not using the right measures, we’re not adjusting for confounding factors, we don’t have the right study, we should tie up with medical studies, etc. I would make the point that we need to know the answer to the question—If you do modulate periodontal disease, is it going to make a difference? It may be expensive, but I think a lot of these other studies are expensive, too, and we may be fiddling for another 20 years and never answer that central question. The best epidemiologic study is not going to answer the question. They’re not experimental and hypothesis-driven. We need randomized clinical trials.

K. Joshipura – I’d like to address the issue of clinical trials versus epidemiologic studies in general. Yes, clinical trials are considered the best form of evidence and you can test in clinical trials what the treatment does. Once you have periodontal disease, you treat it, and how does that impact the outcome? You cannot test the appearance of periodontal disease. You cannot test the difference between people who have a lot of periodontal disease and those who don’t; that needs observational study. And that is the key question for prevention and public health implications. A public health approach is an alternative to waiting to treat everybody with periodontal disease. I think clinical trials are worth doing, but so are epidemiologic studies.

P. Hujoel – As far as clinical trials now or waiting and getting better epidemiological evidence, obviously what’s going to happen is a political decision, not a scientific decision. But if you look at the history of epidemiology and what happened in the 20th century, it’s always the epidemiology, the consistent association identified across different populations, whether it’s for liver cancer, cigarette smoking, cholera. Strong epidemiological associations have always guided clinical research and clinical trials. Typically, it was only decades later that effective interventions were found. So we can turn this around and say, we’ll go straight for clinical trials. It’s a very unusual and high-gamble approach. There are many diseases associated with periodontitis, such as peripheral artery disease. Maybe that’s where everything is happening. We need to identify different populations and see if that’s the one that consistently occurs among populations of “never-smokers.” Once we do find that, great. Then we can start looking into biological mechanisms and start thinking about clinical trials. Now we have about six or seven diseases associated with periodontal disease. All the evidence is pretty inconsistent. Of course we can decide "Yes, we’re going to do clinical trials now," but it’s a very high-risk gamble that has not paid off for other chronic diseases.

Question
Frank Scannapieco – I wanted to ask a question of the epidemiologists in the audience. Is it possible to overcorrect the data we obtain and by doing so might we eliminate associations that in fact exist?

Answer
B. Burt – The answer is yes, just as it’s possible to underadjust, one can overadjust and try too many things in this world. There are some set conventions as to how to choose variables for a model to test a particular question, but like anything else, a lab technique or treatment for a patient, it can be overdone and wrong.

J. Beck – I think certainly you can overadjust. A lot of decisions are based on what you do statistically, but a lot are based on what you know already. Those of us in dentistry, who have gotten into cardiovascular disease research and systemic disease research as it relates to oral infection, didn’t know a lot when we started. I think that’s the value of other studies that are going on at the same time, such as the animal studies, that can cause you to question what you really should be adjusting for. For example, we tend to adjust for something like lipids because if you’re doing a cardiovascular study it’s one of the risk factors. We’re seeing evidence that lipid levels go up in animals that get periodontal infection. We didn’t know that before. So the question is, if periodontal infection is also affecting lipid levels, you probably want to think carefully whether you want to adjust for those lipids if they’re on the causal pathway. Part of the problem we’ve had is that we adjusted for what we thought was related, but we may not be adjusting for the right thing. That’s a way of overadjusting also.

Question
Gordon Douglass – I want to thank everybody for excellent presentations today…like that of Roy Page this morning when he described the bio-film we deal with and how it’s resistant to antibiotics and can only be treated with mechanical removal. And Dr. Hujoel, I appreciate your comments on the casual use of antibiotics in the treatment of periodontal disease. If you take the logic from Roy’s presentation this morning to what you were suggesting this afternoon, there may not be a place right now for the use of antibiotics. But is there a place in the management of patients with periodontitis who have systemic illness with traditional mechanical periodontal therapy? Is there a risk for scaling and root planing in this individual, or might they be a little better off because their inflammatory profiles will improve? Are we better off providing them with oral hygiene instructions, and what is the risk of making that recommendation?

Answer
J. Gunsolley – There’s no evidence until we do a randomized clinical trial. With some certainty, I hope, we can say that we can improve their periodontal condition. In fact, in some work that we’re doing, I can suggest that antibiotics may improve their response, as do local therapies, and low-dose tetracyclines.

