International Poster Journal of Dentistry and Oral Medicine
Login:
username:

password:

Plattform:

Forgotten password?

Registration

Int Poster J Dent Oral Med 6 (2004), No. 2     15. June 2004

Int Poster J Dent Oral Med 2004, Vol 6 No 02, Poster 221

Comparison of different DIAGNOdent cut-offs for in vivo detection of occlusal caries

Language: English

Authors:
Dr. med. dent. Jan Kühnisch,
Prof. Dr. med. dent. Roswitha Heinrich-Weltzien,
Prof. Dr. Lutz Stößer,
Department of Preventive Dentistry, University of Jena, Germany Dr. med. dent. Thomas Oehme,
Dental Office, Lichtenstein, Germany
Andreas Ziehe,
Fraunhofer Institute FIRST, Berlin, Germany

Date/Event/Venue:
July, 3-6th, 2003
ORCA 2003
Konstanz/Germany

Objectives

Dental practitioners who use the DIAGNOdent device (KaVo, Biberach, Germany) were confronted with different clinical guidelines of the manufacturer as well as cut-offs from clinical and laboratory studies (Tab. 1). Therefore this in vivo study aimed to test recently recommended cut-off limits for the laser fluorescence measurement for occlusal caries detection.

No Study Sound Enamel lesion Dentin lesion
D0 D1-2 D3 D4
  Cut-offs recommended from clinical studies
I Lussi & others (1999b) 0-15 16-25 > 25
II Verdonschot, Abdo & Frankenmolen (1999) - > 19 > 19
III Lussi (2000) 0-14 15-20 > 20
IV Lussi & others (2001) 0-13 14-20 > 20
  Cut-offs proposed by the manufacturer
V Clinical Guidelines, KaVo (1998a) 0-4 5-10 11-20 > 20
VI Clinical Guidelines, KaVo (1998b) 0-4 5-10 11-24 > 24 or 30
VII Clinical Guidelines, KaVo (1999 and 2002) 0-4 5-25 26-34 > 34
VIII Clinical Guidelines, KaVo (2001) 0-9 10-17 > 17
  Cut-offs reprted from laboratory studies
IX Lussi & others 0-4 5-10 11-18 > 18
X Shi, Welander & Angmar-Månsson (2002) 0-7 8-21 > 21
XI Pereira, Verdonschot & Huysman (2001) - > 10 > 10
Table 1: Recommended cut-offs for the use of the DIAGNOdent (KaVo) to detect different stages of occlusal lesion extension.

Material and Methods

Study population

  • 248 permanent molars of 94 patients (mean age 19.2 yrs) of a general dental practice were included.

Laser fluorescence examination

  • After professional tooth cleaning and cotton roll isolation the teeth were examined by the DIAGNOdent device (probe A).
  • The maximum DIAGNOdent reading (0 to 99) obtained from the whole fissure pattern was recorded.

Validation by fissure opening

  • The extent of occlusal lesions (gold standard) was determined after minimal operative intervention with a fissurotomy bur (SS White, USA).
  • To quantify the lesion extent a probe with a 2-mm-graduation (P 2-12 YX, Hilite/ American Eagle, USA) was used for measuring the distance between the cavity floor and the outer enamel margin of the cavity. Validation ratings were obtained according to the greatest extent of the lesion on a scale from D0 to D4.
  • All cavities were restored with a composite material.

Results

  • Upon fissure opening 24 teeth had an enamel caries and 224 teeth revealed dentin caries of which 58 and 166 were up to half or beyond half the dentin.
  • The comparison with DIAGNOdent cut-offs revealed considerable variations of the performance (Table 2 and 3). Recommended cut-offs between 17 to 21 for superficial dentin lesions were in the same order of magnitude (κ = 0.48 - 0.51). On the D4 level only the manufacturers cut-off of > 34 achieved the best performance (κ = 0.51).
  Lesion extent D1-2 versus D3-4  
No Cut-off SE SP % correctly diagnosed lesions κ
I, VII > 25 89 63 87 0,40
II > 19 94 58 90 0,48
III, IV > 20 93 63 90 0,49
V,VI > 10 99 17 91 0,24
VIII > 17 96 54 91 0,51
IX, XI (in vitro) > 10 99 13 92 0,18
X (in vitro) > 21 93 63 90 0,49
Optimal cut-off in this study > 18 95 58 91 0,51
Table 2: Performance of the different DIAGNOdent (KaVo) cut-off limits for detecting superficial dentin lesions (D3 level)

  Lesion extent D1-3 versus D4  
No Cut-off SE SP % correctly diagnosed lesions κ
V > 20 95 28 73 0,28
VI > 24 93 33 73 0,30
VII > 34 85 66 79 0,51
IX (in vitro) > 18 98 27 74 0,30
Optimal cut-off in this study > 37 84 70 79 0,54
Table 3: Performance of the different DIAGNOdent (KaVo) cut-off limits for detecting deep dentin lesions (D4 level)
  • The optimal cut-off limits based on the highest κ-values were > 18 for superficial dentinal caries and > 37 for deep dentinal caries (Figure 1 and 2).
Figure 1: Sensitivity, Specifity, sum of Sensitivity and Specifity and Kappa of different DIAGNOdent values for superficial dentin lesions. Figure 2: Sensitivity, Specifity, sum of Sensitivity and Specifity and Kappa of different DIAGNOdent values for deep dentin lesions.
  • According to the κ-values and the area under the ROC curves (Figure 3) the agreement between validated caries extent and laser fluorescence value is still unsatisfactory.
Figure 3: ROC-Curves of superficial (D3-4) and deep dentin lesions (D4) for DIAGNOdent (KaVo) readings.

Conclusions

Regarding the limited data base cut-off values for the clinical use of DIAGNOdent should be interpreted cautiously until more in vivo studies are available.

This poster was submitted by Dr. med dent. Jan Kühnisch.

Correspondence address:
Dr. med. dent. Jan Kühnisch
Department of Preventive Dentistry
University of Jena
An der Alten Post 4
07743 Jena
Germany