Vital and Health Statistics

,

. *  For sale by the Su In mid-December 1962 the U.S. Public Health Service completed the first cycle of the nationwide Health Examination Survey which it had begun more than 2 years earlier. The examinations, without parallel in the United States, were given by invitation to persons selected by probability sampling to represent the civilian, noninstitutional population 18-79 years of age. The conduct of examinations and selection of sample persons have been described in detail in previously published reports.lT 2 This report is based upon dental findings from the Health Examination Survey. It contains national estimates of the prevalence and distribution of selected conditions by age, sex, and race. It also includes a description of the dental examination, an account of examiner training and an assess ment of the scope and effect of nonresponse and examiner variability.

SOURCE OF DATA
The Health Examination Survey began as the result of enactment in 1956 of the NationalHealth Survey Act (P.L. 652, 84th Congress) charging the Public Health Service with securing "accurate and current statistical information on the amount, distribution, and effects of illness and disability in the United States and the services received for or because of such conditions." There are three programs within the National Center for Health Statistics that gather statistical information on health from a broad range of sources. As their names indicate, each program obtains information from essentially different sources, using and frequently developing appropriate collecting tech niques and procedures. The Health Interview Sur vey collects health data from respondents by household interview; the Health Records Survey is concerned with information obtained from the records of facilities or establishments that provide hospital, medical, nursing, personal, or residential services; and the Health Examination Survey produces data by the direct examinationof probability samples of the U.S. population.
The initial cycle of health examinations was undertaken to obtain national estimates on the prevalence of rheumatism and arthritis, diabetes, and cardiovascular disease, on certain dental conditions, and on distributions of several anthro pometric and sensory characteristics in the U.S. civilian population. A sample of 7,710 men and women was drawn from the noninstitutional adult population 18-79 years of age. Approximately 160 sample persons were examined in each of 42 areas (a total of 6,672 persons) within the continental United States. At each Survey location, standard ized examinations lasting nearly 2 hours were conducted in mobile centers by traveling staffs of physicians, dentists, nurses, and technicians.

DENTAL EXAMINATION PROCEDURE
Each dental examination was conducted in a prescribed, uniform manner (Appendix I). The examination procedure differed in some respects from the examination ordinarily given to patients seeking dental care. X-rays were not taken, teeth under inspection were not dried or isolated, oral calculus and debris were not removed prior to observation, and tooth surfaces were not probed unless overt signs of decay were observed. Moreover, to further increase agreement both within and between examiners, many questionable or borderline conditions were by design not recorded.
While these measures undoubtedly increased the comparability of the Survey findings, they also reduced the sensitivity of the examination. As a result, Survey findings in some instances are con servative in contrast to what would be obtained by clinical evaluations. While filled, nonfunctional, and, especially, decayed teeth are systematically underestimated from a clinical viewpoint, com plementary counts of normal and functional teeth are correspondingly overestimated.
Estimates of other conditions, however, such as number of edentulous persons, full dentures in use, and missing teeth, are derived from objective counts highly comparable by any examination procedure.
Most dental examinations were completed in about 10 minutes. An adjustable examining chair and standard light source were used during the examination of teeth and gums by mouth mirror and explorer.
The condition of each tooth was recorded and assessments were made of perio dontal disease and oral hygiene. The evaluation of oral hygiene was based upon amounts of debris and calculus on selected tooth surfaces, while the evaluation of periodontal disease was determined by the presence and extent of gingival inflamma tion and pocket formation.

