Sbu • Statens beredning för medicinsk utvärdering The Swedish Council on Technology Assessment in Health Care

Framtida forskning

Det finns ett stort behov av standardisering av forskningsdesign för  

in vitro-tester för kariesdiagnostik. Det gäller bl a krav på metoder för 

validering och definitioner av vad som är karies vid valideringen. Karies-

prevalensen i tandmaterialet ska uppges. Dessutom behövs en standard 

för hur olika tänder och fördelningen av kariesskador med olika omfatt-

ning och djup ska vara representerade. Nya diagnostiska metoder som 

lanseras bör karakteriseras genom en sådan in vitro-testning innan de  

tas i kliniskt bruk.

Tandläkare har behov av att validera sina egna positiva kariesdiagnoser, 

särskilt om dessa behandlas med fyllningsterapi. En systematisk regist-

rering av egna kliniska fynd avseende skadans utsträckning kan ge god 

feedback på den egna kvaliteten i behandlingen. Praxisnära forskning 

baserad på insamling av sådan information från många kliniker skulle 

kunna ge en standard för ”god” diagnostik. Detta skulle kunna utgöra 

ett rättesnöre i den kliniska verksamheten.

Det är ett känt fenomen att kariesdiagnostik varierar från kliniker till klini-

ker med resultatet att patienter får olika behandling. Behovet av standard-

isering är därför betydande. Ökad kunskap om vilka strategier och metoder 

som är mest effektiva för kalibrering av kliniker kräver ny forskning.

K A R I E S   –   d I A G N O S T I K ,   R I S K B E d ö M N I N G   O C H   I C K E - I N vA S I v   B E H A N d L I N G

140

Table 4.2.2 Approximal caries. Studies with high and medium quality and relevance.

Author

Year,  

reference

Country

Type of 

study

Num-

ber of 

obser-

vers

Observer  

reliability

Population

Type of lesion  

and prevalence

Number  

of teeth 

and sur-

faces

Method Com-

parison 

method

Outcome 

measure

Results 

enamel 

caries

Results 

dentine 

caries

Study 

quality and 

relevance

Comments

Approximal caries

Mileman et al

1990 [30]*

The  

Netherlands

In vitro

276

Mean kappa: 0.53 Extracted 

teeth

Approximal caries, 

dentine caries 

45/105 (43%)

105 sur-

faces

BW

Histology

Se, Sp 

Caries in 

outer 1/2 

of dentine

Se: 54%

Sp: 97%

Medium

Copies of 

radiographs 

used. Insuf-

ficient data 

reporting

Ricketts et al

1997 [21]*

United 

Kingdom

In vitro

5

% agreement

intra-observer: 

94

Freshly  

extracted 

teeth

Approximal caries, 

24% enamel caries,

13% dentine caries.

11 (12%) of lesions

were cavitated

96 teeth 

(48 pre-

molars,  

48 molars)

180 sur-

faces

BW

Histology

Se, Sp,  

% agree-

ment

d1

Se: 22%

Sp: 97%

d3

Se: 13%

Sp: 100%

Medium

Insufficient

description 

of lesion 

severities

Occlusal caries

Angnes et al

2005 [27]

Brazil

In vivo

2

Kappa inter-

observer 0.63

Adult 

patients 

(19–35 yrs)

Occlusal caries, 

24/110 outer 1/2

enamel, 50 inner 

1/2–outer 1/3 den-

tine, 14 middle  

1/3 dentine,  

2 inner 1/3 dentine  

(16 dentine lesions)

38 

patients, 57 

3rd molars, 

110 occlu-

sal sites, 

no frank 

cavitation

BW

Histology, 

hemi-

section 

(blinded,

1 obser-

ver)

Se, Sp, 

AUC, 

kappa 

statistics

By radio-

graphy: 

caries in 

outer 1/3 

of dentine

Se: 3%

Sp: 98%

Medium

Not repre-

sentative 

teeth (eg

3rd molars)

Cayley et al

1997 [35]*

United 

Kingdom

In vitro

11

Intra-examiner 

(n=11)  

Before calib: 0.4

After calib: 0.62

Inter-examiner

Before calib: 0.30

After calib: 0.27

Extracted 

teeth

Occlusal caries, 33 

of 60 had no caries 

or caries restricted 

to enamel = sound,

27/60 had caries in

dentine

60 molars, 

30 upper 

and 30 

lower with-

out obvious 

cavitation 

or resto-

ration  

60 surfaces

Combi-

ned VI

and BW

Histology 

(cut off:

dentine 

caries)