P. Hujoel – I think the best approach is to recommend scaling and root planning, provide oral hygiene instructions, and improve periodontal health.

B. Burt – I’m not a clinician but I remember in dental school many years ago we were taught that patients who had suffered rheumatic fever, whenever an extraction was to take place needed antibiotic cover. And I believe that protocol is still in place today. So, if we assume that scaling and root planing results in a bacterial shower for a person who already has a history of disease, do we do antibiotic cover and if not, why not? What is the difference between the two approaches?

Question
Unknown – I have an immunological question so it may be more directed at Dr. Van Dyke. In terms of cytokines, these are low levels that don’t simulate complement, or sending signals to MHC class two cells, or is this a more complex process and we’re just getting little snapshots of what’s known at this points? My follow-up to that is, what do you know about vaccine development against these pathogens that were presented?

Answer
Thomas Van Dyke – I would say we’re looking at snapshots of things and in the end you try to put them together and make greater sense of the whole. That was the basis of my comment this morning referring to Dr. Page’s comments that it was the macrophage that was central to the pathogenesis of periodontitis. I found that hard to reconcile based on all the available data pertaining to where these mediators are coming from and what cells are producing them. If you isolate one cell, you can measure a mediator profile for that cell, but if you isolate another cell, you can show that the profile will be similar for the second cell. It’s up to the investigator to draw the arrows of our flow chart based on our bias. So, I’m, not sure how we do that in a rational way.

Harvey Schenkein – I was going to comment on the cytokine story. Van Dyke said, "You look at these pathways and everything is so interactive and multifunctional that they’re really hard to isolate." But in some cases you can look at these as indicators of the sorts of immune reactions that are most prevalent in patients. For example, if you look at individuals who are genetically programmed to produce hyper-inflammatory responses, one can get a picture of their risk for disease. The pathways are interesting and we need to know the functions of ctyokines and molecules, but as an overall estimate of an individual’s susceptibility. Why is only 10% of the population susceptible to severe periodontitis? It may be a combination of risk factors related to the disease pathogenesis, and these may include cytokines and other molecules, prostenoids, immunoglobulins, and others. That’s where these things fit in to us as clinicians and, hopefully, looking to the future in developing risk profiles.

R. Genco – Regarding vaccines, there have been extensive studies in rodents, both mice and rats, with P. gingivalis whole cells, outer membrane peptides, hemoagglutenins and proteases, and they’re all protective. Recently there are studies with B. forsythus, again in mouse and rat models, using whole cells and adhesins. Two groups of investigators have used whole cells P. gingivalis and have gotten some protection. So there’s an extensive animal model literature, plus the observation that humans who have periodontal disease have high levels of antibodies to P. gingivalis. And the LJP aggressive periodontitis patients have high levels of antibodies to A. actinomycetemcomitans, suggesting that there may be some antibody effects in humans.

Question
Mark Mallatt – Good job. Every presentation today was thought-provoking. I’m Mark Mallatt, state dental director for Indiana and professor at the Indiana University School of Dentistry, and as such I’ve participated in a lot of clinical trials, probably 70 over the last 20 years. A lot of these were chemo-therapeutic trials but we were looking at gingivitis, bleeding sites, and plaque in most. We have a pretty good definition of periodontal disease. We saw it a couple of times as 5mm pockets, 2mm attachment loss or whatever. But we need to back up a little and talk about gingivitis and nip this in the bud because this is a public health problem. After doing about 35,000 gingivitis exams with a probe, I’m not sure I know what “healthy people” gingivitis is. My point is, we have an age group of 35 to 44 and the Healthy People 2010 objective is to reduce gingivitis by so much, but I’m not sure we’re all in the same boat as far as what gingivitis is. Does that mean two sites, 10 sites, bleeding? There are a lot of indexes and there’s been a lot of time, effort, and money spent to get the ideal gingivitis index. Also, when you do exams on adults, you don’t have a captive audience as you do in a school setting. It’s difficult to get a 35 to 44 age group. Also, in this early stage of the disease, you’ve introduced informed consent. Whenever you do that, you’ve skewed the results. Are we ever going to agree on gingivitis as far as amount or extent? Brian, probably you could comment.

