Dr. Michael Glick’s August JADA editorial, "Risky Liaisons: What Is the Relationship Between Oral Conditions and Nonoral Diseases?" (
JADA 2007;138[8]:1056–9[Free Full Text]
), was excellent, with much needed candor on the "oral-systemic relationship," but a few more issues might have been addressed.
The difference between risk markers and risk factors is enormous. Cardiovascular disease is associated with upward of several hundred risk markers, but only nine major risk factors: family history, aging, smoking, diabetes mellitus, essential hypertension, central (waist) obesity, high low-density lipoprotein cholesterol, high total cholesterol and a sedentary life.
Altering risk factors may substantially improve cardiovascular health; altering markers is unlikely to change the disease pattern. To date, virtually all attempts to link periodontal disease with cardiovascular disease have addressed only risk markers and, if any, risk factors only for short periods in a disease with a lifetime of activity.
Secondly, relative risk calculations should never be used for clinical studies, as they greatly exaggerate risk versus the more appropriate absolute risk rate. A relative risk rate (RRR) of 100 percent (rarely seen in any study) is alarming, until one realizes that the absolute risk rate (ARR) of the same study may be increased by a "massive" 1/100,000 to 2/100,000. This is a doubling of the RRR, but only a change from 1 to 2/100,000 for the ARR.
Lastly, much greater attention must be paid to a "numbers needed to treat" (NNT) calculation for any proposed clinical therapy. The NNT is the number of subjects in the study (or a projected population) who need to be treated to get one positive result from the treatment. This could be 1/5 or 1/100 or 1/1,000, ad infinitum. Since this one benefited patient cannot be foretold, depending on the NNT, the costs could be relatively inexpensive ($100 per person with an NNT of 1/5 for $500) or costly if the NNT is 1/10,000 and the treatment costs were $1,000 ($10 million). What if the "treatment" has no benefit (the NNT is infinity)?
If the cost to ensure one treatment benefit were $50,000 or more (as has been calculated in some scenarios for statin treatment to prevent one heart attack in a hundred), would this be a reasonable cost-benefit ratio for a single person with periodontal disease to reduce a 1/100 risk of a heart attack with an as yet unknown NNT? What do you tell the patient (or his or her attorney), who will likely never benefit from this health expenditure, and all the monies spent go for naught?
Do you give them their money back after you sold them on the idea that periodontal treatment will prevent a myocardial infarction, and the NNT turns out to be infinity or maybe a little less? Do you tell them anything? Not today, you don’t.
