The Periodontal Disease Index (PDI)

The Periodontal Disease Index (PDI)

BY SIGURD P. RAMFJORD*

THE PERIODONTAL Disease Index is aclinician's modification of Russell's PI in-dex for epidemiological surveys of perio-dontal disease. The PDI index is primarilyconcerned with an accurate assessment ofthe periodontal status of the individual per-son. Emphasis is placed on recording ofthe attachment level of the periodontal tis-sues relative to the C-E junction. Such ac-

curate measurable assessments are essentialfor longitudinal studies of periodontal dis-ease and as a scientific basis for clinicaltrials in Periodontology.Objectives of the PDI IndexThe following objectives were incorpo-

rated into the design of the index:1. To assess prevalence and severity of

gingivitis and Periodontitis within the indi-vidual dentitions and in population groups.2. To provide an accurate basis for inci-

dence and longitudinal studies of perio-dontal disease.

3. To provide a meaningful basis for es-timate of need for periodontal therapy inselected population groups.4. To provide accurate recordings for

clinical trials of preventive and therapeuticprocedures in periodontics.5. To provide measurable reference data

for assessment of correlations with factorsof potential significance in the etiology ofperiodontal disease.

Scoring MethodsIn the 20 minutes allowed me for this

presentation I will mainly confine my re-marks to clarification of the scoring meth-

*The University of Michigan School of Dentistry,Ann Arbor, Michigan.

S. Ramfjord presenting.

ods. Accuracy and reproducibility of scor-ing becomes increasingly more dependenton the training of the person who is goingto do the scoring as demands for accuracyare increased, and the merits of this indexshould not be evaluated until a person hashad considerable training and experiencein the use of the system. Assessment of de-gree of periodontal disease includes a sub-jective assessment of color, form, density,and bleeding tendency of the gingival tis-sues. But by far the most important featureof the PDI index is measurement of thelevel of the periodontal attachment relatedto the C-E junction of the teeth. The loca-tion of the reference marks needed for themeasurements however is based on touch

Page 30/602

Periodontal Disease Index Page 31/603

rather than on vision, which means that theinvestigator has to be trained to locate theC-E junction and the bottom of the epi-thelial attachment accurately by touch. Ifthese reference marks are covered over bycalculus they have to be exposed. In a fewinstances the original cementum-enameljunction may have been lost due to exces-

sive abrasion or dental restorative proce-dures. However, in the great majority ofindividuals the cementum-enamel junctioncan be located by a well trained investi-gator. It should be emphasized that thismethod is entirely unreliable in the handsof an untrained investigator.

Maximal accuracy also depends on stand-ardized optimal lighting and standardizedthickness of the measuring probes. Only 6selected teeth are scored for assessment ofthe periodontal status of the mouth; how-ever, for short term clinical trials andwhere a limited number of patients are

available, one may concern all of the teethin the mouth. The 6 selected teeth are:

tooth #3 (maxillary right first molar),tooth #9 (maxillary left central incisor),tooth #12 (maxillary left first bicuspid),tooth #19 (mandibular left first molar),tooth #25 (mandibular right central in-cisor) and tooth #28 (mandibular rightfirst bicuspid). It has been shown byJamison and in a number of published andunpublished studies from our own institu-tion that these 6 teeth provide basis for asurprisingly accurate assessment of the to-tal periodontal status of the individual asexpressed in scoring of all of the teeth.

The gingival status is scored first. Themethod and the assigned value representsessentially a combination of the PMA andthe PI index, with the following definitionsof criteria:

0 = absence of signs of inflammation1—

mild to moderate inflammatory gingivalchanges, not extending around the tooth

2 = mild to moderately severe gingivitis ex-tending all around the tooth

3—

severe gingivitis characterized bymarked redness, swelling, tendency tobleed and ulceration.

Since all these criteria are based on sub-jective values and examiners judgment Iwould like to describe in detail the basisfor such judgment. The gingivae aroundthe teeth to be scored are first dried super-ficially by gently touching with absorbingcotton. Changes in color are evaluated byobserving the color of the gingiva aroundthe tooth to be scored and comparing thecolor corresponding to the buccal, lingualand interproximal surfaces with each other,as well as comparing it with the color ofthe gingiva around adjacent teeth. We paymore attention to the uniformity or lack ofuniformity of color than to color shades orhues, since we assume that the physiologicfactors which determine different colorshades of gingiva are constant around thetooth and for the adjacent teeth. Colorchanges are usually towards redness butthere may also be changes towards a bluishor purplish hue.