Answer
B. Burt – You mean to say what is a case of gingivitis? It’s mostly now based on bleeding and probing, the standard protocol. One site that bleeds is a case of gingivitis and someone who has 32 sites that bleed is also a case of gingivitis. If you start throwing number of sites into things like the goals for Healthy People 2010, you complicate it enormously. As it reads at present, it’s fairly simple. It would be fairly understandable to any member of the public, so in that context I would not favor going too deeply into a definition of gingivitis. But the same applies across the board. Periodontitis, if we’re going to talk about how much loss of attachment, if we’re going to measure pocket depth, how many sites per teeth, etc., there is no consensus on these things. There are various case definitions such as those Jim Beck described, combinations of loss of attachment and pocket depth, but everyone who uses these knows they are relatively arbitrary.

M. Mallatt – I found it interesting that one of the slides used the Russell Index. In fact the probe is not used in that index, which is probably one of its shortcomings.

P. Hujoel – I think your question touches on a key issue in dental research today. All of our definitions of gingivitis and periodontitis are based on surrogate measures. We really need to move to patient measures. I’ll use one example. Recently an article about knee surgery trial was in the New England Journal of Medicine. If we use surrogate measures to assess this knee function, you can devise all kinds of tests as to what you can do with your knee after the surgery. That was not used as the primary outcome. The primary outcome was a self-assessment—a questionnaire asking people how well their knee performed. People are going to say "Oh, my God, that’s tremendously inaccurate; it’s much better that we do a knee function test, and see how many pounds it can push.” All surrogate measures have a fundamental problem in that we do not know whether they reflect outcomes that are of tangible patient benefit. What we want is to achieve outcomes that provide tangible benefit to the patient. All our measures of gingivitis and periodontitis are surrogate measures that have not been validated. It’s a key problem that I think will be solved by moving to true endpoints. And it’s already happening. For instance, they’re starting to use self-reported measures of gingivitis with simple questions such as “Do you see blood on the toothbrush after brushing your teeth?" The closer you move to patient-assisted outcomes, the more you can be assured that you are indeed providing a tangible patient benefit.

Question
Poul Eric Petersen – I think this program has been most interesting and has provided an analysis of the state of the science in periodontal health in relation to general health. The discussion has focused on the possible association between periodontal health and general health, particularly the causal interrelationship that may be observed. But I wonder whether the practical or public health intervention flowing from whether or not there is a causal interrelationship would have an impact. It is also evident from the discussion that periodontal diseases and general diseases are highly related to risk factors, particularly smoking. We’re a little in doubt on the role of oral hygiene in periodontal health. But if it is the case that tobacco use is the most important risk factor for periodontal disease and general health, it would probably not affect public health intervention programs that much. You would probably focus on tobacco use for preventing periodontal health problems and cardiovascular disease, and not justify that people have clean teeth to prevent dying from cardiovascular disease. So the practical implication is maybe not reflecting that much as to whether there is a causal interrelationship but it’s important from a scientific point of view, of course. That leads to my question, particularly to John and Brian, because for public health administrators and for the World Health Organization (WHO), it is important to establish health information systems that can help us monitor disease conditions and risk factors. It was evident from the discussion that there may be a need for standardized information that could be included in such a health information system. Can you recommend four or five indicators that would be included for surveillance of such disease conditions in relation to periodontal health? Is it possible to dream up such robust instruments that are simple and easy to apply at the field level but also easy to interpret and translate for public health administrators?

Answer
J. Gunsolley – As a periodontist, and from a clinician standpoint, I advocate periodontal measures. You ask a different question in what is a screening-type exam that would be appropriate. In some epidemiology studies where they’ve done a couple of sites or half the mouth, they’ve cut down costs, but I think without some type of measuring with a probe, you can’t measure the disease. The other thing you brought up is about impact; it’s interesting that things like smoking cessation plans, where we as dentists often see patients more often than physicians, are incredibly important. Even though the success rates of smoking cessation programs, are not great, we have an opportunity and should be well-positioned and aggressively pursue it. In our dental education programs, we should point out to dentists that not only should they support the programs, but it is their responsibility as clinicians to do that.

B. Burt – Beyond tobacco, in which there was a major global initiative in process just last week, I would hesitate to say we could get global agreement on indicators at this stage. I don’t think we’re far enough along to say that with any confidence. But if we can get the world to stop smoking, I think that’s a reasonably good achievement. Let’s do that first and do the rest later.

