Reliablity of Nyvads

Reliability of the Nyvad criteria forcaries assessment in primary teeth

Mariana C. Sllos, Vera M. SovieroDepartment of Preventive and CommunityDentistry, School of Dentistry, Universidade doEstado do Rio de Janeiro, Rio de Janeiro,Brazil

One of the purposes of diagnosing dental caries is todetect and classify the lesions, in order to select the mostappropriate treatment for each tooth surface (1). Con-sidering the slower rate of dental caries progression incontemporary populations, the assessment of non-cavitated lesions is of great importance for measuring theeectiveness of preventive approaches in longitudinalstudies (26). If caries detection is restricted to countingthe number of tooth surfaces with cavities, the earlystages of the disease process are not considered. There-fore, caries prevalence is underestimated and thebehaviour of individual caries lesions cannot be moni-tored (7, 8). Many studies have conrmed that highinterexaminer and intra-examiner agreement can beachieved even when precavitated lesions are recorded(812), and since the 1950s, the inclusion of non-cavitated lesions has been recommended (1318).However, the World Health Organization decided not toinclude enamel caries lesions in their caries index, mainlybecause the diagnosis may not be reliable when thediagnostic threshold is set at the non-cavitated level, justas it was considered that the inclusion of enamel carieslesions would make a comparison of the caries situationamongst countries worldwide more dicult (19).Besides the inclusion of lesions at the non-cavitated

stage, the assessment of lesion activity is of majorimportance. The decision on clinical treatment will varydepending on the classication of the lesion as beingactive or inactive. Inactive or arrested lesions may notneed any intervention, whereas active lesions have to be

treated using non-operative procedures (such as oralhygiene improvement and topical uorides) or operativetreatment (6, 7, 20). The Nyvad caries diagnostic systemwas the rst classication system to dene clear criteriafor the activity assessment of both non-cavitated andcavitated lesions (8). Recently, an International CariesDetection and Assessment System (ICDAS) has beensuggested (4). This system is focused on the estimation oflesion depth and does not include lesion activity in theirprimary caries codes. A second score system to assessactivity has been suggested for use in combination withthe primary ICDAS codes (21).Clinical studies have shown that the Nyvad classi-

cation system has construct validity because activityassessment reected the expected eects of uoridetoothpaste on caries lesions (11, 22). Moreover, it hasalso been observed that caries activity assessment haspredictive validity because active non-cavitated lesionshave a signicantly greater risk of progressing to cavitythan do inactive lesions (22). With the diagnosticthreshold set at active vs. inactive lesions, the criteriahave also been shown to be reliable, with kappa coe-cient values in permanent teeth ranging between 0.68 and0.80 for intra-examiner agreement and between 0.74 and0.78 for interexaminer agreement (8). So far, the Nyvadcaries classication system has been used successfully inclinical studies (11, 2224), but more studies on its reli-ability in both permanent and primary teeth are needed.The purpose of the present clinical study was to assess

the interexaminer and intra-examiner reliability of the

Sellos MC, Soviero VM. Reliability of the Nyvad criteria for caries assessment inprimary teeth.Eur J Oral Sci 2011; 119: 225231. 2011 Eur J Oral Sci

This study assessed the interexaminer and intra-examiner reliability of the Nyvadcaries classication system in primary teeth and calculated the mean examination time.The criteria were based on visual and tactile examinations to dierentiate active andinactive lesions at cavitated and non-cavitated levels. Eighty children (37 yr of age)were examined under standardized conditions by calibrated examiners. At the toothsurface level, reliability was expressed as percentage agreement and kappa coecient,using four diagnostic thresholds: sound vs. diseased; sound or inactive lesion vs. activelesion; intact surface vs. surface discontinuity; and sound or non-cavitated lesion vs.cavitated lesion. Interexaminer and intra-examiner kappa values were, respectively:0.82/0.86; 0.80/0.86; 0.90/0.94; and 0.95/0.98. At the individual level, reliability ofestimates of the caries prevalence and of the decayed or lled surface (dfs) counts wereassessed at three diagnostic thresholds: sound vs. diseased; sound or inactive lesion vs.active lesion; and sound or non-cavitated lesion vs. cavitated lesion. For caries prev-alence, interexaminer and intra-examiner kappa values were, respectively: 0.84/0.94;0.69/0.74; and 0.95/0.97. The mean examination time was 226.5 s (SD = 128.5). Theuse of the Nyvad caries diagnostic criteria in primary teeth showed reliable results. Theexamination time was acceptable.