EXAMINER VARIABILITY
The dental findings of the Health Examination Survey were obtained by only five examiners who, before joining the examining staff, were carefully trained in a standardized examination procedure (Appendix II). Training included a series of at least 140 replicate dental examinations given independently by each new examiner and by at least one of the two original examiners, A and B. A second set of replicate examinations also was completed by each examiner at or near the end of his field assignment.
Findings of original examiners served as a standard. The replicate examinations provided measures of interexaminer differences.
No attempt was made to assign sample persons randomly or equally among examiners. At most places all dental examinations were performed by a single examiner. The two original examiners, A and B, examined substantially fewer persons than did the other three; Examiner D, the dentist with the largest number, examined nearly one in three of all sample persons. Moreover, the proportion of examinations by individual dentists, even in major geographic areas, varied widely. Examiners B and E, for instance, examined relatively few persons in the West, while Exam iners A and C examined only a small number in the South (table A).
Interexaminer differences cannot be sepa rated from inherent differences existing in the various groups of examined persons. It nonethe less is true that differences in some observations largely reflect the dissimilar dental status of assigned groups. For instance, proportions of edentulous persons seen by individual exam iners-observations virtually free of examiner error-varied significantly. Examiner C found 19.1 percent without any permanent teeth but Examiner E found only 16.2 percent. Relatively little of the observed difference is due to dis similar age-sex composition since adjustedrates indicate that four examiners saw slightly more edentulous adults than expected while Examiner E saw by contrast slightly fewer than expected (table B).
Counts of decayed, missing, and filled (DMF) teeth also are relatively objective and, when determined by trained examiners, highly com parable. In the series of replicate examinations conducted on nonsample persons, counts differed within narrow ranges (Appendix II). Although determinations of decayed teeth probably show greater examiner variability than do counts of either missing or filled teeth, decayed teeth in the sample population accounted for only about 6 of each 100 DMF teeth. As a result, it seems likely that the differing DMF rates reported by the five examiners were due predominantly to real differences in the dental status of examined sample persons, not to appreciable differences between examiners.
Oral hygiene scores range from a possible ' minimum value of 0 in persons with little or no oral calculus and debris to a maximum of 6.0 in persons with copious amounts covering specified teeth. Yet, variations between examiners in the Simplified Oral Hygiene Index 3 of sample persons were comparatively slight. The highest value, 1.68, was recorded by Examiner D and the low est 1.44, by Examiner E. Moreover, differences between the reported mean value for males and for    females were remarkably consistent from one examiner to another, ranging from a low of 0.44 to a high of only 0.53 (table C).
Comparatively wide variations in the preva lence and severity of periodontal disease were obtained by the five examiners. The Periodontal Index, a method of assessing periodontal disease quantitatively, may range from a low of 0 for individuals with normal tissues supporting the teeth to a high of 8.0 for individuals with severe destructive disease involving each tooth. The highest mean Peridontal Index, 1.59, recorded by Examiner A, was slightly more than twice as great as that recorded by Examiner E. In nearly every age-sex group, index values of Examiner E were lower than the corresponding values of other examiners .
While the training of examiners, as well as the examination procedure, was designed to mini mize examiner variability, the actual extent of measurement error in the examination of sample persons cannot be determined. Comparison of the findings of individual examiners suggests that agreement was at a relatively high level. But, re gardless of magnitude, the possible effect of interexaminer differences, for example in the evalua tion of periodontal disease, varies with each characteristic. Since all examiners, for instance, examined approximately the same proportion in each age-sex group, the effect of examiner dis-