Change in form is initially a blunting orrounding of the margin of the gingiva andthickening of papilla; however, gingivitis isnever scored on the basis of a slight con-tour change alone since this may not neces-sarily indicate the presence of disease. Verylittle significance is given to the presenceor absence of stippling since this does notnecessarily relate to presence or absence ofgingival inflammation.

Change of consistency or density is de-tected by applying gentle pressure with theside of the periodontal probe against thegingiva to determine if there is a soft orspongy consistency. If there is a clearly de-tectable color change indicating gingivitis,the consistency is not tested. Any minorchange either in contour, stippling or con-sistency alone is not considered to be a defi-nite manifestation of gingivitis.The score of 3 is based on evidence of

ulceration of the gingiva with bleeding, ifthe gingiva is touched gently with the sideof a periodontal probe, or if there is severeredness and marked change in contour. Thescore of 3 is given even if these changes donot extend all around the tooth.

Page 32/604 Ramfjord

The next step in the scoring procedureis recording of crevice depth related to theC-E junction. For this purpose we use a

University of Michigan # 0 probe, made bythe Premier Mfg. Co. in Philadelphia. Wehave encountered great problems in stand-ardizing the manufacturing of this probeand it still is not the instrument that wewould like to have. We have had problemswith variation in thickness, with inaccurateplacement of the reference marks, varia-tions in clarity of definition of the referencemarks, and variations in angulations. Be-sides the obvious errors that can resultfrom variation in placement of the refer-ence marks, a slight variation in thicknessmay influence the results when populationswith fairly normal and dense gingival tis-sues are scored. A thin probe under thesecircumstances will penetrate deeper than a

thicker probe and consequently give higherscores.

The probe should be held with a lightgrasp similar to the manner of holding a

pencil and balanced well in the hand so itcan be moved and directed by very smallforces. The end of the probe should beplaced against the enamel surface coronallyto the margin of the gingiva so that theangle formed by the working end of theprobe and the long axis of the crown of thetooth is approximately 45°. A minimal forceshould be used to pass the probe in apicaldirection maintaining contact with thetooth. The angle between the probe andthe tooth may have to be decreased slightlywhen the probe touches the gingiva toavoid pressure on the gingiva when theprobe is inserted in the gingival crevice.

Since the surfaces of the enamel andthe cementum have different inclines, thechange in direction of the movement of thepoint of the probe is detectable when itmoves from the enamel to the cementum.The texture or surface characteristics arealso different since the cementum is dis-tinctly rougher than enamel. A very im-portant point is to use a light grip on theprobe otherwise the keen sense of touch in

the fingertips becomes impaired. The probeshould always be pointed towards the apexof the tooth or the central axis of multi-rooted teeth. After the distance from thefree gingival margin to the cementumenamel junction has been measured, an at-tempt should be made to move the probealong the cemental surface. This of coursecan be achieved only if there has been lossof periodontal attachment. If calculus cov-ers the cementum enamel junction it has tobe removed before the C-E junction can belocalized. Occasionally it is also necessaryto remove heavy deposits of supragingivalcalculus to gain access to the gingival crev-ice. The University of Michigan #0 probeis graduated at 3, 6, and 8 mm. from theend, making it necessary to estimate inter-vening measurements. In our experience,reproducibility is better and eye strain lessfollowing proper training with this probethan with probes that have marks for everymm.

All measurements are rounded to thenearest mm.; except that anything close toVi a mm. is always rounded to the lowerwhole number. It has been found in ourcombined histometric and clinical studiesthat there is a tendency with a thin probeto record a slightly greater depth than tothe coronal level of the connective tissueattachment to teeth. Thus we do not re-cord for instance 1 mm. of pocket depthbelow the C-E junction and indicate loss ofattachment unless we are sure that theprobing extends definitely more than Vimm. from the C-E junction. By assigningall the doubtful measurements to the lowerscore the reproducibility is much greaterthan if a more accurate determination ofVx mm. were attempted.