Prof. Vera Mendes Soviero, Faculdade deOdontologia, Clnica de Odontopediatria,Universidade do Estado do Rio de Janeiro UERJ, Av. 28 de Setembro, 157 (2/ andar), VilaIsabel, 20511-030 Rio de Janeiro RJ, Brazil

Telefax: +552128686372E-mail: [email protected]

Key words: dental caries; diagnosis; primaryteeth; reliability

Accepted for publication March 2011

Eur J Oral Sci 2011; 119: 225231DOI: 10.1111/j.1600-0722.2011.00827.xPrinted in Singapore. All rights reserved

2011 Eur J Oral Sci

European Journal ofOral Sciences

Nyvad caries classication system for caries assessmentand classication in primary teeth and to calculate themean examination time.

Material and methodsThe study sample consisted of 80 children [45 boys and 35girls; 37 yr of age (mean age 5.1 yr)], with a high cariesexperience, from a government school in Rio de Janeiro,Brazil. Children had to have at least four primary incisorsremaining to be included in the study. All children in thestudy sample were born and raised in an urban area thathad a uoridated water supply (0.41.5 mg l)1 of uoride).Informed consent was obtained from parents, and the studywas approved by the Committee for Ethics in Research atthe Rio de Janeiro State University.Dental examinations were carried out in a dental chair

under standardized conditions (compressed air for 35 s,articial light, cotton rolls, a dental mirror, and a sharpprobe) by two calibrated examiners (V.S. and M.S.), inde-pendently. Children had their teeth brushed by the rstexaminer before the examination. Then, they werepositioned in the dental chair and, if necessary, additional

biolm was removed during the examination using aprobe. Only primary teeth were recorded at the tooth sur-face level.The Nyvad criteria are based on visual and tactile

diagnoses to assess caries lesion activity at three progressionstages: the non-cavitated stage; the enamel discontinuitystage; and the cavitated stage (8, 25). The codes used toclassify the criteria in primary teeth are shown in Fig. 1. Inthe event of doubt, examiners were instructed to choose thecode representing the less severe status. However, in thepresence of two or more caries lesions on the same toothsurface, themost severe caries lesion was registered accordingto the following severity scale: active lesion>inactive lesion,and cavitated lesion>surface discontinuity>non-cavitatedlesion. Each child was examined on two dierent days. Theexaminer M.S. performed the rst examination, and theexamination time was measured using a digital chronometer,which was started as soon as the dental mirror was placed inthe childs mouth. Immediately following the rst examina-tion, V.S. carried out the second examination.Oneweek later,M.S. repeated the examination.The examiners were trained for 2 wk by two of the

authors of the criteria (Drs B. Nyvad and V.Machiulskiene),with the training being based on clinical examinations and

A-a

B-a

C-a

A-b

B-b F-a F-b

G

I

H

J

C-b

A-c E-a E-b E-c

D-a D-b D-c

Fig. 1. Clinical aspects of the Nyvad caries diagnostic codes in primary teeth. A-a, A-b, and A-c: code 0, sound surfaces; B-a and B-b:code 1, active non-cavitated lesions; C-a and C-b: code 2, active enamel discontinuity; D-a, D-b, and D-c: code 3, active cavitatedlesions; E-a, E-b, and E-c: code 4, inactive non-cavitated lesions; F-a and F-b: code 5, inactive enamel discontinuity; G: code 6, inactivecavitated lesion; H: code 7, lling; I: code 8, lling associated with an active lesion; J: code 9, lling associated with an inactive lesion.

