Numerous studies have considered the advantages of using radiographs and laser fluorescence in addition to visual exams to detect caries, and there has been ongoing debate over the effectiveness of these devices compared with visual exams using the International Caries Detection and Assessment System (ICDAS).
For example, while some researchers have reported good performance of the Diagnodent and Diagnodent laser fluorescence pen (LFPen; KaVo) in detecting occlusal caries, others have observed a large number of false-positive results with these tools, which limits their use as a principal diagnostic tool, according to a study in the April Journal of the American Dental Association (April 2012, Vol. 143:4, pp. 339-350).
And a recent study in Community Dentistry and Oral Epidemiology (June 2012, Vol. 40:3, pp. 257-266) found no benefit in additional radiographic examination when compared with a clinical exam for detecting approximal caries lesions in permanent teeth.
Despite these mixed results, many clinical guidelines advocate the benefits of adjunct methods used simultaneously with visual inspection -- primarily the increase in net sensitivity -- and recommend that dentists take two bitewing radiographs in children to detect caries lesions in primary molars.
Now a new study in Caries Research (August 16, 2012) concludes that adjunct radiographic and laser fluorescence methods "offer no benefits to the detection of caries in primary teeth in comparison to visual inspection alone," prompting the authors to recommend that current clinical guidelines be re-evaluated.
"Considering the current low caries prevalence in several countries, the actual necessity of caries detection methods should be evaluated," the study authors wrote. "The increase in sensitivity that is provided by additional inspection methods consequently decreases specificity; hence, the number of errors tends to be higher in populations with low caries prevalence."
For this study, the researchers -- from the University of SĂŁo Paulo School of Dentistry -- used visual inspection alone or in combination with radiographic exams and the Diagnodent LFPen to detect occlusal and approximal caries in the primary molars of 126 children.
Visual inspection was performed with a plane buccal mirror and ball-point probe using ICDAS. The radiographic exams included two bitewings taken using Kodak Insight film. The Diagnodent was used according to the manufacturer's instructions; tip 1 was used for approximal surfaces and tip 2 was used for occlusal surfaces. Two examiners assessed a total of 1,213 approximal surfaces and 407 occlusal surfaces.
Of the approximal surfaces, they found 1,162 sound or noncavitated approximal caries lesions and 51 cavitated lesions with an intact marginal ridge. For occlusal surfaces, the researchers observed 386 surfaces that were sound or with enamel caries lesions and 21 surfaces with dentine caries lesions.
"For approximal surfaces, we observed that the radiographic and LFpen methods, alone or combined simultaneously with visual inspection, presented significantly higher sensitivities than visual inspection alone," the study authors wrote. "Nevertheless, these strategies presented lower specificities."
In the occlusal study, "no differences among the strategies were observed in terms of sensitivity; however, LFpen used alone and both the simultaneous testing approaches exhibited lower specificities and accuracies compared to the visual inspection," they noted.
These findings led them to conclude that simultaneous combined strategies increased sensitivities but decreased specificities. "Notwithstanding the increase of sensitivity, the LFpen and radiographic methods do not offer any benefits to the detection of nonevident occlusal and approximal caries lesions in primary molars," they concluded.
Overcoming the limitations
This is not the first study to question the role of adjunct methods in caries detection, noted Joel Berg, DDS, president of the American Academy of Pediatric Dentistry (AAPD) and a professor of pediatric dentistry at the University of Washington School of Dentistry. "It is not that different than some other papers that have addressed the same subject," he told DrBicuspid.com. "The point is that some of the new technologies are sensitive but not specific, so there is a lot of noise and lot of false positives."
If a dentist is examining a patient and takes only one piece of information and isolates it, they might get the wrong result, he added.
"It's about synthesizing all the information," Dr. Berg said. "That's why only the dentist can make the diagnosis. A machine gives you information but it can't make the diagnosis. The dentist makes the diagnosis based on multiple pieces of information. Any single piece of information in isolation might lead you astray."
Gustavo Oliveira, DDS, an assistant professor in the department of general dentistry and oral medicine at the University of Louisville School of Dentistry found this study "very interesting" and its conclusions "bold and to the point." However, the "minutiae of the methods" should be fully understood to better interpret the findings, he emphasized in an email to DrBicuspid.com.
"The visual inspection of proximal lesions was achieved with the use of temporary separation of teeth by orthodontic rubber rings, which were placed around contact points for seven days. It is not difficult to comprehend that this procedure will allow enough separation of teeth to be adequately evaluated by direct vision," Dr. Oliveira said.
On the other hand, it is widely known that the bitewing radiograph view is associated with low sensitivity for both proximal and occlusal surfaces, he added.
"In other words, the findings are not a surprise when a diagnostic-aid instrument such as a radiograph has already an inherent poor performance on detecting caries, and the visual examination is facilitated with teeth separation, surfaces properly cleaned, and appropriate training in detecting lesions systematically," he stated.
Dr. Oliveira also noted that the LFpen was used according to the manufacturer's instructions, which means that the cutoff number was used in accordance with the instrument's manual.
"Previous studies have reported a mismatch between the cutoff numbers suggested by the manufacturer and the cutoff numbers empirically found that more accurately correspond to a clinical finding," he said. "The number used is much lower than what is scientifically accepted to be an optimal cutoff, which drastically increases the sensitivity. This fact alone may have significantly influenced their findings, since this instrument was previously found to be of good assistance in detecting occlusal lesions, when properly used."
Even taking into account this discrepancy, the study findings speak to the need for better diagnostic tools that are more specific, Dr. Berg added.
"There are many devices that are being developed using various methods that can identify carious lesions, not just noise," he said. "I think in time we will find a device that is very sensitive and that only gives useful information, not the noise of things that aren't there."