We have during the last two years modi-fied the PDI index so we are now scoringonly buccal and mesial measurements. Ithas been found both by Jamison and Ashin separate extensive analysis, althoughyet unpublished, that there is no significantloss of accuracy in the PDI index fromomitting the lingual and distal scores. Omit-


S. Ramfjord examining.

ting the lingual and distal scores has madeit much easier to achieve reproducibility inbuccal and mesial scoring than when we in-cluded distal and lingual scoring. However,in some clinical trials where there are a

limited number of cases available and theinvestigators are well trained, we may stilluse all 4 measurements for each tooth.

The crevicular measurements are re-

corded in the following manner: The dis-tance from the free gingival margin to thecementum enamel junction and the dis-tance from the free gingival margin to thebottom of the gingival crevice or pocket ismeasured for the buccal and mesial aspectof each tooth examined. The buccal meas-urements should be made at the middle ofthe buccal surfaces. The mesial measuringshould be made at the buccal aspect of theinterproximal contact area with the probetouching both teeth if there is a neighbortooth present and the probe pointing in the

direction of the long axis of the tooth to bescored. Follow this routine for recording:

A. If the gingival margin is on enamel,measure from gum margin to cementum-enamel junction and record the measure-ment. If the epithelial attachment is on thecrown and the cementum enamel junctioncannot be felt by the probe, record thedepth of the gingival crevice on the crown.Then record the distance from the gingivalmargin to the bottom of the pocket ifthe probe can be moved apically to thecementum-enamel junction without resist-ance or pain. The distance from the ce-mentum-enamel junction to the bottom ofthe pocket can then be found by subtract-ing the first from the second measurement.

B. If the gingival margin is on cemen-

tum, record the distance from the cemen-

tum-enamel junction to the gingival marginas a minus value. Then record the distancefrom the cementum-enamel junction to thebottom of the gingival crevice as a plusvalue. Both loss of attachment and actualcrevice depth can easily be assessed fromthese scores.

Scoring of Calculus

Calculus is scored on the basis of thefollowing criteria:

0 = absence of calculus1 = supragingival calculus extending only

slightly below the free gingival margin(not more than 1 mm.)

2 = moderate amount of supra and subgin-gival calculus or subgingival calculusalone

3 = an abundance of supra and subgingivalcalculus.

Subgingival calculus apparently is muchmore important in the pathogenesis of Peri-odontitis than supragingival calculus andconsequently it has been given a higherscore than supragingival calculus. Subgin-gival calculus is located with a #17 probeif there is uncertainty when probing withthe University of Michigan # 0 probe.


Scoring of PlaqueScoring of plaque is done after staining

with Bismarck brown solution. Bismarckbrown solution is placed in a dappen dish,and two Richmond cotton pellets #0 are

placed in the dish until they appear com-pletely saturated with the solution. Thepatient is asked to swallow and one satu-rated pellet is removed with a cotton plierand touched gently onto the lingual andbuccal surfaces of the mandibular teeth.The second saturated pellet is touched ontothe palatal and buccal surfaces of the max-illary teeth. The occlusal surfaces are alsorubbed with the pellet so the disclosingsolution flows over all the surfaces of theteeth. The patient is then instructed to spitin the cuspidor and rinse his mouth thor-oughly twice. The scoring is now doneaccording to the following criteria:0 = no plaque present1 = plaque present on some but not on all

interproximal buccal and linqual sur-faces of the tooth

2 = plaque present on all interproximal,buccal and lingual surfaces, but cover-ing less than one half of these surfaces

3 = plaque extending over all interproximal,buccal and lingual surfaces, and cover-ing more than one half of these surfaces.

Only full erupted teeth should be scoredand missing teeth should not be substituted.

The scoring of the calculus and plaquedoes not constitute a part of the PDI index,but since an extremely high correlation hasbeen established between periodontal dis-ease and the presence of plaque and calcu-lus we usually include plaque and calculusscore in total assessment of periodontalstatus.