226 Sellos & Soviero

discussions. Calibration was performed in a pilot studyinvolving 30 children, which was carried out under the sameconditions as the present study.

Evaluation

The agreement was rst analyzed at the tooth surface level.Interexaminer and intra-examiner reliability of the cariesdiagnostic codes (09) was assessed. Then, the codes weredichotomized at four category thresholds: (i) sound (code 0)vs. diseased (codes 19); (ii) sound or inactive lesion (codes0, 4, 5, 6, 7, and 9) vs. active lesion (codes 1, 2, 3, and 8); (iii)intact surface (codes 0, 1, 4, 7, and 9) vs. surface disconti-nuity (codes 2, 3, 5, 6, and 8); and (iv) sound or non-cavitated lesion (codes 0, 1, 2, 4, 5, 7, and 9) vs. cavitatedlesion (codes 3, 6, and 8). The results were expressed aspercentage agreement and Cohens kappa coecient foreach diagnostic threshold.At the individual level, the caries prevalence (the per-

centage of children with caries) and the extent [the decayedor lled surface (dfs) count for each child] was estimatedconsidering: (i) all caries lesions (codes 19); (ii) active carieslesions (codes 1, 2, 3, and 8); and (iii) cavitated caries lesions(codes 3, 6, and 8). For the dfs, interexaminer andintra-examiner agreement was assessed using the methoddescribed by Bland & Altman (26). In this analysis, thedierence between the dfs counts obtained in the rst andsecond examination is plotted against the mean count ofboth examinations. Based on the SD of the dierences,upper and lower limits of agreement are calculated, whichresults in an interval where 95% of the dierences betweenthe examinations are found. For caries prevalence, agree-ment was assessed by percentage agreement and Cohenskappa coecient for each diagnostic threshold.

The MannWhitney U-test was used to verify the inu-ence of the childrens age and caries experience on the meanexamination time. The signicance level for all of theanalyses was set at 5% (a = 0.05).

Results

A total of 6,400 tooth surfaces from 80 children weresuitable for the study. Table 1 presents the intra-examiner and interexaminer percentage agreement andkappa values at the tooth-surface level. In general, kappavalues were 0.80 or higher in all analyses, with theexception of a kappa value of 0.76 when considering allcodes from 0 to 9. The percentage agreement varied from0.96 to 0.99. The highest kappa values (0.95 and 0.98)were observed when a positive diagnosis was based onthe presence of cavitation. Lower kappa values (0.80 and0.86) were observed when a positive diagnosis was basedon the presence of an active lesion. In all the analyses theintra-examiner agreement had slightly higher values thanthe interexaminer agreement.Among the 6,400 tooth surfaces examined, disagree-

ments between examiners were observed in 242 (3.8%)cases. Most of these disagreements (65.3%; 158/242)concerned the dierentiation between sound surfaces andnon-cavitated lesions, representing 2.5% (158/6,400) ofall tooth surfaces. From the total number of disagree-ments, 33.5% (81/242) were related to the dierentiationbetween sound surfaces and inactive non-cavitatedlesions, representing 1.3% (81/6,400) of all tooth sur-faces; 26.0% (63/242) concerned disagreement between

Table 1

Percentage agreement, kappa coecient values, and respective 95% CIs for interexaminer (V.S. M.S. 1st) and intra-examiner(M.S. 1st M.S. 2nd) examinations considering all codes and four diagnostic thresholds (n = 6,400 tooth surfaces)

Interexaminer(V.S. M.S. 1st)

Intra-examiner(M.S. 1st M.S. 2nd)

Nyvad criteria Percentage agreement = 96 Percentage agreement = 97(codes 09) j = 0.76 j = 0.83Diagnostic thresholds ) + ) +1. Sound vs. diseased (n) ) 5794 91 5813 58

