Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
tjaoftai JcmmmlcfOnJudaula 11 (1989) 309-320 C 19*9 European Orthodontic Society The development of an index of orthodontic treatment priority * University College Hospital, London ** University Dental Hospital of Manchester, England SUMMARY The aim of this study was to develop a valid and reproducible index of orthodontic treatment priority. After reviewing the available literature, it was felt that this could be best achieved by using two separate components to record firstly the dental health and functional indications for treatment, and secondly the aesthetic impairment caused by the malocclusion. A modification of the index used by the Swedish Dental Health Board was used to record the need for orthodontic treatment on dental health and functional grounds. This index was modified by defining five grades, with precise dividing lines between each grade. An illustrated 10-point scale was used to assess independently the aesthetic treatment need of the patients. This scale was constructed using dental photographs of 12-year-olds collected during a large multi-disciplinary survey. Six non-dental judges rated these photographs on a visual analogue scale, and at equal intervals along the judged range, representative photographs were chosen. To test the index in use, two sample populations were defined; a group of patients referred for treatment, and a random sample of 11 -12-year-old schoolchildren. Both samples were examined using the index and satisfactory levels of intra- and inter-examiner agreement were obtained. Introduction Whilst many indices exist to record malocclusion, it is important to distinguish those that classify malocclusions into types (Angle, 1899) and those that record prevalence in epidemiological studies (Bjork et al., 1964), from those indices that attempt to record treatment need or priority. Furthermore, indices used to record treatment success and treatment difficulty will have differing requirements. Many indices have been developed with the intention of categorizing malocclusions into various groups, according to the urgency and need for treatment (Summers, 1971; Salzmann, 1968; Linder-Aronson, 1974; Lundstrdm, 1977; Grainger, 1967; Draker, 1960). Individuals with greatest treatment need can then be assigned priority when orthodontic resources are limited, and when the availability of treatment is unevenly spread. Similarly, individuals with little need for treatment can be safeguarded from the potential risks of treatment (Shaw, 1988). Recent research has called into question many of the previously held views on the benefits of orthodontic treatment (Shaw et al., 1980). There may be small effects on the susceptibility to temporomandibular dysfunction (Roth, 1973; Mohlin and Thilander, 1984) and periodontal disease (Horupe/ al., 1987;Sandali, 1973;Davies etal, 1988; Addy etal, 1988). However, so many studies have been undertaken on these subjects, with differing conclusions, and often only weak statistical associations, that it is difficult to believe that the effect, with a small number of specific exceptions, can be anything but minor. There will be a reduced incidence of trauma to incisors where treatment reduces their prominence. However, treatment needs to be carried out early (before the child is 10-years-old) if the peak incidence of trauma is to be avoided (Jarvinen, 1979). The avoidance of tooth impaction is also desirable. The main benefit to the patient of orthodontic treatment may be in improved aesthetics and social-psychological well-being, and additionally the effect this may have on attitudes to dental health. This has important implications in the Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 Peter H. Brook* and William C. S h a w " 310 construction of any treatment priority index. Such an index must involve an aesthetic assessment, and allow appropriate weighting for this component. Subjects and materials Development of the dental health component With these criteria in mind, the index of treatment priority used by the Swedish Dental Board (Linder-Aronson, 1974) was used as the basis for grading the functional and dental health indications for treatment. There arefivegrades, grade 1 representing little or no need for treatment and grade 5 representing great need of treatment (Table 1). An attempt was made to try to establish from the literature meaningful values for cut-off points between grades for each occlusal trait that represents a quantifiable threat to the dentition. Most of the traits are recorded using a millimetre rule, modified to incorporate a device for angular measurements. Crowding was recorded by measuring the largest displacement between teeth in the arches, using a modified version of the index described by Lau et al. (1984). In