Alan Cross – The CDC has asked a panel to look at the issue of measures for public health surveillance. Scott Tomar gave a description of the workings of that group. So with a lot of hard work and a little luck, maybe in 2 years we might have something.

Question
Ann Battrell – I’m Ann Battrell, director of education for the American Dental Hygienists Association. One of you had a slide about Healthy People 2010 objectives that do not address gingivitis in children, and a previous speaker addressed that issue. One objective of this workshop is to look at population-based interventions for preventing and controlling periodontal disease. We focus so much on quantitative data and clinical trials and argue about what’s right and wrong or what should be included. I was wondering about your comments about the qualitative side, the behavioral side of this very important issue, because the heart of what dental hygienists do is to prevent these diseases, when you’re talking tobacco cessation, oral hygiene instruction, scaling, and root planing. And so much of what the CDC is about is public education. Oral health is so lacking in that area right now, yet much of what we’re talking about is only the science. That applies directly to the issue of how do we agree on what is gingivitis. Right now in this country, practitioners don’t even have a dental insurance code to treat gingivitis, yet there’s a prevalence. We have it to treat periodontal disease and someone with a healthy mouth, but the gingivitis code is completely lacking. The practical considerations of this workshop are huge, but also on the behavioral side, not just on the quantitative side.

Answer
T. Van Dyke – Isn’t that an issue for the American Dental Association (ADA), which makes the codes? To my knowledge there is no diagnostic code at all.

A. Battrell – I think it’s not only the ADA, it’s all of us. That’s part of the problem; we lack diagnostic codes, so what are practitioners to do?

B. Burt – On the good side, change does take place over time. If we look at tobacco, in the l960s something like 60% of American men smoked. Now it’s 25%, a huge change. Use of seatbelts in cars has been increased by a combination of legislation, fines, and education. Nearly every state has seatbelt legislation now, and it’s a great triumph of people complying. You think, "My goodness, what does it take?" Here is a measure that clearly can save your life. We’ve been at it for 30 years with a mix of carrot and stick. This is a life-saving measure and we’re up to about 70% of compliance. With all due respect, when you look at those things and then say "What do we have to do to get people to brush their teeth properly?" maybe it’s not quite in the same league. But again, it takes time. I think we generally believe that gingivitis levels have improved, not necessarily for reasons of oral health but more for grooming issues, attractiveness, and so on, and will probably continue to do so. I’m frankly skeptical about having procedure codes for treating gingivitis; I’m not personally convinced that a treatment code would necessarily have the desired effect.

Mary Foley – My name is Mary Foley and I’m the state dental director for Massachusetts. I want to thank all of you for your presentations today. They’ve been very interesting and informative. I have a public health perspective, where evidence increasingly drives public health programs. I want to thank you for the discussion, particularly the importance of making sure that the studies we do are epidemiologically sound, and that we are very careful about publishing those studies. In public health, we work with many professionals who are not dental—funders and other folks who do not have the tools to critique the literature. So we have to be able to show them that the literature is valid. We appreciate all the work you’re doing in this conference and we hope that in the future there will be some work that we can move forward with respect to periodontal disease in our public health programs.

Question
Scott Presson – I wanted to ask for a few more comments about targeting. If we have this 10% to 15% that are at high risk, until we have the genetic test available, what about the value of trying to identify this group early in life, say in late adolescence or maybe their late 20s? How valuable might that exercise be?

Answer
P. Hujoel – For the 15%, targeting will be relatively easy for 80% of them, which are people that smoke. The remainder will be genetic. Smoking will be an easy way to target people who will develop periodontitis.

H. Schienken – There’s still a group that’s probably 3%, 5%, or 7% that are severely involved at younger ages for which a screening program might be very appropriate. Those folks get severe periodontal disease and need to be treated and monitored. It’s a devastating disease in some of those folks.

B. Burt – Yes, people with 3mm to 4mm loss of attachment in even one side at age 16 or 17—given that we know past disease is the best predictor of future disease—would seem to be identified as people who have to get to a dentist to be seen. But that and the smoking would seem to be good indicators.

R. Genco – Many of us are involved with developing risk profiles. I’ll talk a little about this tomorrow, but certainly most of our data show that 40% or 50% of periodontal patients are smokers. That may account for a significant amount. But I’m not sure about 80%.