+ 77 438 74 455Percentage agreement (95% CI) 97 (9798) 98 (9798)Kappa (CI) 0.82 (0.800.85) 0.86 (0.840.86)

2. Sound or inactive lesions vs. active lesions (n) ) 6020 76 6031 35+ 46 258 53 281

Percentage agreement (95% CI) 98 (9798) 99 (9899)Kappa (95% CI) 0.80 (0.760.83) 0.86 (0.830.89)

3. Intact surface vs. surface discontinuity (n) ) 6090 35 6096 12+ 18 257 21 271


4. Sound or non-cavitated lesion vs. cavitated lesion (n) ) 6192 9 6198 5+ 11 188 3 194


Code 0, sound surfaces; code 1, active non-cavitated lesions; code 2, active enamel discontinuity; code 3, active cavitated lesions; code4, inactive non-cavitated lesions; code 5, inactive enamel discontinuity; code 6, inactive cavitated lesion; code 7, lling; code 8, llingassociated with an active lesion; code 9, lling associated with an inactive lesion; +, positive diagnoses; ), negative diagnoses.

Caries assessment in primary teeth 227

sound surfaces and active non-cavitated lesions, repre-senting 0.9% (63/6,400) of all tooth surfaces; and 5.8%(14/242) concerned disagreement between active non-cavitated lesions and inactive non-cavitated lesions,representing 0.2% (14/6,400) of all tooth surfaces. Theremaining disagreements (34.7%; 84/242) were related toother combinations. The cross-tabulation showing theinterexaminer agreement is available as supportinginformation (Table S1).Disagreements in intra-examiner examinations were

observed in 171 (2.7%) of the 6,400 tooth surfaces andthe majority (70.7%; 121/171) were again related to thedierentiation between sound surfaces and non-cavitatedlesions. The cross-tabulation showing the intra-examineragreement is available as supporting information(Table S2).Comparison between examinations was also made

based on caries prevalence (Table 2) and dfs count(Table 3), which are variables at the individual level.Considering the presence of at least one surface withcaries (codes 19) or with cavitated lesions (scores 3, 6, or8), the percentage agreement on caries prevalence washigh, ranging from 93.8% to 98.7% with correspondingkappa values ranging from 0.84 to 0.97. The lowestagreement was observed when caries diagnosis at theindividual level was based on the presence of at least onesurface with an active lesion (codes 1, 2, 3, or 8). In thiscase, the interexaminer percentage agreement was 85%(j = 0.69) and the intra-examiner percentage agreementwas 87.5% (j = 0.74).Table 3 shows the mean dfs (i.e. the number of tooth

surfaces aected by caries per subject) for both examin-

ers according to three diagnostic thresholds. In no situ-ation was the mean dfs statistically signicantly dierentamong examinations. Table 4 shows the data related tothe analysis of the interexaminer and intra-examineragreement on dfs counts. The mean dfs dierencebetween examinations was below 1 in all situations,indicating a high level of agreement for the three diag-nostic thresholds. The limits of agreement denote theinterval which holds 95% of the dierence: the smallerthe range between the limits, the higher the agreement.The narrowest interval was observed when the diagnosticthreshold was set at the cavity level (from )0.72 to 0.67surfaces for intra-examiner analysis, and from )1.40 to1.60 surfaces for interexaminer analysis), indicating veryhigh agreement. The wider interval (from )4.90 to 4.60surfaces) was seen for interexaminer agreement when alltypes of lesions were included in the dfs counts. Whenonly active caries lesions were included in the dfs counts,the limits of agreement ranged from )3.90 to 3.20surfaces for interexaminer analysis and from )3.20 to3.60 surfaces for intra-examiner analysis.The mean examination time was 226.5 s (SD =

128.5 s) and was not inuenced by the childrens age,whereas it was inuenced by their caries experience(Table 5).