RESPONSE
Of the 7.7 10 sample adults 6,672 were actually examined, of whom 6,653 received dental exami nations. The response rate of 86.5 percent reflects unflagging efforts throughout the 2?&year-exami nation period to obtain the cooperation of all sample persons. In contrast with most other com parable surveys which, though not nationwide, involved the examination of large, broadlyrepre sentative sample populations in the United States, the response rate attained by the Health Exami nation Survey seems remarkably high.
Yet, 13.5 percent of the sample were not examined, and so large a number of nonrespond ents could introduce serious bias. In this Survey, response was lower in the older age groups than in the younger and lower among women than among men. Minor response differentials also were as sociated with race. Bias from nonresponse arises, however, only when respondents and nonrespond ents differ significantly in parameters that are to be estimated. In the instance of the Health Exami nation Survey, careful analysis of household in terviews and physician inquiries indicates that examined and nonexamined sample persons share many important health and demographic characteristics.2 Since the dental examination was but a part of a multiphase health survey, it seems unlikely that an appreciable share of nonresponse is attributable to the dental examination. However, no information was obtained about dental con ditions in nonexamined sample persons.
In each of the Survey's sample households, information was obtained by interview on both examined and nonexamined sample persons. As figure 1 shows, the proportions of examined and nonexamined persons who reported visits within specified intervals were about the same. Although relatively more nonexamined persons either failed to respond or answered "Don't know," the 96nonexamined persons in these categories represented only 1.2 percent of the identified sample.
Interval since last dental visit, an indicator of frequency of visits, undoubtedly reflects to a large degree the dental status of respondents. Data collected by the Health Interview Survey during July 1957-June 1958 show that nearly 60 percent of all edentulous persons had not seen a dentist within the past 5 years, compared with only about 8 percent of all other persons.* The response to two other questions asked by the Health Examination Survey further points out the similarity of the examined and nonexamined All races------- DonV know No answer dentist Figure  3. Percent of examined and of nonexamined adults accord ing to whether they have Q regular dentist.
permanent teeth and nearly 10 million more had lost all 16 teeth in either the upper or the lower jaw. This means that among every 100 adults an estimated 18 had no natural teeth at all, while 9 others had natural teeth in only one jaw (table 1). Only 1.3 percent of men and 1.4 percent of women in the 18-24 year age group were edentu lous. In each succeeding age group, the percent of both men and women who had lost all permanent teeth was increasingly higher. At ages 35-44, only 5.9 percent of men and 10.0 percent of women were edentulous, but by ages 75-79, the percent of edentulous persons had risen to 55.7 and 64.8, respectively. In each of the seven age groups represented in the sample, relatively more women than men had lost all of their natural teeth ( fig. 4).
Rates of edentulous persons also differed greatly by race. While 19.2 percent of the total white population had lost all their permanent teeth, only 11.3 percent of Negro adults were 60 AGE I rigwe 4. Percent of men and women who ore edentulcvs, by age. edentulous. In nearly every age group, pro portionately more white than Negro adults had lost all their natural teeth (tables 2 and 3, fig. 5).
In each of the age groups under 45 years approximately the same number of men and women had lost all their teeth in one jaw as in bbth jaws. At age 45, the percent of edentulous persons increased abruptly and continued to rise in each succeeding age group. In fact, at age 75-79 years men were more than twice as likely and women were five times as likely to have no teeth in either jaw as to have no teeth in one jaw.
The estimated overall prevalence of edentu lous persons based on the Health Examination Survey was 19.6 per 100 persons aged 25-74 years, only slightly lower than a comparable estimateof 20.5 derived from the Health Interview Survey for these age groups4 The latter report was based on an independent survey during the period July 1957-June 1958 in which interviews were con . . ducted in approximately 36,000 households and included about 115,000 persons. In all but the youngest age group, the Health Interview Survey estimates are slightly higher than those of the Health Examination Survey. The higher rates reported by the Health Interview Survey m ight have resulted from including with edentulous persons some who were actually edentulous in only one jaw.
DMF Teeth DMF teeth, first explained and used by Klein and Palmer, 5 are the sum of permanent teeth in a person's mouth that are decayed, filled, and m issing or indicated for extraction. In younger adults where relatively few teeth are lost from periodontal disease, DMF teeth are a measure of the cumulative toll of dental decay. In older adults who lose an increasing number of teeth becauseof advanced periodontal disease DMF teeth provide a convenient summary of the lifetime toll of dental disease in the permanent dentition. DMF counts in this report are based on 32 teeth, including third molars. Since unerxpted third molars were not identified they are included, along with extracted molars, in the counts of m issing teeth. This procedure results in overestimates of DMF counts as a measure of the impact of dental dis ease, especially for the younger adults, many of whom have unerupted third molars (Appendix III).
Dental examinations indicated that the 111 m illion adults in the United States had the stag gering total of 2% billion decayed, m issing, and filled teeth. O f the average 20.4 DMF teeth per person, 13.5 teeth were classified as m issing, 5.7 filled, and 1.2 decayed. Mean scores for men and women of comparable age were similar, but those for women generally were one or two DMF teeth higher. As m ight be expected, the overall DMF count, as well as the m issing component, increased in both sexes with advancing age, whereas, by contrast, the number of decayed teeth decreased. The average number of filled teeth reached a peak at ages 25-34 and decreased eo 40 50 60 70 80 AGE I Figure 6. Average DMF and components for adults, by age.
thereafter. The exclusion of edentulous persons from the counts reduces the level by about oneeighth, but leaves the relationship by age, sex, and race essentially unchanged (tables 4 and5, figs. 6 and 7). From 18-44 years of age persons of both sexes had approximately seven to eight filled teeth. Thereafter, the mean number of filled teeth decreased sharply. By age 65, the average number of filled teeth was 1.9 for menand2.6 for women, and for the oldest age group (75-79 years) the figures dropped to 0.9 for men and 1.5 for women.
The average DMF count for white adults was D AGE 0 1  I  I  I  I   I   20  30  40  50  60  70  60 AGE approximately half again as high as that for Negro adults, with an average per person of 21.2 and 14.5, respectively (tables 6 and 7, figs. 7 and 8). This higher DMF count resulted from a much larger number of fiJ.Ied teeth for white adults, 6.3, as compared with only 1.3 for Negro adults, as well as to a slight excess in the number of missing teeth, 13.9 compared with 11.3. On the other hand, Negro adults averaged somewhat higher numbers of decayed teeth than did white adults ( fig. 8). In both races DMF counts were higher for women than for men, with the dif ferences more pronounced in the Negro race.