Computation of Indices fromRecorded DataThe recorded data can be utilized as a

basis for individual indices for gingivitis,for total periodontal disease (PDI), forcalculus and for plaque or for calculus andplaque combination. The gingivitis index isderived at by adding the scores for gingivi-tis for all of the examined teeth and then

divide the sum by the number of teeth ex-amined. The PDI index for the patient isthe average of the PDI index for the ex-amined teeth, which is based on the follow-ing tabulation: If the gingival crevice innone of the measured areas extended api-cally to the cementum-enamel junction therecorded score for gingivitis is the PDIscore for that tooth. If the gingival crevicein any of the two measured areas extendedapically to the cementum-enamel junctionbut not more than 3 mm. (including 3 mm.in any area) the tooth is assigned a PDIscore of 4. The score for gingivitis then isdisregarded in the PDI score for that tooth.If the gingival crevice in any of the tworecorded areas of the tooth extends apicallyto from 3 to 6 mm. (including 6 mm.) inrelation to the cementum-enamel junctionthe tooth is assigned the PDI score of 5(again the gingivitis score is disregarded).Whenever the gingival crevice extends morethan 6 mm. apically to the cementum-enamel junction in any of the measuredareas of the tooth, the score of 6 is as-

signed as the PDI score for that tooth(again disregarding the gingivitis score).

The total PDI score for the patient isthen computed by addition of the individ-ual scores of the teeth and the sum isdivided by the number of scored teeth.Whether or not the periodontal supporthas been lost because of Periodontitis or

atrophy is not considered in the PDI index.

If some of the teeth which were sched-uled to be examined were missing or un-eruptsd so they could not be examined,then individual scores for each of the ex-amined teeth should be added and dividedby the number of teeth examined andthere should be no substitution for absentteeth.

Indices for calculus and plaque are tabu-lated in the same manner as the PDI index.The scores for calculus for each individualtooth examined are added and the sumdivided by the number of teeth examinedto yield the person's index for calculus.


The scores for plaque for each tooth ex-

amined also are added and divided by thenumber of examined teeth, thus the indexof plaque hygiene is tabulated.

The mean of the calculus and the plaqueindex will give an expression for the com-bined potential irritation from these twomain etiological factors related to lack oforal hygiene or an "irritation index" as in-dicated by O'Leary.Examiner Scoring Deviationand Calibration

The observers ability to assess repeat-edly in a reproducible way the scoring cri-teria in actual clinical situations have to bedetermined by errors tests prior to the study.The examiner should also be retested dur-ing a study that extends over any prolongedperiod of time. Control of accurate adher-ence to the scoring criteria and reproduci-bility is especially important for clinicaltrials and longitudinal studies.

One way of assessing scoring deviationis as follows: After the investigator haslearned to use the index to the point thatscoring becomes automatic and he can calloff the figures without any hesitation, thefollowing scoring test can be applied. Select7 adult patients at random and score eachpatient 5 times (not consecutively) for gin-givitis, crevice depth, calculus and plaque.The first and second, second and third,third and fourth, fourth and fifth, and firstand fifth scores for the individual teeth ofeach patient can be compared for gingivitis,periodontal crevice depth, calculus andplaque. No change is recorded as 0, in-creases as +1 or +2 or decreases as —1 or—2 in individual scores for the examinedteeth based on scoring units. In one recentsuch test administered to a dental hygienistdoing a toothbrush study her average devi-ation for plaque score was ±0.062; ±0.018for calculus; ±0.033 for gingivitis; ±0.000for periodontal scores and for crevice depth±0.004. A greater number of such testshave been directed by Dr. Ash and in gen-eral the accuracy figures are very high for

crevice depth determinations, slightly lowerfor calculus and plaque and lowest for gin-givitis. However, the reproducibility figuresfor gingivitis still are fairly good as can beseen from the data I just quoted to you.

Another method is also used to deter-mine variability in scoring for a single testcalibration and for tests before, during, andafter a study has been completed. Usuallysome change will take place, the amountwill depend upon initial training. Initialtraining should include a wide range ofsubjects with a sufficient range in value ofscores to prevent presentation of a numberof difficult scoring problems.

In a test scoring situation using 7 sub-jects and 5 observations the usual proce-dure is to determine the pooled estimate ofvariance of the measurement errors (Table1). The variances from two tests, one atthe beginning and another at the end ofthe study, can be tested on the basis of anF ratio. It is obvious that a knowledgeof variation and scoring error should beknown throughout a study if comparisonsbetween studies or more than one observeris to be used.

Besides the individual examiners error

one of course has to consider calibrationfor interexaminer errors. It is not withinthe scope of this paper to indicate howsuch tests should be organized; however, ithas to be emphasized that calibration testsare of essential importance in all forms ofclinical studies.