Discussion

This study assessed the interexaminer and intra-examineragreement on caries diagnosis in primary teeth using theNyvad classication system. Usually, agreement in caries

Table 2

Interexaminer (V.S. M.S. 1st) and intra-examiner (M.S. 1st M.S. 2nd) reliability in the assessment of caries prevalence(percentage of children with caries) according to three diagnostic thresholds (n = 80 subjects)

Diagnostic thresholds

Interexaminer estimate Intra-examiner estimate

V.S.M.S.1st Agreement j

M.S.1st

M.S.2nd Agreement j

At least one surface aected by caries (codes 19) 72.5 73.8 93.8 0.84 73.8 73.8 97.5 0.94At least one surface with active caries lesions (codes 1, 2, 3, or 8) 58.8 61.3 85.0 0.69 61.3 61.3 87.5 0.74At least one surface with cavitated caries lesions (codes 3, 6, or 8) 40.0 37.5 97.5 0.95 37.5 38.8 98.7 0.97

Code 0, sound surfaces; code 1, active non-cavitated lesions; code 2, active enamel discontinuity; code 3, active cavitated lesions; code4, inactive non-cavitated lesions; code 5, inactive enamel discontinuity; code 6, inactive cavitated lesion; code 7, lling; code 8, llingassociated with an active lesion; code 9, lling associated with an inactive lesion.

Table 3

Mean decayed or lled surface (dfs) and SD for examiners V.S. and M.S (1st and 2nd examinations) according to three diagnosticthresholds (n = 80 children)

Diagnostic thresholds

Mean dfs (SD)

V.S. M.S. 1st M.S. 2nd

Tooth surfaces aected by caries (codes 19) 6.59 (8.57) 6.71 (8.44) 6.53 (8.33)Tooth surfaces with active lesions (codes 1, 2, 3, or 8) 3.80 (5.85) 4.18 (5.99) 3.95 (5.86)Tooth surfaces with cavitated lesions (codes 3, 6, or 8) 2.55 (4.63) 2.46 (5.53) 2.49 (4.53)



diagnosis is analyzed at the tooth surface level whencodes are assigned to each surface. Although this type ofanalysis is important to identify how reproducible themethod is, it does not show the impact of agreement levelon epidemiological data.No two examiners, and not even the same examiner,

are expected to agree completely or give identical resultsin repeated examinations. However, it is most importantto know whether this disagreement could cause problemsin clinical practice (i.e. cause the clinician to choose aninappropriate treatment, or bring about a misleadingconclusion to a clinical study) (26).The high reproducibility of the Nyvad caries classi-

cation system in the present study was in accordancewith results from studies carried out on permanent teethin young individuals (8, 24). As also observed in previousstudies (8), disagreements were concentrated on thedierentiation between sound surfaces (code 0) andnon-cavitated lesions (codes 1 and 4). Therefore,non-cavitated stages still represent the major problem inclinical diagnosis.Despite its common usage, the Cohen kappa coe-

cient is not always the most suitable indicator forassessing agreement. It was originally proposed to mea-sure agreement between two examiners when subjects areclassied into two nominal categories. The extension ofits application to multicategory data may result in mis-

leading interpretations. When kappa is used for multi-category data, the categories should be grouped tobecome dichotomies. Even so, kappa values can behighly inuenced by the way in which multicategoryclassications are grouped. (27)A more comprehensive analysis should also focus on

the disagreements and their possible impact on the -nal outcomes. Thereby, it would provide a betterunderstanding of the diculties related to each level ofdiagnosis and allow for a more profound comparisonbetween studies. A detailed analysis of disagreementsmay identify some of their eects on clinical researchand practice. In the present study, most of the dis-agreements were between sound surfaces (code 0) andnon-cavitated lesions (codes 1 and 4). In the case ofinactive lesions, this would not result in any over-treatment in a clinical situation, because neither soundsurfaces nor inactive lesions require any intervention.This disagreement has often been related to stainedssures on occlusal surfaces. In the case of active le-sions, it could be a relevant disagreement in terms ofthe clinical practice as it could change the treatmentdecision. Although total agreement is not expected,disagreements must be avoided as much as possible. Itis important to reinforce that improvement in agree-ment is highly related to the quality of training andexperience of the examiners (8, 10, 22, 23).