Periodontal Disease
The prevalence and severity of periodontal disease in the U.S. adult population were measured by the Periodontal Index (PI), a method firstpro posed by Russell in 1956.6 Index classification of individuals is determined by visual appraisal of the condition of the tissues supporting each tooth in the mouth. Since by definition periodontal scores cannot be assigned to missing teeth, the lifetime experience of individuals with periodontal disease may be underestimated.
However, individuals who already have lost teeth because of periodontal disease very likely will show extensive disease involving their remaining teeth.
Zero is assigned when there is no evidence of periodontal disease. When overt signs of perio dontal disease are present, teeth are assigned scores of 1, 2, 6, or 8, corresponding to the ob served extent of gingival inflammation and de structive disease. The values of all teeth in the mouth then are averaged to obtain an individual's score or his PI. The periodontal score is not a clinical diagnosis. According to Russell, however, most persons with clinically diagnosed gingivitis have scores ranging "from 0.1 to 1.0, those with frankly established destructive disease from 1.5 to 5.0, and those with disease in terminal stages from about 4.0 to 8.0."' Periodontal scores were estimated for the approximately 90 million adults who had at least one natural tooth (table 8). The average score for all persons 18-79 years of age was 1.13. The average score increased with advancing age from a low of 0.62 and 0.48 for men and women, re spectively, in the youngest age group to a high of 2.91 for men and 2.94 for women in the oldest age group. W ithin each of the various agegroups, average scores generally were higher for men than for women.
The mean periodontal score for Negro adults was higher than that for white adults, 1.60 com pared with 1.06 ( fig. 9). In corresponding age groups of both races, the average for males was greater than that for females.
Matching the pattern of the PeriodontalIndex findings the percent of persons with destructive periodontal disease-that is, disease which had advanced to the point of pocket formation-increased rapidly with age (table 9, fig. 10). For men and women alike, the prevalence rose from approximately 1 in 10 for 18-24-year-oldpersons to more than 5 in 10 for the oldest age groups. In every age group destructive periodontal disease was more common for men than for women.
Half again as many Negro adults as white had one or more periodontal pockets, 36.0 versus 23.9 percent (table E). The percent for Negromen was 38.9 as compared with 28.9 for white men. The difference by race between women was even greater, 33.3 and 19.2,respectively.

