The article by Drs. Marc Heft, Gregg Gilbert, Teresa Dolan and Ulrich Foerster ("Restoration Fractures, Cusp Fractures and Root Fragments in a Diverse Sample of Adults: 24-Month Incidence," October JADA) has two serious flaws. These flaws should be identified to prevent readers from drawing incorrect conclusions from the article.
First, the article incorrectly refers to the information reported as "incidence" or "rates of occurrence" for cusp and restoration fractures. The authors correctly point out in the Abstract and the Discussion sections that the article actually reports consecutive prevalence comparisons, based on clinical examinations two years apart.
Consecutive prevalence studies can only be used to calculate incidence when the event of interest is permanent and cannot be obscured by subsequent treatment before the next examination. Clearly this is not the case for cusp or restoration fracture.
Unfortunately, the term "incidence" and related terms are used indiscriminately throughout all sections of the article, including the title, leading less cautious readers to assume that incidence data have been determined, when in fact they have not.
The second flaw also is related to the method used for determining "incidence." The conclusion states, "Blacks and people who seek care on a problem-oriented basis are at greater risk of developing [fractures]." This observation is obviously confounded by the effect of treatment done on teeth that fractured between the two examinations. Clearly, those who seek care less often will be less likely to receive treatment for fractures and, hence, will be more likely to exhibit fractured teeth at the second exam.
Thus, the conclusion must be considered flawed. We urge readers to regard the study’s findings with skepticism, and The Journal’s reviewers to be more attentive to such fundamental flaws in the future.