Table 5

Mean examination time in seconds (s) according to age and number of aected tooth surfaces

n Mean time (s) SD Min. Max.

Age*3650 months 24 203.29a 136.17 47 6015170 months 29 224.17b 134.75 54 586Older than 71 months 27 249.74c 114.78 66 525

Number of aected tooth surfaces**0 22 106.64a 45.57 47 22616 29 191.72b 58.38 74 313 7 29 352.31c 112.74 178 601

Max., longest examination time; Min., shortest examination time.*MannWhitney U-test (a b; a c; b c: P > 0.05).**MannWhitney U-test (a b; a c: P < 0.01); (b c: P = 0.01).

Table 4

Interexaminer (V.S. M.S. 1st) and intra-examiner (M.S. 1st M.S. 2nd) agreement in the assessment of number of tooth surfacesaected by caries per subject [decayed or lled surface (dfs)] according to three dierent thresholds (n = 80 subjects)

Diagnostic thresholds Examiners

Mean dfsdierenceper subject

Limits ofagreement

Range of the dfsdierences betweenthe two examinations

Surfaces aected by caries (codes 19) V.S. M.S. 1st )0.13 [)4.90; 4.60] [)6; 14]M.S. 1st M.S. 2nd 0.20 [)3.50; 3.90] [)9; 6]

Surfaces with active lesions (codes 1, 2, 3, or 8) V.S. M.S. 1st )0.40 [)3.90; 3.20] [)8; 6]M.S. 1st M.S. 2nd 0.22 [)3.20; 3.60] [)10; 6]

Surfaces with cavitated lesions (codes 3, 6, or 8) V.S. M.S. 1st 0.03 [)1.01; 1.06] [)2; 5]M.S. 1st M.S. 2nd )0.03 [)0.72; 0.67] [)2; 1]



For numerical variables, such as dfs counts, the Bland& Altman method (26) is considered an appropriatemethod with which to assess agreement between repeatedmeasurements: the narrower the interval, the higher theagreement between measurements. As expected, thenarrowest interval of agreement was found to be whenthe cut-o point was set at cavitated lesions. However,for the other two cut-o points, considering all codes oronly active lesions, the limits of agreements were notexcessively wide and the nal mean dfs count did notdier signicantly between examiners or within anexaminer (Tables 4 and 5). These results encourage theuse of more comprehensive assessment of carious lesionsin clinical studies.In our opinion, the main advantage of the Nyvad caries

classication system is the ability to assess the progressionstage and activity of the lesions simultaneously, using avery straightforward code system. In addition, thedierentiation between enamel discontinuity and a dentincavity provides important information for monitoringcaries progression. The theoretical underpinning of theNyvad caries classication system is to provide a dis-cernible link between the diagnosis and the best treatmentoption.Although dental biolm has to be removed from the

tooth surfaces to allow proper visualization of earlycarious lesions, professional cleaning of the teeth is notrecommended before examination using the Nyvad sys-tem. Most of the biolm is usually removed duringtoothbrushing, and only those surfaces where the patientdoes not usually clean eectively will remain covered bybiolm. In the present study, children were assisted bythe examiner during toothbrushing because they werevery young.When visual examination combining surface features

(opacity, roughness, colour, and location of the lesion) isnot enough to classify activity, tactile examination usinga probe is recommended (8). For a proper understandingof the surface texture, the probe must be sharp.However, the intention is not to test whether the probecatches irregularities on the enamel, but rather to feelthe texture or the consistency of the lesion. For enamellesions, the tip of the probe must be placed at an angle ofabout 30 to the tooth surface and be moved gentlyacross the lesion, so that the dierence between thesmooth texture of a sound surface or inactive lesion andthe rough texture of an active lesion can be felt. Fordentin lesions, tactile examination, using slight pressure,dierentiates hard tissue from soft tissue or a leatheryconsistency. In fact, surface texture has been consideredas a better indicator of activity than colour (7, 20). Thatis why activity assessment may not be based only on thecolour of the lesion, especially for dentin lesions. Manydark-brown dentin lesions have a leathery consistency,indicating that they are still active. As the transitionbetween active and inactive stages does not occurinstantaneously, mixed lesions must be considered asactive. We believe that all these considerations onactivity assessment were of great importance for the highlevels of agreement achieved on activity assessment in thepresent study. Disagreement on the dierentiation