Se, Sp, 

kappa 

(5-graded

confidence 

score 

about 

caries 

being 

present)

Se: 65%

Sp: 73%

Medium

Ekstrand 

et al

1997 [34]*

Denmark,

United  

Kingdom

In vitro

3

Observer 0.79;

inter-observer 

0.59

Extracted 

teeth

Occlusal caries 

13/100 outer 1/2

enamel, 24 between

inner 1/2 enamel

and outer 1/3 den-

tine, 23 middle 1/3

dentine, 16 inner  

1/3 dentine

80 molars, 

20 pre-

molars

100 surfaces

BW

Histology 

(2 obser-

vers)

Se, Sp,  

corre-

lation,  

% agree-

ment

Histological caries 

threshold betw score 2

(inner enamel–outer  

3rd of dentine) and

score 3.

BW 

Se: 54%

Sp: 100%

High

141

K A P I T E L   4   •   d I A g n o s T I K

Author

Year,  

reference

Country

Type of 

study

Num-

ber of 

obser-

vers

Observer  

reliability

Population

Type of lesion  

and prevalence

Number  

of teeth 

and sur-

faces

Method Com-

parison 

method

Outcome 

measure

Results 

enamel 

caries

Results 

dentine 

caries

Study 

quality and 

relevance

Comments

Approximal caries

Mileman et al

1990 [30]*

The  

Netherlands

In vitro

276

Mean kappa: 0.53 Extracted 

teeth

Approximal caries, 

dentine caries 

45/105 (43%)

105 sur-

faces

BW

Histology

Se, Sp 

Caries in 

outer 1/2 

of dentine

Se: 54%

Sp: 97%

Medium

Copies of 

radiographs 

used. Insuf-

ficient data 

reporting

Ricketts et al

1997 [21]*

United 

Kingdom

In vitro

5

% agreement

intra-observer: 

94

Freshly  

extracted 

teeth

Approximal caries, 

24% enamel caries,

13% dentine caries.

11 (12%) of lesions

were cavitated

96 teeth 

(48 pre-

molars,  

48 molars)

180 sur-

faces

BW

Histology

Se, Sp,  

% agree-

ment

d1

Se: 22%

Sp: 97%

d3

Se: 13%

Sp: 100%

Medium

Insufficient

description 

of lesion 

severities

Occlusal caries

Angnes et al

2005 [27]

Brazil

In vivo

2

Kappa inter-

observer 0.63

Adult 

patients 

(19–35 yrs)

Occlusal caries, 

24/110 outer 1/2

enamel, 50 inner 

1/2–outer 1/3 den-

tine, 14 middle  

1/3 dentine,  

2 inner 1/3 dentine  

(16 dentine lesions)

38 

patients, 57 

3rd molars, 

110 occlu-

sal sites, 

no frank 

cavitation

BW

Histology, 

hemi-

section 

(blinded,

1 obser-

ver)

Se, Sp, 

AUC, 

kappa 

statistics

By radio-

graphy: 

caries in 

outer 1/3 

of dentine

Se: 3%

Sp: 98%

Medium

Not repre-

sentative 

teeth (eg

3rd molars)

Cayley et al

1997 [35]*

United 

Kingdom

In vitro

11

Intra-examiner 

(n=11)  

Before calib: 0.4

After calib: 0.62

Inter-examiner

Before calib: 0.30

After calib: 0.27

Extracted 

teeth

Occlusal caries, 33 

of 60 had no caries 

or caries restricted 

to enamel = sound,

27/60 had caries in

dentine

60 molars, 

30 upper 

and 30 

lower with-

out obvious 

cavitation 

or resto-

ration  

60 surfaces

Combi-

ned VI

and BW

Histology 

(cut off:

dentine 

caries)

Se, Sp, 

kappa 

(5-graded

confidence 

score 

about 

caries 

being 

present)

Se: 65%

Sp: 73%

Medium

Ekstrand 

et al

1997 [34]*

Denmark,

United  

Kingdom

In vitro

3

Observer 0.79;

inter-observer 

0.59

Extracted 

teeth

Occlusal caries 

13/100 outer 1/2

enamel, 24 between

inner 1/2 enamel

and outer 1/3 den-

tine, 23 middle 1/3

dentine, 16 inner  

1/3 dentine

80 molars, 

20 pre-

molars

100 surfaces

BW

Histology 

(2 obser-

vers)

Se, Sp,  

corre-

lation,  

% agree-

ment

Histological caries 

threshold betw score 2

(inner enamel–outer  

3rd of dentine) and

score 3.