use, only the highest scoring trait need be recorded, as this determines the grading of the patient. The aesthetic component The second part of the overall assessment of treatment priority, was to record the aesthetic impairment contributed by the malocclusion. For this component the SCAN Index (Standardized Continuum of Aesthetic Need) was utilized (Evans and Shaw, 1987). This scale was constructed using dental photographs of 1000 12-year-olds collected during a large multi-disciplinary survey. Six non-dental judges rated these photographs on a visual analogue scale, and at equal intervals along the judged range, representative photographs were chosen giving a 10-point scale from 0.5 (attractive dental appearance) to 5.0 (unattractive dental appearance) (Fig. 1). Testing the index Two-hundred and twenty-two patients referred to a regional orthodontic centre for advice or treatment were examined under ideal lighting conditions with radiographs available. Both components of the index were applied and the patients were also asked to give their own rating on the aesthetic scale. Intra-examiner error was estimated by the same examiner seeing 67 of these patients on two occasions, at least one week apart, without reference to notes. A second examiner assessed 72 of the patients independently of the first examiner to estimate inter-examiner error. To simulate the use of the indices in a screening programme, 333, 11-12-year-old school children were examined. A cross section of schools attended by children from a broad range of social backgrounds were visited, and all available children in thefirstyear of secondary education were examined. Each child was examined in the school medical room using an angle-poise lamp for lighting, a simple millimeter rule and a dental mirror. Again, both components of the index were applied, and the self rating recorded on the SCAN scale. In addition a dental surgery assistant recorded her rating on the SCAN scale. Forty-six children were chosen at random for reexamination. Amongst the school sample 58 (17.4 per cent) of the children were undergoing, or had completed orthodontic treatment. The orthodontist involved in the treatment of each of these children was contacted so that details of the original malocclusion and information from the Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 The following criteria were considered to be important in the development of a new index of orthodontic treatment need. 1. Separate components to record: (a) Functional and dental health indications for treatment. (b) Aesthetic impairment. 2. For the functional and dental health component, each occlusal trait thought to contribute to the longevity and satisfactory functioning of the dentition, needs to be defined, and easily measurable cut-off points between each grading need to be established. In view of the uncertainty of the relative contribution that each occlusal trait makes to the longevity and satisfactory functioning of the dentition and indeed the doubt surrounding the importance of aesthetics in the provision of orthodontic care, the indices needed to be sufficiently flexible to allow for adjustment of cut-off points and relativities between the categories, as the results from adequate longitudinal research become available. PETER H. BROOK AND WILLIAM C. SHAW 311 ORTHODONTIC TREATMENT PRIORITY Table 1 Index of orthodontic treatment need dental health component: for use on patients. Grade 4—Great Increased overjet greater than 6 mm but less than or equal to 9 mm. Reverse overjet greater than 3.5 mm with no reported masticatory or speech difficulties. Reverse overjet greater than 1 mm but less than or equal to 3.5 mm with reported masticatory or speech difficulties. Anterior or posterior CTOssbites with greater than 2 mm displacement between retruded contact position and intercuspal position. Posterior lingual crossbites with no occlusal contact in one or both buccal segments. Servere displacement of teeth greater than 4 mm. Extreme lateral or anterior open bite greater than 4 mm. Increased and complete overbite causing notable indentations on the palate or labial gingivae. Patient referred by colleague for collaborative care e.g. periodontal, restorative or TMJ considerations. Less extensive hypodontia requiring pre-restorative orthodontics or orthodontic space closure to obviate the need for a prosthesis (not more than 1 tooth missing in any quadrant). Grade 3—Moderate Increased overjet greater than 3.5 mm but less than or equal to 6 mm with incompetent lips at rest. Reverse overjet greater than 1 mm but less than or equal to 3.5 mm. Increased and complete overbite with gingival contact but without indentations or signs of trauma. Anterior or posterior crossbite with less than or equal to 2 mm but greater than 1 mm