between active and inactive lesions was very infrequent,indicating a good reproducibility of activity assessmentwhen a consensus on the presence of the lesion wasreached.For proper detection of non-cavitated lesions, tooth

surfaces must be dried and visualized under goodillumination. The requirement of compressed air andarticial light is therefore a prerequisite for the Nyvadcaries classication system, as well as for other cariesclassication systems that aim to detect initial cariouslesions; otherwise, non-cavitated lesions are underesti-mated.The mean examination time is not frequently men-

tioned in reliability studies in the literature. In thepresent study, the mean time needed for the examina-tion was less than expected. The average examinationtime for each child was 3 min and 46 s. In a previousstudy, with older children and adolescents, the exami-nation time was estimated to be between 5 and 8 min(8). This dierence was probably because more toothsurfaces are present in mixed or permanent dentitionscompared with the primary dentition. We expected thatyounger children would need more time for examina-tion because it is often more dicult to control theirbehaviour in the dental chair. However, age did notinuence examination time. Probably, behaviour con-trol was not a problem in the present study becauseboth examiners were specialized in paediatric dentistry.On the other hand, the more the tooth surfaces wereaected by caries the longer the examination took be-cause, if a tooth surface was not sound, examinersneeded additional time to reect and assign a code tothat surface.This present study concludes that the Nyvad caries

classication system showed a high level of agreementand suitable examination time, and may be consideredreliable for dental caries clinical studies in primary teeth.More studies on the reliability of the system should beperformed with larger samples and in populations withdierent rates of caries prevalence to conrm thesendings.

Acknowledgements We gratefully acknowledge Professors BenteNyvad (Denmark) and Vita Machiulskiene (Lithuania) for their

generous introduction to the clinical use of the criteria, Prof. Vibeke

Baelum (Denmark) for her relevant suggestions on statistical

analysis and Soraya Leal (Brazil) for her contribution to the nal

version of this manuscript.

References1. Baelum V, Heidmann J, Nyvad B. Dental caries paradigms in

diagnosis and diagnostic research. Eur J Oral Sci 2006; 114:263277.

2. Ismail AI. Clinical diagnosis of precavitated carious lesions.Community Dent Oral Epidemiol 1997; 25: 1323.

3. Nyvad B. Diagnosis versus detection of caries. Caries Res 2004;38: 192198.

4. Pitts NB. ICDAS an international system for cariesdetection and assessment being developed to facilitate cariesepidemiology, research and appropriate clinical management.Community Dent Health 2004; 21: 193198.


5. Ismail AI. Visual and visuo-tactile detection of dental caries.J Dent Res 2004; 83: C56C66.

6. Thylstrup A, Bruun C, Holmen L. In vivo caries models mechanisms for caries initiation and arrestment. Adv Dent Res1994; 8: 144157.

7. Nyvad B, Fejerskov O. Assessing the stage of caries lesionactivity on the basis of clinical and microbiological examina-tion. Community Dent Oral Epidemiol 1997; 25: 6975.

8. Nyvad B, Machiulskine V, Baelum V. Reliability of a newcaries diagnostic system differentiating between active andinactive caries lesions. Caries Res 1999; 33: 252260.

9. Machiulskine V, Nyvad B, Baelum V. A comparison ofclinical and radiographic caries diagnoses in posterior teeth of12yearold Lithuanian children. Caries Res 1999; 33: 340348.

