Community Dent. Oral Epidemiol 1977: 5; 116-120

(Key words: epideviiology, oral; tooth, .stain)Methodology

A new index for measuringextrinsic stain in clinical trials

LINDA SHAW AND J. J. MURRAY

Department oj Chitdren's Dentistry, Institute oj Dental Surgery, Eastman Dentat Hospitat,London, Engtand

ABSTRACT - A Staining index has been proposed which is simple to use clinically and yet issensitive enough to detect small changes in staining levels between different groups. Accu-rate scale drawings from An attas oj tooth jorm were reproduced. Outlines of the labial andlingual surfaces of all eight incisor teeth were enlarged to scale (magnification X 4) andeach tooth face divided into 4-mni squares. All areas of extrinsic stain were drawn by theexaminer on to the grid system. This method has been tested in three clinical trials aud thereproducibility investigated. In 161 duplicate examinations a total of 1,830 stained sc[uareswere scored by one examiner, compared with 1,853 squares by the second examiner; the re-producibility ratio was 0.155 and the coefficient of correlation was 0.956. The methodproved sufficiently sensitive to record differences in staining levels iu groups using two den-tifrices; one was a normal commercial product with an abrasivity against dentin approxi-mately two-thirds that of the other paste. The proposed Extrinsic Stain Index provides datawhich can be analyzed by applying parametric or nonparametric tests.

(Aeeepted jor pubtieation 15 January 1977) ^

In a recent review of indices for measurementof soft accumulations in clinical studies of oralhygiene, MANDEL^" states that examination andscoring of stain are very difficult and that the dataon techniques for in vivo measurement are toomeagre to attempt definitive recommendations. Sev-eral investigators report the presence of pigmentationof the teeth 'with the use of topical applications ofstannous fluoride''''''^-"•'-'''. SNYDER^' describes a"Diversified Dental Index (Pigmentation)" andHOROWITZ & CHAMBERLIN" developed a "DentalPigmentation Index" for recording staining that oc-curred in clinical trials of stannous fluoride agentson children. Stain was assessed by tooth surface butquantitative estimates of the area covered were notmade. An attempt at an accurate and detailed eval-uation of extrinsic stain diagnosed clinically in adultswas made by LOBENE'*. This index, with mitiormodifications, has also been used by ABBE, BRIDGE,

RIBBONS, DEAN & LAZAROU'. However, accordingto LOBENE" a major disadvantage of the method isthat moderately to highly stained teeth are necessaryto permit statistical verification of results. DAVIS &

describe a method of evaluating stain by es-titnating the percentage area of labial incisor sur-faces covered by stain and checking this esti-mate against planimeter measurements from photo-graphs. Limited information is available on the re-producibility and sensitivity of this method.

There is a need to formulate an index to measureextrinsic stain clinically which fulfils the general re-quirements for all indices suggested by DAVIES^.

Thus, a staining index should be: 1) simple to useand should permit the study of a large number ofpersons in a limited amount of time and at mini-nrnm cost; 2) highly reproducible both within andbetween examiners; 3) sensitive enough to detectsmall changes in staining levels between differentgroups; and 4) amenable to statistical analysis.

MATERIAL AND METHODSDESCRIPTION OF EXTRINSIG STAIN INDEXAccurate scale drawings from An attas oj tooth jorm^^^were reproduced. Outlines of the labial and lingual sur-faces of all eight incisor teeth were enlarged to scale (mag-nification X 4). Each tooth face was divided into 4-mm

New index jor extrinsic stain 117

51LABIAL

76 76

f

H i

V

\

14

/ ,

/-

I

\

37 42

j

1

78

\

//

/

51

pN\\

44 38 38

LINGUALFig. 1. Design of chart.

squares There was a total of 412 squares on the labialsurfaces and 422 squares ou the lingual surfaces. The de-sign of the chart used is shown in Fig. 1. The subjectswere assessed for staining with a dental light and planemouth mirror; the teeth were dried by compressed air be-fore examination and any gross accumulation of plaque ordebris was removed with a gauze dental napkin. All areasof extrinsic discoloration on the incisor teeth were drawnby the examiner on to the grid system and shaded in. Thenumber of squares covered by stain was totaled separatelyfor labial and lingual surfaces. If a square was not entirelycovered by stain, only those where the shaded portion ex-ceeded half the area were counted.