F. Scannapieco – I’ve been confused about Philippe’s comments, too. We have some data on chronic periodontitis and found about 40%, which is twice the rate of people smoking in the normal population. Now, among individuals who smoke, I don’t disagree that you can attribute most of their periodontal disease to smoking, but there are still people who have never smoked who get periodontal disease, and I don’t see where it’s 80%.

P. Hujoel – That’s what I’m referring to. Scott Tomar published 54% in the U.S. population as attributable risk, meaning that if you have a person with periodontitis sitting in your chair, there’s a 54% chance that he is a smoker and the disease is attributable to his smoking. The 80% comes from some studies in the Boston area, where they sampled periodontal offices and general dental offices and dug up the number of 80%. In Sweden, the number was 90%. Part of this variability in estimates is due to the definition of periodontitis as discussed in the presentations.

R. Genco – But it’s more complex. We don’t know all the risk factors. You heard about some of the genetic factors; we’re just beginning to learn about those. There are many other factors, such as stress, inadequate coping, and diabetes. Impaired glucose tolerance and obesity seem to be related, as well as alcohol use. So there will be an expanding list of risk factors, some of which may be more important than others. But to do attributable risks, you need a panel of the most important. I don’t think we have that panel. Another [factor] is osteoporosis and calcium in the diet. There will be a list, but it’s premature at this point to say we know all the risk factors. We all have that goal in mind of identifying the 25- or 30-year-old with this risk profile that might include 8–10 risk factors. Another thing we can’t forget is the flora. In some of our studies, and in other studies that are epidemiologic, microbiologic studies, it’s been shown that two or three organisms also increase risk. The presence of those organisms, even prior to disease, may also be an indicator. We have a way to go, but that’s the goal—to take that set of risk factors, do some attributable risk, put the important ones in there, and develop a risk profile.

Scott Tomar – I just wanted to revisit a question that Brian had asked before and really wasn’t addressed. I believe the American Heart Association and certainly the ADA have promoted the use of prophylactic antibiotic coverage for certain types of prosthetic devises. Is there a concern that, if we recommend aggressive periodontal treatment for people with heart disease, that could actually put them at increased risk for an adverse event?

R. Genco - The question is, “Do you increase or decrease that risk?” The one thing about prophylaxis that’s always been a quandary to me is that if you have somebody who has periodontal disease, they can get bacteremias by oral care. Not ours, theirs.

Unknown – The use of prophylactic antibiotics is for bacterial endocarditis, which is a specific colonization of the heart valve and cardiac tissue that is different from the cardiovascular disease that’s been discussed. So this not predisposing to that component of it; it relates to the infectious etiology of bacterial endocarditis. That is why those antibiotics are used prophylactically.

R. Genco – I think you all know there’s a pilot trial supported by the National Institutes of Health (NIH) at five centers to look at this treatment and management of periodontal disease in patients who have heart disease and periodontal disease. We have labored with that specific question. In a heart patient, if you do a thorough, deep scaling, we know that in about 25% or 30% you’re going to get bacteremia. Is that detrimental in a heart patient? We don’t know the answer. We suspect that it isn’t, and maybe that’s because we hope it isn’t. But we don’t know. I would say that’s another reason for doing a trial. The main outcome of the trial will be the effect of periodontal therapy or management prevention on cardiovascular disease. But one of the outcomes may be, "Is there an increase in heart disease in the treated?" That’s certainly a possibility. If it’s built into the trial, we’ll see it if it’s of sufficient size. Clearly, these randomized control trials are very powerful for answering a lot of questions, including the public health-related question, which is, "If you modulate the disease, will you actually affect a chronic disease like heart disease or diabetes?"

J. Gunsolley – I’d like to build on that comment with a question. Endocarditis is an interesting example of a widespread public health measure that was instituted without any outcome data, based on articles presented in Mayo Clinic proceedings. I don’t know how it fits into prophylactic care…but it goes to a point I made a couple of times and to Bob’s [Genco] discussion about the need for some clinical studies. The reason that decision was made at the time is that the morbidity associated with endocarditis was so incredibly high compared with morbidity associated with treating people with antibiotics to prevent their malady. I don’t think it was a bad decision.

G. Rozier – I’d like to thank all of you for your attention to these excellent presentations and I’d like to thank all the panel members for their presentations and discussion.

Back to Chronic Periodontal Infections Conference

Historical Document
Page last reviewed: February 2, 2005
Content source: Division of Oral Health, National Center for Chronic Disease Prevention and Health Promotion