BW 

Se: 54%

Sp: 100%

High

The table continues on the next page

K A R I E S   –   d I A G N O S T I K ,   R I S K B E d ö M N I N G   O C H   I C K E - I N vA S I v   B E H A N d L I N G

142

Author

Year,  

reference

Country

Type of 

study

Num-

ber of 

obser-

vers

Observer  

reliability

Population

Type of lesion 

and prevalence

Number 

of teeth 

and sur-

faces

Method Com-

parison 

method

Out-

come 

measure

Results 

enamel 

caries

Results 

dentine 

caries

Study 

quality and 

relevance

Comments

Lussi 1993

[32]*

Switzerland

In vitro

24

BW inter-obser-

ver 0.45; intra-

observer 0.55

Extracted 

teeth

Occlusal caries 

clinically intact, 

9/63 in enamel,

33/63 dentine

caries

52 molars 

and 11 

premolars 

with occlu-

sal surfaces 

which were

clinically 

intact

BW

Histology

Se, Sp, 

LR+,

kappa

Dentine 

caries

Se: 45%

Sp: 83%

High

Lussi

1993 [32]*

Switzerland

In vitro

10

VI+BW inter-

observer 0.46

Extracted 

teeth

Occlusal caries 

clinically intact, 

9/63 in enamel,

33/63 dentine

caries

52 molars 

and 11 

premolars 

with occlu-

sal surfaces 

which were

clinically 

intact

Combi-

ned VI

and BW

Histology 

(joint

decision 

by 2 

obser-

vers)

Se, Sp, 

LR+,

kappa

Dentine 

caries

Se: 49%

Sp: 87%

High

Ricketts

et al

1997 [21]*

United 

Kingdom

In vitro

5

% agreement

intra-observer 

88%

Freshly 

extracted 

teeth

Occlusal caries, 

40% enamel caries,

39% dentine caries

96 teeth 

(48 molars,

48 pre-

molars)

96 surfaces

BW

Histology

Se, Sp,  

% agree-

ment

d1

Se: 12%

Sp: 93%

d3

Se: 13%

Sp: 92%

Medium

Insufficient

description 

of lesion 

severities

Rocha et al

2003 [26]

Brazil

In vivo

2

Kappa, inter-

observer 0.53, 

intra-observer 

0.14–0.46

29 patients 

10–11 yr, pri-

mary teeth 

indicated 

for extrac-

tion for 

orthodontic 

reasons

Occlusal caries, 

16/50 in enamel,

(some enamel

lesions may extend 

into outer 1/3 of

dentine, 13/50 in

dentine)

30 1st 

and 2nd 

molars,  

50 sites

BW

Histology

Cross 

tables. Se, 

Sp, kappa

All lesions

Se: 62%

Sp: 73%

Dentine lesions

Se: 96%

Sp: 81%

Medium

Few sub-

jects/teeth

Table 4.2.2 continued

* Study selected because the paper was included in the review by [23].

AUC = Area under the curve; BW = Bitewing radiographs; d1 = Enamel caries;  

d3 = Dentine caries; LR = Likelihood ratio; Se = Sensitivity; Sp = Specificity; 

VI = Visual inspection

143

K A P I T E L   4   •   d I A g n o s T I K

Author

Year,  

reference

Country

Type of 

study

Num-

ber of 

obser-

vers

Observer  

reliability

Population

Type of lesion 

and prevalence

Number 

of teeth 

and sur-

faces

Method Com-

parison 

method

Out-

come 

measure

Results 

enamel 

caries

Results 

dentine 

caries

Study 

quality and 

relevance

Comments

Lussi 1993

[32]*

Switzerland

In vitro

24

BW inter-obser-

ver 0.45; intra-

observer 0.55

Extracted 

teeth

Occlusal caries 

clinically intact, 

9/63 in enamel,

33/63 dentine

caries

52 molars 

and 11 

premolars 

with occlu-

sal surfaces 

which were

clinically 

intact

BW

Histology

Se, Sp, 

LR+,

kappa

Dentine 

caries

Se: 45%

Sp: 83%

High

Lussi

1993 [32]*

Switzerland

In vitro

10

VI+BW inter-

observer 0.46

Extracted 

teeth

Occlusal caries 

clinically intact, 

9/63 in enamel,

33/63 dentine

caries

52 molars 

and 11 

premolars 

with occlu-

sal surfaces 

which were

clinically 

intact

Combi-

ned VI

and BW

Histology 

(joint

decision 

by 2 

obser-

vers)