CLINICAL USE OF THE INDEXThe Index has been used in three double-blind clinicaltrials. Two intensive short-term studies involved 19 dentalsurgery assistants (average age 21 years, range 18-25

years) and 40 non-deutal persounel from a technical col-lege (average age 20.3 years, range 16-26). All partici-pants were non-smokers; they received oral hygiene in-struction and scaling and polishing of their teeth. ThePlaque Index^^ and Gingival Index'' scores were thenmonitored and the oral hygiene instructions reinforced asnecessary until minimal scores were recorded in order toachieve a "start from clean" situation. A final prophylaxiswas then given to remove all stain and the subjects weredivided into two groups balanced with respect to theirinitial plaque, gingivitis and staining scores. One groupwas allocated a standard commercial toothpaste and theother group a paste of exactly the same formula but withan abrasion level of approximately two-thirds that of thecommercial paste. Examinations using the stain index werecarried out at monthly intervals for 6 months iu the den-tal surgery assistants trial and for 3 months in the tech-nical college trial. All the examinations for both short-tenn studies were carried out by Examiner 1 (L.S.). Thethird clinical trial involves 1,432 Berkshire schoolchildren(average age 11.9 years, range 11-13 years) in an unsuper-vised 3-year study. Toothpastes with identical formulationsto those used in the short-term trials have been allocatedand the children are being reexamined at yearly intervals.The trial period has not yet been completed and thereforeonly the reproducibility data from this study are available.

REPRODUCIBILITY OF MEASUREMENTDuplicate examinations were carried out at all stages ofthe three clinical trials in order to assess the reproduci-bility of the Index. Of the adults involved in the twoshort-term studies, reexaminations were performed on 83occasions; from the third study, 203 children were re-examined.

Two clinicians (L.S. and J.J.M.) independently exam-ined 44 adult participants and 117 children in order toassess interexaminer reproducibility. Intraexaminer repro-ducibility was evaluated by Examiner 1 (L.S.) who dupli-cated assessments on 20 adult subjects and 44 children,and Examiner 2 (J.J.M.) who reexamined 19 adults and42 children. The subjects involved in the reproducibiltystudies were chosen at random by a third person and theexaminers were unaware that a duplicate assessment was

Table 1. Extent of staining in groups using low- and mod-erate-abrasive toothpastes

Dental surgery Technical collegeassistants staff and students

No. of Stained No. of Stainedsubjects squares subjects squares

Low-abrasivetoothpaste 9 22.4 21 48.4

Moderate-abrasivetoothpaste 10 6.7* 19 13.1

* F < 0 . 0 5 .

118 SHAW AND MURRAY

Table 2. Interexaminer comparison of number of stainedsquares recorded at 44 duplicate examinations on adults

Examiner 1 (L.S.)

Nostain

StainTotalstainedsquares

Examiner 2(JJ-M.)

No stain

Stain

Totalstainedsquares

34,841

54

41

926

967

980

being carried out. The time interval between examinationswas approximately 4-6 hours.

RESULTSCLINICAL TRIALSThe results of the use of the Staining Index in thetwo completed clinical trials are given in Table 1.After 6 months' use of a low-abrasive paste the den-tal surgery assistants had a mean of 22.4 stainedsquares compared with a mean of 6.7 squares inthose participants using a moderately abrasivetoothpaste. The differences between the groups weresignificant {t = 2.43, P < 0 . 0 5 ) . The results fromthe study that involved non-dental personnel from atechnical college show a similar trend but withhigher levels of stain. The 21 subjects allocated thelow-abrasive paste had a mean of 48.4 stainedsquares after 3 months, while the mean for the 19participants in the moderate-abrasive toothpastegroup was 13.1. The differences between the groupswere significant {t = 2.31, P < 0 . 0 5 ) .

Table 3. Interexaminer comparison of number of stainedsquares recorded at 117 duplicate exaniinalions ou chil-dren

Nostain

Examiner 1 (L.S.

StainTotalstainedsquares

Examiner 2(JJ-M.)

