I read with interest the March JADA article, "The Clinical Performance of Professionally Dispensed Bleaching Gel With Added Amorphous Calcium Phosphate [ACP]." There are numerous flaws, some so egregious in nature that the article should have been rejected.
First, the referenced literature is cited incorrectly. The article states that peroxide contact with dentin will cause outward fluid flow in dentinal tubules consequent to retraction of odontoblastic processes, and references Martin Brännström.1
Brännström, the architect of the hydrodynamic theory of tooth sensitivity, wrote that outward fluid flow stimulated the A fibers on the periphery of the pulp-dentin junction, resulting in the perception of pain. Odontoblastic retraction was not part of the pain mechanism. The reason why peroxide would cause sensitivity when placed on dentin is not explained. Peroxide contact with dentin does not cause pain because of its chemical nature; it would cause pain by virtue of its osmolarity.
The authors write, "However, a 1999 [American Dental Association Foundation] study ... demonstrated that this compound can also make teeth less sensitive to hot, cold, air pressure and tactile stimulation. ..."2 The article cited did no such thing. It cited a previous article in which a "small-scale clinical feasibility study" was undertaken.
The authors write, "[ACP] also may have the ability to directly depolarize nerve endings," and cite Yates and colleagues.3 Yates and colleagues mention nothing of direct depolarization of nerve endings and, for such an event to even occur, it would mean that ACP would have to be able to travel inward the length of the dentinal tubule and assert its effect within the pulp. In fact, Yates and colleagues concluded that "either the amorphous calcium phosphate was without therapeutic effect or the activity was masked by the placebo response in the control teeth."
There are concerns with the experimental design. The authors state that "the ACP-containing gel we used was similar to the control...." What does "similar" mean? Scientific method dictates that variables be carefully controlled. It is difficult to understand how authors could use similar solutions and arrive at definite conclusions. One thinks that it would have been appropriate to use solutions identical in nature, save for the inclusion of ACP.
The authors write regarding the ACP-containing test gel that the "manufacturer reformulated the individual components of the gel to incorporate the calcium ion (as CaNO3) and the phosphate ion (as K4P2O7) ..." ACP is one of the products of this combination; potassium nitrate is another. How did the authors discount the presence and consequent effect of potassium nitrate as a desensitizing agent in the experimental solution?
The article claims that the mechanism of ACP desensitization is via the obliteration of dentinal tubules with ACP. How could that be a factor within this study if subjects with any recession or sensitivity were eliminated from the study? If there was no dentin exposed in any of the patients, how do the authors propose ACP travel through the enamel to gain access to dentinal tubules to effect reduction in sensitivity? How can ACP obliterate tubules if there are no tubules available to obliterate?
Perhaps the most egregious error in the article is the statistical treatment of the data. The thermal-sensitivity, tactile-sensitivity data were all nonparametric subjective data. Yet, for some unknown reason, they are treated with parametric statistical analysis. The application of Kruskal-Wallis analysis would have been appropriate for this nonparametric data.
Further examination shows compounding of this defective action. The scales used for the dentinal and tactile sensitivity measurements were scaled from 0 to 4, while the subjective self-assessment of unstimulated tooth sensitivity was offered on a scale of 0 to 10. The scales for thermal sensitivity were described as "absence of pain, but perceiving stimulus," "slight pain," "pain during application of stimulus" and "pain during application of stimulus and immediately thereafter." Tactile sensitivity was scaled as "none," "slight," "moderate" and "intense." On page 389, the authors describe these data as "objective."
The tactile and thermal sensitivity data, subjectively recorded on a scale of 0 to 4, were calculated to two decimal places. None of the treatments resulted in changes greater than 1 full point on the scale provided. It is my opinion that calculating this kind of subjective data to two decimal places is preposterous, and treating it with parametric analysis is blatantly wrong. To draw conclusions of significance based on a parametric analysis of nonparametric data on a difference of 0.13 and 0.22 units (Figure 3) on a subjective scale from 0 to 4 is absurd.
One is left wondering how these glaring deficiencies escaped the peer-review process. Does it perhaps have something to do with the fact that the ADA benefits from the sale of ACP?
TOP
REFERENCES