Se, Sp, 

LR+,

kappa

Dentine 

caries

Se: 49%

Sp: 87%

High

Ricketts

et al

1997 [21]*

United 

Kingdom

In vitro

5

% agreement

intra-observer 

88%

Freshly 

extracted 

teeth

Occlusal caries, 

40% enamel caries,

39% dentine caries

96 teeth 

(48 molars,

48 pre-

molars)

96 surfaces

BW

Histology

Se, Sp,  

% agree-

ment

d1

Se: 12%

Sp: 93%

d3

Se: 13%

Sp: 92%

Medium

Insufficient

description 

of lesion 

severities

Rocha et al

2003 [26]

Brazil

In vivo

2

Kappa, inter-

observer 0.53, 

intra-observer 

0.14–0.46

29 patients 

10–11 yr, pri-

mary teeth 

indicated 

for extrac-

tion for 

orthodontic 

reasons

Occlusal caries, 

16/50 in enamel,

(some enamel

lesions may extend 

into outer 1/3 of

dentine, 13/50 in

dentine)

30 1st 

and 2nd 

molars,  

50 sites

BW

Histology

Cross 

tables. Se, 

Sp, kappa

All lesions

Se: 62%

Sp: 73%

Dentine lesions

Se: 96%

Sp: 81%

Medium

Few sub-

jects/teeth

K A R I E S   –   d I A G N O S T I K ,   R I S K B E d ö M N I N G   O C H   I C K E - I N vA S I v   B E H A N d L I N G

144

Table 4.2.3 Studies with low quality and relevance.

Author

Year,  

reference

Country

Type 

of 

study

Num-

ber of 

obser-

vers

Observer 

reliability

Popu-

lation

Type of 

lesion and 

prevalence

Number of 

teeth and  

surfaces

Method Com- 

parison 

method

Outcome 

measure

Results  

enamel 

caries

Results  

dentine  

caries

Study quality 

and relevance

Comments

Bamzahim 

et al

2004 [45]

Sweden

In vitro 5 

No infor-

mation

Extrac-

ted teeth 

with

restora-

tions

Secondary 

caries, 

amalgam 

with caries

24/48 (50%),

tooth-

coloured 

restorations 

with caries

10/18 (56%)

66 teeth, 48 

with amalgam

(32 molars, 16

premolars; 28

class I restora-

tions, 11 class II

restorations).

18 with tooth-

coloured 

restorations 

(13 molars, 5

premolars; 7

class I restora-

tions, 11 class II

restorations)

BW

Clinical 

inspection 

after remo-

val of the 

restoration, 

joint deci-

sion by 2 

observers

Se, Sp, 

AUC, 

logistic 

regression

BW

Se: 65%

Sp: 81%

Low

No reliability 

test. Insuf-

ficient descrip-

tion of material 

and methods

No comment 

of the possible 

“hiding” of the

lesion by the 

restorations

Bamzahim 

et al

2005 [46]

Sweden

In vivo

5, 

majo-

rity 

scores

Posterior 

teeth 

with

amalgam 

resto-

rations, 

preval-

ence 

according 

to gold 

standard 

25/51

(49%)

Secondary 

caries

21 patients 

(19–45 yrs), 51

restorations. 

Location of

caries not spe-

cified and that 

is crucial point

BW

Clinical 

inspection 

after remo-

val of the 

restoration, 

joint deci-

sion by 2 

observers

Se, Sp

Radiography 

(gold standard: 

caries visually 

and tactile 

consistency)

Se: 56%

Sp: 92%

Low

Few subjects/

teeth. No relia-

bility test

No comment 

of “hiding” of

the lesion by 

radiopaque 

restoration 

(amalgam)

Grossman 

et al

2002 [47]

South 

Africa, 

Brazil, 

Canada

In vitro 4

Repeated

diagnosis, 

but no 

informa-

tion about 

reliability

Extrac-

ted teeth

Occlusal 

caries, 

36/139 (26%)

in enamel, 

44 (32%) in

dentine

139 occlusal 

surfaces

BW

Histology, 

kappa 0.82

Se, Sp, 

PPV, NPV

Enamel 

diagnosis

Se: 26%

Sp: 93%

PPV: 0.82

NPV: 0.53

Dentine 

diagnosis

Se: 38%

Sp: 93%

PPV: 0.76

NPV: 0.74

Low

No reliability 

test. Insuf-

ficient descrip-

tion of material 

and methods

145

K A P I T E L   4   •   d I A g n o s T I K