No stain

Stain

Totalstainedsquares

96,020

85

85

788

873

873

REPRODUCIBILITYInterexarniner reproducibility - The results for in-terexaminer reproducibility using the Staining In-dex in adults are given in Table 2. Both examinersscored 926 stained squares on the 44 adult subjects.Examiner 1 scored 41 squares with stain that hadnot been recorded by Examiner 2, and Examiner 2designated 54 squares as stained that had not beenscored by Examiner 1. Therefore, a total of 967stained squares was recorded by Examiner 1, anda total of 980 squares by Examiner 2. Using the re-producibility ratio as recommended by the EDI=a 95 '

— = - 5 ^ = 0.103. The theory of measurement re-liability is built up arouud the notion of correlation.The scores recorded by Examiner 1 for each subjectwere therefore correlated against the scores ob-tained by Examiner 2. The resulting coefficient ofcorrelation was -̂ 0.988.

The results for the 117 ehildren assessed by bothexaminers are shown in Table 3. The reproducibil-

Table 4. Intraexaminer comparison of number of stained scjuares recorded at duplicate examinations on adults

Examiuer 1 (L.S.) 20 reexaminations Examiner 2 (J.J.M.) 19 reexaminations

2ndexamination

No stain

Stain

Total stain

Nostain

1st examination

StainTotalstain

16,314

12

34

320

354

3322ndexamination

No stain

Stain

Total stain

Nostain

1st examination

StainTotalstain

15,459

37

42

308

350

345

fti^' . New index jor extrinsic stain 119

Table 5. Intraexaniiuer comparison of number of stained squares recorded at duplicate examinations on children

Examiner 1 (L.S.) 44 reexaminations

2ndexamination

No stain

Stain

Total stain

1st examination

Nostain Stain

Totalstain

36,060

28

36

486

522

514

Examiuer 2 (J.J.M.) 42 reexaminations

2 n d •

examination

No stain

Stain

Total stain

Nostain

1st examination

StainTotalstain

34,844

23

19

142

161

165

ity ratio was = 0.216, and the coefficient of788

correlation was -1-0.940. •Intraexaminer reprodueibility - The results for

intraexaminer reproducibility in adults are given inTable 4. The reproducibility ratio for Examiner 1was ^^ = 0 144, and the correlation coefficient

320

was -1-0.986. The reproducibility ratio for Exam-

iner 2 was- - = 0.256, and the correlation coef-308

ficient was -f-0.958.Intraexaminer reproducibility in children is dem-

onstrated in Table 5. Eor Examiner 1 the repro-64

ducibility ratio was T O F = 0.132, and the corre-

lation coefficient was -f 0.993. The reproducibility

ratio for Examiner 2, on a subsample with very lowlevels of stain,49

= 0.296, and the coefficient

of correlation was -f- 0.928.

DISCUSSIONThe Extrinsic Stain Index is simple to use and re-quires no specialized equipment. The three clinicalstudies have involved 1,499 subjects and in the ma-jority of cases only a short examination time was re-quired. However, even in those with extensive stain-ing the examination time did not exceed 3 min.Therefore many subjects may be studied in a mini-mum time and at minimum cost.

The reproducibility of the method was investi-gated in different populations with a range of agesand propensity to form stain. The reproducibilityratio^ concentrates on inconsistency of diagnosis

rather than on the agreement in diagnosis. Never-theless, the ratios for diagnosis of stain using theExtrinsic Stain Index were more favourable thanthose reported for caries diagnosis. The highcorrelation coefficients show that the examinerswere able to reproduce well the ranking order ofthe subjects. The data from Examiner 2 indieatesthat the between-examiner variation was veryshghtly lower than the within-examiner variation.This may be due to the larger numbers of subjectsreassessed for interexaminer reproducibility than forintraexaminer reproducibility, and the lower preva-lence of stain in the subsample of children seen byExaminer 2. Reproducibility of diagnosis is par-tially dependent on the level of disease.

Indices that are sensitive enough to measuresmall changes in the factor under examination areassociated with greater variability, and often sensi-tivity is reduced in order to achieve better reprodu-cibility of diagnosis*. The clinical trials that werecarried out using the Extrinsic Stain Index were de-signed to test its sensitivity in recording staining ondifferent populations with a range of ages and oralhygiene standards. The results showed significantdifferences between groups using toothpastes withslightly differing abrasiveness. This finding indi-cates that the stain index was sensitive in detectingsmall amounts and small changes in staining.