Recommended Use of the PDI Indexand Associated Data

Periodontal indices are sometimes clas-sified as irreversible and reversible indices.An irreversible index assesses the damagecaused by a disease rather than the diseaseitself. A classical example of such an indexis the DMF index for dental caries. ThePDI index has often been referred to as an

irreversible index since it measures the de-struction caused by periodontal disease and


Patient

Table 1Test

Observation2 3 4 /v 2* Sx2 df

404123216

7101311939

1120352517317

1.20.01.20.80.81.2ck86.0

Difference betweensuccessive steps

16

2S(x-x)2df

= .2143= .463

6£2.8

it has been assumed that the destructioncaused by periodontal disease was irre-versible.

A reversible index indicates the possi-bility for the score to return to zero, as aresult of cure of the disease that was in-dexed. In other words a reversible indexassesses an active disease which may becompletely cured while irreversible indicesassess permanent damage caused by a dis-ease. It has been implied that for clinicaltrials of treatment for periodontal diseaseone would need reversible indices.

This is an entirely meaningless charac-terization of indices because if the indexreports the status of health or disease ac-

curately, reversals or lack of reversals willappear when the data are compared re-gardless whether the tag of reversible orirreversible has been placed on the index.

It is however, extremely important toknow how the data obtained from the ex-amination, scoring and indexing can bestbe utilized and applied to problem solvingin clinical trials. The first objective of thePDI index was to find a numerical expres-sion of the status of periodontal health ordisease in a given individual. Then this

information could be used for assessmentof the periodontal status of populationgroups following proper sampling. Lookedupon from this standpoint the PDI index isessentially a clinician's modification of theRussell PI index and for this purpose itbehaves essentially the same way as the PIindex. The similarity in behavior of thesetwo indices has been documented by com-parison of data from Greene's and mystudy in India and from Jamison's work.

Being a clinician rather than an epi-demiologist I mainly like to discuss withyou the clinical significance of the PDIindexing. It serves to map the clinical be-havior of periodontal disease in any givenindividual as well as in population groups.Thus it may be used as a guide in assessingneed for treatment and for evaluation ofresults of treatment. More specifically itcan be stated that data accumulated asso-ciated with PDI indexing are useful for thefollowing purposes:

1. It records gingivitis for the selectedteeth and provides an average value forgingivitis in the entire mouth. It is gener-ally accepted that simple gingivitis is areversible disease. Cure of gingivitis or fluc-tuation in severity of gingivitis both for

Periodontal Disease Index Page 37/609individual teeth and for the entire mouthcan be assessed from the gingivitis scores.Other indices may of course provide suchdata equally well or better.

2. It provides the data needed for assess-ment of prevalence of both gingivitis andPeriodontitis separately, besides measuringthe severity of these conditions. It may beof great clinical significance to know atwhat age Periodontitis (periodontal de-struction) starts.

3. It provides meaningful data for as-sessment of the total need of periodontaltreatment for individuals and for popu-lations.

4. It establishes an accurate record ofthe level of periodontal support for theselected teeth at the time of the indexing.It was found in a recent study that thispart of the index is not necessarily as irre-versible as it has been claimed since perio-dontal support may be regained as a resultof treatment. The PDI index or the re-corded data from such indexing can pin-point such gain or loss of support.

If the recorded data are going to be usedfor clinical trials, separate analysis shouldbe done for the impact of the clinical trialupon gingivitis, and for the impact of thesame clinical trial upon crevice depth andlevel of periodontal attachment around theteeth. As an example it can be mentionedthat data from a recent clinical trial involv-ing periodontal therapy showed some gainof periodontal support in the interproximalareas following curettage, while a surgicallowering of buccal and lingual attachmentwas followed by gradual return to the pre-surgery level.

5. Clinical research concerning patho-genesis of periodontal disease with experi-mentally induced irritation requires thegreatest possible accuracy in accumulationof measurable data. Information from thePDI indexing provides measurable data re-

garding pocket formation related to loss ofattachment.

6. Testing of methods and devices fororal hygiene and preventive periodonticsshould include accurate information bothregarding the impact on gingivitis and ef-fectiveness in prevention of periodontaldestruction. Data from the PDI indexingwill provide basis for separate analysis ofthese two factors.