10. Ismail AI, Sohn W, Tellez M, Amaya A, Sen A, Hasson H,Pitts NB. The International Caries Detection and AssessmentSystem (ICDAS): an integrated system for measuring dentalcaries. Community Dent Oral Epidemiol 2007; 35: 170178.

11. Lima TJ, Ribeiro CC, Tenuta LM, Cury JA. Low-fluoridedentifrice and caries lesion control in children with differentcaries experience: a randomized clinical trial. Caries Res 2008;42: 4650.

12. Assaf AV, Meneghim MC, Zanin L, Mialhe FL, PereiraAC, Ambrosano GMB. Assessment of different methods fordiagnosing dental caries in epidemiological surveys. CommunityDent Oral Epidemiol 2004; 32: 418425.

13. Parfitt GJ. A standard clinical examination of the teeth.Br Dent J 1954; 96: 296300.

14. Backer-Dirks O. Longitudinal dental caries study in children915 years of age. Arch Oral Biol 1961; 6: 94108.

15. Pitts NB, Fyffe HE. The effect of varying diagnostic thresh-olds upon clinical caries data for a low prevalence group. J DentRes 1988; 67: 592596.

16. Neilson A, Pitts NB. The clinical behaviour of free smoothsurface carious lesions monitored over 2 years in a group ofScottish children. Br Dent J 1991; 171: 313318.

17. Pitts NB. Diagnostic tools and measurements impact onappropriate care. Community Dent Oral Epidemiol 1997; 25: 2435.

18. EkstrandKR, Ricketts DNJ, Kidd EAM, Qvist V, Schou S.Detection, diagnosing, monitoring and logical treatment ofocclusal caries in relation to lesion activity and severity: an invivo examination with histological validation. Caries Res 1998;32: 247254.

19. WHO. Oral Health Surveys: basic methods, 4 edn. Geneva:WHO, 1997.

20. Holmen L, Thylstrup A, Artun J. Clinical and histologicalfeatures observed during arrestment of active enamel cariouslesions in vivo. Caries Res 1987; 21: 546554.

21. Ekstrand KR, Martignon S, Ricketts DJN, Qvist V.Detection and activity assessment of primary coronal carieslesions: a methodologic study. Oper Dent 2007; 32: 225235.

22. Nyvad B, Machiulskine V, Baelum V. Construct and pre-dictive validity of clinical caries diagnostic criteria assessinglesion activity. J Dent Res 2003; 82: 117122.

23. Gonzalez MC, Ruz JA, Fajardo MC, Gomez AD, MorenoCS, Ochoa MJ, Rojas LM. Comparison of the def index withNyvads caries diagnostic criteria in 3-and 4-year-old Colom-bian children. Pediatr Dent 2003; 25: 132136.

24. Machiulskine V, Baelum V, Fejerskov O, Nyvad B. Preva-lence and extent of dental caries, dental fluorosis, and devel-opmental enamel defects in Lithuanian teenage populationswith different fluoride exposures. Eur J Oral Sci 2009; 117: 154160.

25. Machiulskine V, Nyvad B, Baelum V. Prevalence andseverity of dental caries in 12-year-old children in Kaunas,Lithuania 1995. Caries Res 1998; 32: 175180.

26. Bland JM, Altman DG. Statistical methods for assessingagreement between two methods of clinical measurement.Lancet 1986; 327: 307310.

27. Maclure M, Willett WC. Misinterpretation and misuse ofthe kappa statistic. Am J Epidemiol 1987; 126: 161169.

Supporting informationAdditional Supporting Information may be found in the onlineversion of this article:

Table S1. Distribution of diagnoses at the inter-examiner examin-ations based on Nyvad criteria.

Table S2. Distribution of diagnoses at the intra-examiner examin-ations based on Nyvad criteria.

Please note: Wiley-Blackwell is not responsible for the content orfunctionality of any supporting materials supplied by the authors.Any queries (other than missing material) should be directed to thecorresponding author for the article.


This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy.Users should refer to the original published version of the material.

Documents

Reliablity of Nyvads