Simplified
Oral Hygiene Index The Simplified Oral Hygiene Index is a method devised by Greene and Vermillion3 for recording degrees of debris and calculus formation in six AGE different areas of the mouth and then averaging them to provide a final whole mouth score between 0 (no debris, stain, or calculus present) and 6 (more than two-thirds of the exposed tooth sur faces of the six specified teeth covered with debris and calculus). In this Survey oral hygiene evaluations were made only if at least one of the six specified teeth was present. Accordingly, edentulous persons and persons with none of the preselected teeth were excluded from these ratings. The average Simplified Oral Hygiene Index for all adults was 1.5. The index for white persons was lower than that for Negroes-l.5 compared with 2.2. In both races the findings for men were generally slightly higher than for women, and the index increased with advancing age (  Both sexes------- and filled teeth showed a twofold increase from the youngest to the oldest age groups, rising from a low of about 14 teeth to a high of about 29. The prevalence and severity of periodontal disease in persons with natural teeth and therefore still susceptible to the disease also increased sharply with age. At ages 75-79 years, no fewer than 9 of every 10 such persons showed evidence of periodontal disease and more than half had evi dence of destructive disease. More women than men had lost all their permanent teeth. The mean number of decayed, missing, and filled teeth also was generally higher for women than for men of comparable age. Perio dontal disease, by contrast, was less severe and less prevalent in women. Within most agegroups, differences by sex were not large.
Substantial differences in dental status were found between white andNegro adults. White adults were twice as likely as Negro to have lost all their natural teeth, either in one or both jaws. The average number of decayed, missing, and filled teeth in Negroes was only about two-thirds of the number found in white persons, 14.5 and 21.2 teeth, respectively. On the other hand, destructive periodontal disease was half again as prevalent among Negro as among white adults.

REFERENCES
Percent distribution States, of adults, by status of periodontal and age: United                            DMF is the total of these three categories. Third molars are included in the count.       APPENDIX I

THE DENTAL EXAMINATION
The dental examination of the Health Examination Survey is designed to gather comparable information on the dental health status of the population. As a result, the examination procedure has been standardized so that not only the same examiner but different examiners obtain their findings on a uniform basis. The dental examination consists of determining the condition of the teeth, whether, for instance, a tooth is decayed, missing, and filled, and of assessing malocclusion, oral hygiene, and periodontal disease through the use of "indexes." The presence or absence of fluoride and nonfluoride opacities of the maxillary anterior teeth is also recorded. The dental examination is performed by the dentist member of the health survey team. A portable chair and light are used, and the mouth mirror and explorer examination of the teeth and gums usually requires about 10 minutes.
To determine the condition of individual teeth on a uniform basis and to restrict the examining dentist's judgment to as narrow a range as possible, objective criteria have been set up and are followed throughout the examination procedure. The various criteria repre sent a line drawn at a high common denominator of specific conditions, a line or denominator which, in most instances, is visible evidence of a condition which, when seen by most dentists, would bring agreement that the condition does indeed exist. A tooth, for example, is considered "nonfunctional-loss of supporting structure" when its total mobility labiolingually or buccolingually exceeds three millimeters. Similarly, when determining whether a tooth is carious, the exam iner first looks for evidence of decay-undermined enamel in pits and fissures, opacity of marginal ridges, and decalcified areas on smooth surfaces. Once observed, suspected lesions are considered carious only when a break in the enamel can be demonstrated with an explorer.
The "indexes" which are included in the examination are objective assessments of the oral hygiene status and of the severity of malocclusion and periodontal disease in individuals. The oral hygiene assessment is based upon the amount of oral debris and calculus on selected teeth; the assessments of malocclusion and periodontal disease are based, respectively, upon the number of malaligned teeth and their degree of mala lignment, and the presence and extent of gingival in flammation and pocket formation.

24
"Should see own dentist at an early date"is checked when the individual presents a condition which suggests that an examination by his own dentist is desirable in order to arrive at a clinical diagnosis of the condition and to determine whether or not treatment is needed; otherwise, "at next regular appointment" is checked. Each person examined is informed by the examining dentist that the Survey examination must not be con sidered a substitute for an examination by his own dentist.

Edentulous Arches
An edentulous arch is identified by a check in the appropriate box. The box which indicates the presence or absence of a denture for that arch is then checked. A denture is scored present only when in the examinee's mouth at the time of the examination and not defective.
An arch in which the crown of an erupting tooth can be seen or in which roots only are present is also con sidered edentulous if a full denture is being used. The presence of erupting teeth and roots under a full denture is noted in the remarks.
The box which designates a defective denture should be checked only when there is visible evidence that the denture is causing extensive destruction of theprimary stress-bearing areas of the ridge or palate. Tissue in these areas may be acutely inflamed; bone resorption may have occurred; hypertrophied tissue may be , present. The denture is also defective if it is in the possession of the examinee at the time of the exami nation but not in the mouth.