I want to emphasize that the data ac-cumulated during the PDI indexing lendthemselves very well for separate analysisof gingivitis and level of periodontal at-tachment; thus provides a great amount ofversatility beyond the information that canbe obtained from a simple PDI indexing.The data obtained in millimeters are quan-titative and suitable for the types of mathe-matical manipulations necessary for manyforms of statistical analysis. Analysis ofthe data should be directed towards thetype of information that is desired.

Final Remarks

By now I am sure that you all have con-cluded that as usual I have gone way over-board and in my enthusiasm have tried tosell you a "perfect tool" for clinical studies.You know of course that there is no suchthing as a perfect tool in this field wheresubjective judgments play havoc with math-ematical values. Furthermore, we are fish-ing with a large masked net which does notallow us to catch small fishes. There is anobvious need for refinements in devices andtechniques. We have for example refinedour plaque scores for toothbrush studies inpopulations with relatively good oral hy-giene to get a more accurate assessment ofsmall variations. We are modifying ourdata analysis to suit the various problemsthat we are investigating. We have not beenable to come up with the type of perio-dontal probe that we would like to have,and we are steadily experimenting withmethods for more expedient training of ex-aminers. Beyond all we are looking forbetter ways to catch more fish!!

REFERENCES

1. Ackerman, M. Anne: A clinical study of the Do-minion electric toothbrush. Typed thesis. Univ. of Mich.School of Dentistry, Ann Arbor, 1965. p. 86.2. Greene, J. C: Periodontal disease in India: Re-

port of an epidemiological study. J. dent. Res., 39:302-312, 1960.3. Jamison, Homer: Prevalence and severity of per-

iodontal disease in a sample of a population. Typedthesis. Univ. of Mich. School of Public Health, AnnArbor, 1960. 153 p.4. Ramfjord, S. P.: Indices for prevalence and inci-

dence of periodontal disease. J. Periodont., 30:51-59,1959.5. Ramfjord, S. P., Nissle, R. R., Shick, R. A. and

Cooper, H.: Subgingival curettage versus surgical elimi-nation of periodontal pockets. J. Periodont. (In press).6. Russell, A. L.: A system of classification and

scoring for prevalence surveys of periodontal disease.J. dent. Res., 35:350-359, 1956.7. Smith, W. A. and Ash, M. M., Jr.: A clinical

evaluation of an electric toothbrush. J. Periodont., 35:127-136, 1964.

The Gingival Index, the Plaque Index andthe Retention Index Systems

BY HARALD LÖE

THE GINGIVAL INDEX (Gl)The main purpose of creating the Gin-

gival Index system was to introduce a sys-tem for the assessment of the gingival con-dition which clearly distinguished betweenthe quality of the gingiva (the severity ofthe lesion) and the location (quantity) asrelated to the four (buccal, mesial, distal,lingual) areas which make up the total cir-cumference of the marginal gingiva (Löeand Silness, 1963). At the time the GI wastaken into use the existing index systems,the PMA index (Massler and Schour, 1949)with later modifications, the Periodontal In-dex (Russell, 1957) and the PeriodontalDisease Index (Ramfjord, 1959), did notfulfill this requirement.

The Gingival Index does not considerperiodontal pocket depth, degrees of boneloss or any other quantitative change ofthe periodontium. The criteria are entirelyconfined to qualitative changes in the gin-gival soft tissue.

Department of Periodontology, The Royal DentalCollege, Aarhus, Denmark.

CRITERIA FOR THE GINGIVAL INDEX SYSTEM

0 = Normal gingiva1 = Mild inflammation

—

slight change incolor, slight oedema. No bleeding on

probing2 = Moderate inflammation—redness, oede-

ma and glazing. Bleeding on probing3 = Severe inflammation

—

marked rednessand oedema. Ulceration. Tendency tospontaneous bleeding.

Each of the four gingival areas of thetooth is given a score from 0 to 3; this isthe Gl for the area. The scores from thefour areas of the tooth may be added anddivided by four to give the GI for thetooth. The scores for individual teeth (in-cisors, premolars and molars) may begrouped to designate the GI for the groupof teeth. Finally, by adding the indices forthe teeth and dividing by the total numberof teeth examined, the Gl for the individ-ual is obtained. The index for the subjectis thus an average score for the areas ex-amined.

GI = 0 is given to the gingiva the color ofwhich is pale pink to pink. The

Page 38/610

Documents

The Periodontal Disease Index (PDI)