displacement between retruded contact position and intercuspal position. Moderate lateral or anterior open bite greater than 2 mm but less than or equal to 4 mm. Moderate displacement of teeth greater than 2 mm but less than or equal to 4 mm. Grade 2—Little Increased overjet greater than 3.5 mm but less than or equal to 6 mm with lips competent at rest. Reverse overjet greater than 0 mm but less than or equal to 1 mm. Increased overbite greater than 3.5 mm with no gingival contact. Anterior or posterior crossbite with less than or equal to 1 mm displacement between retruded contact position and intercuspal position. Small lateral or anterior open bites greater than 1 mm but less than or equal to 2 mm. Pre-normal or post-normal occlusions with no other anomalies. Mild displacement of teeth greater than 1 mm but less than or equal to 2 mm. Grade 1—None Other variations in occlusion including displacement less than or equal to 1 mm. study models could be used in place of the recordings taken at the school visits. In practice, as many of the children were only just commencing treatment, the gradings were little changed. Results Reproducibility of the index Dental health component Intra-examiner agreement ranged from a Kappa value of 0.837 for the referred population seen under ideal conditions, to 0.754 for the nonreferred population. In total there were 14 errors out of 118 re-tests and in all cases the disagreement was only by one grade. Guidelines for the interpretation of the Kappa statistic (Landis and Koch, 1977) are shown in Table 2. Inter-examiner agreement ranged from 0.7310.797. In total there were 21 out of 154 measurements that were not agreed. There were only 2 cases where the error was by more than one grade. SCAN component For the referred sample there were three raters; the patient and the two examiners. This gave two patient ratings (PI and P2), two ratings by Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 Grade 5—Very great Defects of deft lip and/or palate. Increased overjet greater than 9 mm. Reverse overjet greater than 3.5 mm with reported masticatory or speech difficulties. Impeded eruption of teeth (with the exception of third molars) due to crowding, displacement, the presence of supernumerary teeth, retained deciduous teeth and any other pathological cause. Extensive hypodontia with restorative implications (more than one tooth missing in any quadrant) requiring pre-restorative orthodontics. 81:4 PETER H. BROOK AND WILLIAM C. SHAW 05 1-5 45 25 Figure 1 The SCAN Scale. Originally presented in colour in a horizontal arrangement. 0.5 extreme left. 5 extreme right. Table 2 Guidelines for the interpretation of Kappa. Kappa statistic Strength of agreement <0.00 0.00-0.20 0.21-0.40 0.41-0.60 0.61-0.80 0.81-1.00 Poor Slight Fair Moderate Substantial Almost perfect Table 3 Examiner variability. Pearson's correlation coefficients for SCAN. Referred Population PB1 PB2 PI examiner 1 (PB1 and PB2), and one rating by examiner 2 (WCS). The examiner reproducibility, and the comparability of patient and examiner ratings, were investigated using Pearson's correlation coefficient. The results are listed in Table 3 with the number of repeat examinations in parenthesis. Whilst the correlations between the orthodon- P2 PB2 PI P2 WCS 0.87 (72) 0.50 (82) 0.45 (72) 0.36 (72) 0.40 (72) 0.67 (72) 0.71 (82) 0.73 (72) 0.37 (82) 0.29 (72) tists were quite high, they were poorer than those obtained by Evans and Shaw (1987), where selfretaining lip retractors were used during the orthodontists' and the patients' assessments. For Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 35 313 ORTHODONTIC TREATMENT PRIORITY General features of the referred and non-referred populations The numbers of patients falling into each Dental Health Index grade for each group are illustrated Table 4 Examiner variability. Pearson's correlation coefficients for SCAN. School Population DSA1 DSA2 SI S2 PB1 PB2 0.78 (46) 0.66 (46) 0.61 (46) 0.70 (46) 0.69 (46) 0.78 (46) 0.80 (46) 0.85 (46) 0.64 (46) 0.68 (46) 0.81 (46) 0.88 (46) 0.61 (46) 0.69 (46) 0.95 (46) DSA2 SI S2 PB1 Table 5 Distribution of Dental Health grades. Dental health grade Referred population School population Numbers Numbers Grade 1 Grade2 Grade 3 Grade 4 Grade 5 1 11 40 112 39 0.5 5.4 19.7 55.2 19.2 24 93 107 92 17 7.2 27.9 32.1 27.6 5.1 Total 203 100.0 333 100.0 Percentage Percentage in Table 5, and Figs. 2 and 3. The SCAN Index scores are illustrated in Table 6, and Figs. 4 and 5. Discussion Reproducibility Dental health component In general, the reproducibility of this index was very good. The same grade was re-chosen 86.4 per cent of the time with different examiners, and in 93 per cent of cases for the same examiner under the more ideal clinic setting (the referred population). The common traits