Status of Tooth Spaces
The status of each tooth space in an arch should be recorded when that arch has at least one tooth or root present and a full denture is not being used. The symbol indicating the condition of the space is written in the upper section of each respective tooth space box.
Primary teeth are numbered and scored the same as permanent teeth but with a circle around the symbol. When the succedaneous tooth is also present, the symbols for the primary tooth are placed above the upper arch, or below the lower, circled, and a line is drawn from the circle to the tooth's position in the permanent arch, The examiner should determine the condition of tooth spaces in accordance with the criteria listed below. Circumstances which in some instances may prevent a reasonable application of the criteria should be ex plained by the examiner under "Remarks." . Filled teeth which are not carious are scored similarly when the restoration is: 1. Loose 2. Temporary 3. Fractured and the base or pulpal wall of the cavity preparation exposed. (XD) Nonfunctional-carious.-When caries has pene trated the pulp chamber of a tooth, that tooth is scored (XD). Teeth are scored so when there is: 1. Visible evidence of periapical abscess or ex posure. 2. Visible evidence of extensive underminingof all enamel walls.
NOTE: All roots are scored (XD) and X is placed in the lower section of the tooth space box.
(XP) Nonfunctional-loss of supporting structure.-When the mobility of a tooth exceeds 3 mm. as measured at the incisal or occlusal third of the crown, or when the tooth is depressible in its E Jcket, the tooth is scored (XP). (X0) Nonfunctional-other .-An (X0) score is entered for all teeth with occlusal surfaces contacting the opposing alveolar ridge when the remaining teeth are in occlusion. (R) Replaced on fixed bridge or removable partial denture.-When a missing tooth is replaced on a fixed bridge or removable partial denture, the space normally occupied by the missing tooth is scored (R). (RD) Replaced defective.-Missing teeth replaced on a defective fixed bridge or a defective removable partial denture are scored (RD).
Fixed bridges are defective: 1. When one of the abutment teeth is nonfunctional due either to caries or ioss of supporting structure, or when there is visible evidence of periapical pathology. 2. When the connection of the pontic with its abut ment is broken. 3. When an abutment crown or inlay is defective due to one of the following reasons: A. Tooth structure exposed by abrasion of the crown or inlay is carious. B. A carious lesion at one of themargins of the restoration has resulted inextensiveunder mining of an enamel wall.
Removable partial dentures are defective: 1. When one of the abutment teeth is nonfunctional due either to caries or loss of supporting structure or when there is visible evidence of periapical pathology. 2. When there is visible evidence that the denture is causing extensive destruction of the stressbearing areas of the ridge or palate.

Periodontal Score6
A periodontal score isrecordedinthelower section of each tooth space box for every tooth, other than roots, that is present. Criteria for the periodontal score: Score 0 -Negative.-There is neither overt in flammation in the investing tissues nor loss of function due to destruction of supporting tissues. 1 -Mild gingivitis.-There is an overt area of inflammation in the free gingivae, but this area does not circumscribe the tooth. 2 -Gingivitis.-Inflammation completely circumscribes the tooth, but there is no apparent break in the epithelial attachment. 6 -Gingivitis with pocket formation.-The epithelial attachment has been broken and there is a pocket (not merely a deepened gingival crevice due to swell ing in the free gingivae). There is no interference with normal masticatory function; the tooth is firm inits socket and has not drifted. 8 -Advanced destruction with loss of masti catory function.-The tooth may be loose; may have drifted; may sound dull on percussion with a metallic in strument; may be depressible in its socket.

Simplified
Oral Hygiene Index3 Selected surfaces of six teeth are used in making this estimation of oral hygiene status. For the purposes of this examination, each surface that is used, buccal or lingual, is considered to encompass half of thecircum ference of the tooth. The buccal surface ofa molar, for example, is considered to include half of the mesial surface and half of the distal.
The posterior teeth used for the assessment are the first fully erupted teeth distal to the bicuspid area on each side of each arch. Inmost cases this will be a first molar, but in others it may be a second or third molar. The buccal surfaces of upper molars and the lingual of lowers are examined. In the anterior portion of the mouth, the labial surfaces of the upper right central in cisor and the lower left central incisor are examined. When one or both of these teeth is missing, the adjacent central incisor is substituted.