Theoretically, the Lobene Index^ in commonwith many other clinical indices, can only be ana-lyzed by nonparametric methods as it imposes anumerical interval scale on a gradation of staining.The Extrinsic Stain Index provides data which canbe analyzed by applying either parametric or non-parametric methods.

The proposed new index is simple to use, repro-ducible, sensitive to small changes, and amenable

120 SHAW AND MURRAY

to statistical analysis. It thus fulfils the general re-quirements of an index suggested by DAVIES^

REFERENCES1. ABBE, N . J. v., BMDOE, A. J., RIBBONS, J. W., DEAN,

P. M. & LAZAROU, J. A.: The effect of dentifrices onextrinsic tooth stains. / . Soe. Cosmet. Chem. 1971: 22.-457.

2. DAVIES, G . N . : The different requirements of perio-dontal indices for prevalence studies and clinical trials.Int. Dent. J. 1968: 18: 560-570.

3. DAVIS, W . B. & REES, D . A.: A parametric test tomeasure the cleaning power of toothpaste. / . Soe.Co.smet. Ctiem. 1975: 26: 217-225.

4. HOROWITZ, H . S.: Clinical trials of preventives for deu-tal caries. / . Public Heatth Dent. 1972: 32: 229-233.

5. HoROwrrz, H. S., BAUME, L. J., DIRKS, O . B., DAVIES,

G. N. & SLACK, G . L . (ed.): Principal requirementsfor controlled clinical trials of caries preventive agentsand procedures. Technical Report No. 1 FDI, 1974.

6. HOROWITZ, H . S. & CHAMBERUN, S. R . : Pigmentation

of teeth following topical applications of stannous fluo-ride in a non-fluoridated area. / . Pubtie Heatlh Dent.1971: 31: 32-37.

7. HYDE, E . J. & MuiiLEtt, J. C : Pigmentation of teethtreated with stannous fluoride and its association withcaries incidence and oral hygiene. / . Can. Dent. Assoe.1963: 29: 514-520.

8. LOBENE, R . R . : Effect of dentifrices on tooth stains

with controlled brushing. / . Am. Dent. Assoe. 1968:77: 849.

9. LOE, H . & SILNESS, J.: Periodontal disease in preg-naucy. I. Prevalence and severity. Aeta Odontot.Scand. 1963: 21: 533-551.

10. MANDEL, I. D.: Indices for measurement of soft ac-cumulations in clinical studies of oral hygiene and pe-riodontal disease. / . Periodontal Res. 1974: 9: Suppl.14: 7-30.

11. Muin.ER, J. C : Effect on gingiva and occurrence ofpigmentation on teeth following the topical applicationof stannous fluoride or stannous chlorofluoride. / . Pe-riodontot. 1957: 28: 281-286.

12. O'CARROLL, F . M . & MuHLEK, J. C : Pigmentation as-sociated with stannous fluoride therapy in teeth cal-cified in a fluoridated area. / . Dent. Ctiitd. 1963: 30:247-253.

13. Stt..NESS, J. & LOE, H . : Periodontal disease in preg-nancy. II . Correlation between oral hygiene and perio-dontal condition. Aeta Odontol. Seand. 1964: 22: 121-135.

14. SNYDER, J. R.: Tooth enamel pigmentation tested witha new deutal index during the 1961-1963 Moorheaddeutal study. Nortti-West Dent. 1964: 43: 151-161.

15. WELLOCK, W . D . , MAn-LAND, A. & BRUDEVOLD, F . :

Caries increments, tooth discolouration, and state oforal hygiene in children given single annual applica-tions of acid phosphate-fluoride aud stannous fluoride.Areh. Orat Biot. 1965: 10: 453-460.

16. WHEELER, R . C : An attas oj tooth morphology, 3rded. W. B. Saunders, Philadelphia 1962.

Address;

Eastrnati Dentat Hospitat256 Gray's Inn RoadLondon WCIX 8LDEngland

'A