causing disagreement, in descending order of frequency were; crowding, increased overjet, crossbites and overbites. Crowding represents a problem in recording when the patient is in the mixed dentition. Further refinement of the index in terms of the mixed dentition analysis of crowding, may lead to an improvement in reproducibility. Evidently, the less than ideal conditions of the school examination resulted in poorer reproducibility. Should reproducibility levels similar to that of the referred population be required, then better lighting, better patient seating facilities and a more relaxed work rate would be required. SCAN component Whilst the correlation coefficients for the SCAN ratings were reasonably high for the school survey, they were less satisfactory for the referred population. It was felt that this may have been due to the omission of the self-retaining lip retractors for this sample. During the original development of the index, both front and side views of the dentition were available. This enabled conditions such as large overjets to be more readily assessed. During the surveys, it was noted that the areas around 1.5 and 4.0 on several raters graphs, showed slight inconsistencies. An inspection of the index demonstrates the difficulties that may be leading to these effects. Both the representations of 2.0 and 4.5 show increased overjets, that a lay person in the absence of a side view, may not find too displeasing. So, in summary, incorporation of side views may assist in identifying large overjets. Some guidelines for assessing the relative attractiveness of features not depicted on the scale, may also be Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 this reason, self-retaining cheek retractors were used routinely for the examination of the nonreferred sample. In the school survey, 46 subjects were rated twice by the orthodontist (PB) on the SCAN scale. Additionally, there was a rating by the dental surgery assistant on two occasions (DSA1 and DSA2), and two subject ratings (SI and S2). Again, these relationships were examined using Pearson's correlation coefficient (Table 4). On this occasion intra-examiner agreement for the orthodontist (PB) was better. The DSA's reproducibility was less good. There was however, superior inter-examiner agreement on this occasion. There was also better correlation between the subjects' and the professionals' ratings. 314 PETER H. BROOK AND WILLIAM C. SHAW REFERRED POPULATION DHI 2 3 4 5 DENTAL HEALTH INDEX SCORE Figure 2 Distribution of ratings for the Dental Health Index obtained from examination of 222 patients referred to a regional orthodontic centre. desirable. Self-retaining cheek retractors are a useful aid to recording dental aesthetics. The referred sample As expected, the referred sample showed a large proportion of patients scoring in the higher grades of the Dental Health Index, with all but 6 per cent in the highest three grades. The orthodontists' SCAN ratings showed a similar shift to the unattractive side. The patient ratings showed this effect to a much smaller extent. The school population There was a much more even spread of patients amongst the grades of the Dental Health Index for this population, with approximately one third of the subjects in grade 3, and one third either side of this. The professional ratings using the SCAN Index, showded a distribution skewed towards the attractive end of the scale. The patient ratings are skewed even further towards the attractive end of the scale, i.e. there was a tendency for subjects to overrate their dental attractiveness. Comparison of the two samples From an overview of the data recorded from the samples, it appears quite obvious that there are significant differences between them. Indeed the median test and the /-test demonstrate this readily for the Dental Health component and the SCAN component respectively (p< 0.0001). However, the findings from the two surveys are not directly comparable. Firstly, the referred sample had a wide age spread, and secondly, when applying the Dental Health component to Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 1 315 ORTHODONTIC TREATMENT PRIORITY SCHOOL POPULATION DHI 2 3 4 5 DENTAL HEALTH INDEX SCORE Figure 3 Distribution of ratings for the Dental Health Index obtained from examination of 333 unselected schoolchildren. the referred population, the examiners had access to radiographs. For the school sample, it was necessary to set criteria for such parameters as unerupted or missing teeth, i.e. except for incisors and first molars, all teeth were assumed to be present, at age 11-12 years, premolars and canines were assumed to be unerupted but not impacted, missing upper lateral incisors and lower incisors were assumed to be developmentally absent whereas missing upper central