A. Examining for Oral Debris
The surface area covered by debris is estimated by running a number five explorer along the surface being examined and noting the occlusal or incisal extent of the surface being examined and noting the occlusal or incisal extent of the debris as it is removed from the tooth surface and adheres to the explorer.
Scoring: 0 -NO debris or stain present. 1 -(a) Soft debris covering not more than the gingival third of the tooth surface, or (b) the presence of the extrinsic stains without debris re gardless of surface area covered. 2 -Soft debris covering more than one-third but not more than twothirds of the exposed tooth sur face. 3 -Soft debris covering more than two-thirds of the exposed tooth surface.

B. Examining for Oral Calculus
A number five explorer is also used to estimate surface area covered by supragingival calculus and to probe for subgingival calculus.
1 -Supragingival calculus cov ering not more thanone-third of the exposed tooth surface. 2 -Supragingival calculus COV ering more than one-third but not more than two-thirds of the exposed tooth surface, and/or the presence of indi vidual flecks of subgingival calculus around the cervical portion of the tooth. 3 -Supragingival calculus cov ering more than two-thirds of exposed tooth surface and/ or a continuous heavy bandof subgingival calculus around the cervical portion of the tooth.

C. Calculating the Index
The debris scores are totaled and divided by the number of surfaces scored. The calculus score is de termined similarly.
The debris and calculus scores are then added to give the examinee's Simplified Oral Hygiene Index.

Nonfluoride Opacities
These lesions are often round or oval. They are clearly differentiated from adjacent normal enamel. They are usually pigmented at the timeof eruption, often creamy-yellow to dark reddish-orange. Any tooth may be affected. They do not follow a standard pattern of distribution.

Score :
None -Less than two of the eight upper anterior teeth are affected. Mild -Two or more of the eight upper anterior teeth are affected but the areas cover less than half of the labial surface. Objectionable -At least half of the labial surfaces of two or more of the eight upper anterior teeth are affected.

Fluorosis
In the space provided, one of the following three classifications is checked. None -The enamel presents the usual translucent andemivitrioform type of structure. The sur face is smooth, glossy, and usually of a pale creamy white color.
M.-At least two of the eight upper anterior teeth have small, opaque, paper-white areas scattered irregularly over them but the areas do not involve as much as approximately 50 percent of the labial surface.
Objectionable.-Half or more of the labial surfaces of at least two of the eight upper anterior teeth are affected and surfaces subject to attrition show marked wear. Brown stain is frequently a dis figuring feature. There may be discrete or confluent pitting. NOTE: When less than two of the eight upper anterior teeth are present, the "not applicable" box is checked.
Malocclusion Score Each tooth is scored individually but subtotals are recorded for six segments of the mouth: 1.

2.
A tooth may be in more than oneof the above posi tions, i.e., both rotation and displacement; if so, it is assessed according to the position of highest score value.
-----------  II  I2  I3  IY  I5  I6   The first of a series of examinations to standardize and V). In all tables, the same examiner is given the and calibrate dental examiners for the Health Examina-same letter. For example, Examiner A in Table I is tion Survey was completed in November 1958. Each of also Examiner A in Table IV. approximately 200 individuals, most of whom were from The findings in table VI are derived from the ex-14 to 17 years of age, was seenby two examiners using amination of the same 56 persons by each examiner mouth mirrors and explorers and a standard dental including a second series by Examiner A. Examiner C light. Unlike the ones to follow, however, these initial completed his examinations as a training exercise prior examinations were not only to train examiners in a to returning to the field following a lo-month clinical standard procedure but, more important, to determine assignment. The examinations by Examiners D and E whether a proposed examination procedure was suf-were performed at or near the close of their field ficiently objective to allow different examiners to ob-assignments. tain comparable findings.
Before the start of the examinations, a small number of persons were seen by both examiners. Any differences in findings recorded by examiners for individuals were Table I. Mean dental findings for 207 persons discussed in an effort to attain as uniform an application examined in November 1958, Health Examination of the examination criteria as was possible. No com-Survey parisons of oral hygiene were madeon replicate exami nations since one dentist's assessment prohibits avalid Examiner appraisal of the same mouth by a second dentist. Exami-Findings nation of the first of 207 persons was then begun with A B the examiners working independently. The results of the two examiners are shown in table I.
To compare the findings resulting from the exami-Decayed, missing, and filled----11.6 11.6 nation procedure with the findings of a trained examiner Total teeth present-------26.9 26.9 experienced in DMF surveys, a dentist from the Epi demiology and Biometry           The Health Examination Survey is designed as a highly stratified multistage sampling of the civilian, noninstitutional population, aged 18-79 years, of the conterminous United States. The first stage of the plan is a sample of the 42 primary sampling units (PSU's) from 1,900 geographic units into which the United States has been divided. A PSU is a county, two or three contiguous counties, or a standard metropolitan statistical area. Later stages result in the random selection of clusters of about four persons from a small neighborhood within the PSU. The total sample Included 7,710 persons in the 42 PSU!s in 29 different States. The detailed structure of the design and the conduct of the Survey have been described in previous reports.'> *

Reliability in Probability Surveys
The methodological strength of the Survey derives especially from its use of scientific probability sampling techniques and of highly standardized and closely controlled measurement processes. This does not imply that statistics from the Survey are exact or without error.
Data presented are imperfect for three im portant reasons: (1) results are subject to sampling error, (2) the actual conduct of a survey never agrees perfectly with the design, and (3) the measurement process itself is inexact, even when standardized and controlled. The faithfulness with which the study design was carried out has been analyzed in a previous report. * Of the total of 7,710 sample persons, 86 percent or 6,672 were examined. Analysis indicates that the ex amined persons are a highly representative sample of the adult civilian, noninstitutional population of the United States. Imputation for the nonrespondents was accomplished by attributing to nonexamined persons the characteristics of comparable examined persons. The specific procedure used * consisted of inflating the sampling weight for each examined person to com pensate for nonexamined sample persons at the same stand and of the same age-sex group.
Of the examined adults there were 19 who did not have the dental examination. These persons are included in the "Total" and "With neither arch eden tulous" columns of detailed tables 1, 2, and 3, and are excluded from alI other tables.

Sampling and Measurement Error
In this report and its appendices, several refer ences have been made to efforts to evaluate both bias and variability of the measurement techniques. The probability design of the 'Survey makes possible the calculation of sampling errors. Traditionally the role of the sampling error has been the determination of how imprecise the survey results may be because they come from a sample rather than from measurement of all elements in the universe.
The task of presenting sampling errors for a study of the type of the Health Examination Surveyis compli cated by at least three factors. (1) Measurement error and "pure" sampling error are confounded in the data; it is not easy to find a procedure which will either completely include both or treat one or the other sepa rately.
(2) The survey design and estimation procedure are complex and accordingly require computationally involved techniques for calculation of variances. (3) Thousands of statistics come from the Survey, many for subclasses of the population for which there are small numbers of sample cases. Estimates of sampling error are obtained from the sample data and are themselves subject to sampling error, which may be large when the number of cases in a cell is small, or even occasionally when the number of cases is substantial.
In the present report, estimates of approximate sampling variability for selected statistics are pre sented in table VII. These estimates have been prepared by a replication technique which yields overall variability through observation of variability among random subsamples of the total sample. The method reflects both "pure" sampling variance and a part of measurement variance.
In accordance with usual practice, a 68 percent con fidence interval may be considered the range within one  These data indicate cent confidence interval the range within two standard that for ages 18-24 the number of missing third errors. molars cannot be more than 2.1 and is almost surely An overestimate of the standard error of a dif-less than 1.7. Thus, it is probable that the missing ference d= X-Y of two statistics x and y is given component of the

18-24 years---------
In some tables magnitudes are shown for cells for which sample size is so small that the sampling error may be several times as great as the statistic itself. Obviously in such instances the statistic has no meaning in itself except to indicate that the true quantity is small. Such numbers, if shown, have been included to  I  I  I  I  I  I  I  I  I  I  I  I  I   I  I  16   19 20 21