incisors and first molars were assumed to have been extracted. In an attempt to overcome this, a sub-group of 25 from the referred sample, with a similar age range to the school sample was selected. Any conditions that would have required radiographs to confirm the diagnosis, were regraded using the Table 6 score). SCAN rating 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Total Distribution of SCAN ratings (Orthodontist Referred Population School 1Population Numbers Numbers 0 2 19 17 28 33 39 48 8 8 202 Percentage 0.0 1.0 Percentage 8.4 13.8 16.3 19.2 23.6 3.9 3.9 16 54 65 59 62 33 26 12 5 1 4.8 16.2 19.5 17.7 18.6 9.9 7.8 3.6 1.5 0.3 100.0 333 100.0 9.4 Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 1 316 PETER H. BROOK AND WILLIAM C. SHAW REFERRED POPULATION SCAN 5 6 7 8 SCAN RATING Figure 4 Distribution ot raungs for the SCAN index obtained from examination of 222 patients referred to a regional orthodontic centre. criteria set for the population seen without radiographs. It was still evident that the samples were drawn from different populations (Mest p< 0.001 for the SCAN ratings; median test p< 0.001 for the Dental Health ratings). These tests seemingly validate the index, at least in terms of the priorities of patients or dentists in bringing about referral to an orthodontist. acceptable as it has been shown that most of the traits can be recorded with a high degree of precision (Helm et al., 1975; Helm, 1977) with up to 80 per cent agreement. However in their pure form they do not record treatment priority. The allocation of weighting factors to traits can give an overall figure that is intended to represent a score of severity, and thus treatment priority. Several indices of this type have been developed (Summers, 1971; Draker, 1960; GraComparison with previous methods of recording inger, 1967). Correlation coefficients for exatreatment priority miner agreement for such indices have ranged Angle's classification (Angle, 1899) has been from a Spearman correlation coefficient of 0.903 shown to have poor reproducibility (Gravely and (Summers, 1971) to as low as 0.34 (Albino et al., Johnson, 1974) and has no usefulness in record1978) in a community screening setting. The ing treatment priority. For epidemiological use, validity of such indices relies on acceptance of the the registration techniques described by Bjork et authors' weightings. al. (1964) and Baume et al. (1973) may be quite Indices based upon the classification of mor- Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 4 317 ORTHODONTIC TREATMENT PRIORITY SCHOOL POPULATION SCAN 30 P E 20 R C E 15 N T A G 10E 19.5 17.7 18.6 9.9 7.8 5 - 3.6 1.5 1 2 3 4 5 6 7 8 0.3 10 SCAN RATING Figure 5 Distribution of ratings for the SCAN index obtained from examination of 333 unselected schoolchildren. phological traits rely on the subjective opinion of an experienced judge to define the dividing lines between each trait (Linder-Aronson, 1974; Lundstrom, 1977; Malmgren, 1980). Used as such, the percentage concordance ranges from 55.9 to 74.6 per cent (Malmgren, 1980). Subjective clinical opinion alone has agreement of about 80 per cent in most studies (Bowden and Davies, 1975; Helm et ai, 1975) but the validity of such judgements depends upon the examiners' knowledge of the harmful effects of malocclusion. In addition, inexperienced examiners will find it difficult to apply such techniques. The orthodontic index of treatment need" described in this report has examiner agreement levels that compare well with any of those previously described (80.5-93 per cent). As its development was based upon a full analysis of the available literature (Brook, 1987) and the experience gained from a longitudinal survey (Shaw et ai, 1986), it is not felt that its validity can be inferior to that of other indices. The inclusion of a separate index to record aesthetic impairment removes the most subjective element from indices of this kind. Good levels of agreement for this component have been demonstrated (Pearson's correlation coefficient values from 0.71-0.95). Some support for the validity of the index comes from the observation that fewer subjects in the lower grades were referred for orthodontic advice. The extent to which it represents common professional opinion is presently being evaluated. However, true validity (i.e. that the index measures what it purports to measure) Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 25- 318 PETER H. BROOK AND WILLIAM C. SHAW Table 7 Interrelationship between the SCAN and Dental Health Index scores. (Cumulative Percentages) SCAN Score 0.5 1.0 1 2 3 4 5 2.1 4.2 4.5 4.8 4.8 Total 4.8 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Total 5.4 15.6 19.5 20.4 21.0 6.3 25.8 37.2 39.6 40.5 6.6 31.8 51.3 57.3 58.5 7.2 34.8 60.9 75.3 76.8 7.2 35.1 64.5 84.9 86.7 7.2 35.1 65.7 91.5 94.5 7.2 35.1 67.2 93.9 98.2 7.2 35.1 67.2 94.8 99.7 7.2 35.1 67.2 94.8 100.0 7.2 35.1 67.2 94.8 100.0 21.0 40.5 58.5 76.8 86.7 94.5 98.2 99.7 100.0 100.0 must await the compilation of a greater body of knowledge than that which is currently available. It may accurately reflect contemporary professional opinion, but this may be erroneous. Epidemiological uses for the index During testing of the index on the school sample, it was felt that it represented a simple, quick and reasonably reproducible method of recording orthodontic treatment need. As all the traits are simple to record, it may be possible for less highly trained personnel to apply the index, following suitable training and calibration. tions could produce a similar percentage. An alternative method would be to combine the scores from the two components to give an overall score, then define limits based on this figure. Eventual definitions of cut-offs must reflect the setting in which treatment would be provided and include a consideration of the success rate of the treatment which would be available, the iatrogenic risks, and the cost (Shaw, 1987). Conclusions An index with two components has been developed to record orthodontic treatment priority. The first of these components records need for As developed so far, the index records the dental treatment on dental health and functional health need for treatment, and the aesthetic grounds. The second component records the impairment, and by implication the socialaesthetic impairment, and by implication, the pschological need for treatment. As yet, no justification for treatment on social-psychologiattempt has been made to combine these into an cal grounds. overall assessment of treatment need, or to define scores below which treatment should be withThe indices were tested on a sample of patients held. referred for orthodontic treatment and advice, To assist discussion in this area, a table and on a random selection of 11-12-year-old showing the cumulative percentages of patients school children. It was easy and quick to use and having varying combinations of the Dental had acceptable reproducibility. True validation Health grades, and the SCAN ratings has been of such an index must await the emergence of constructed from the non-referred sample (Table further research data on the effects of malocclu7). sion, but the present index can be adaptable to new information. Work is currently in progress Many authors quote figures of around 50 per to gauge the extent to which the index reflects cent for the percentage of children who would common professional opinion. benefit from orthodontic treatment (Gardiner, 1956; Haynes, 1982; Foster and Walpole Day, Defining specific ranges within which patients 1973). From Table 7, it can be seen that this should, or should not be offered treatment has number would be obtained if patients scoring 'not been attempted but a mathematical model grade 3 or less on the Dental Health Index and has been suggested that can define combinations 2.0 or less on the SCAN scale were excluded (51.3 of the gradings that will encompass varying per cent), leaving 48.7 per cent. Other combinaproportions of a target population. Further development of the index Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 DHI Grade ORTHODONTIC TREATMENT PRIORITY Acknowledgements Address for correspondence Professor W C Shaw Department of Orthodontics University Dental Hospital of Manchester and Turner Dental School Higher Cambridge Street Manchester M15 6FH England References Addy M, Griffiths G S, Duramer P M H, Kingdon A, Hicks R, Hunter M L, Newcome R G, Shaw W C 1988 The association between tooth irregularity and plaque accumulation, gingivitis, and caries in 11-12-year-old children. European Journal of Orthodontics 10: 76-83 Albino J E, Lewis E A, Slakter M J 1978 Examiner reliability for two methods of assessing malocclusion. Angle Orthodontist 48: 297-302 Angle E H 1899 Classification of malocclusion. Dental Cosmos 41: 248-264 Baume L J, Horowitz H S, Summers C J, Backer Dirks O, Brown W A B, Carlos J P, Freer T J, Harvold E P, Moorrees C F A, Saltzman J A, Schmuth G, Solow B, Taatz H 1973 A method of examining occlusal traits developed by the FDI commission on classification and statistics for oral conditions (COCSTOQ International Dental Journal 23: 530-537 Bjork A, Krebs Aa, Solow B 1964 A method for epidemiological registration of malocclusion. Acta Odontologica Scandinavica 22: 27-41 Bowden D E J, Davies A P 1975 Inter- and intra-examiner variability in assessment of orthodontic treatment need. Community Dentistry and Oral Epidemiology 3: 198-200 Brook P H 1987 The development of an orthodontic treatment priority index. M.Sc. Thesis, University of Manchester Davies T M, Shaw W C, Addy M, Dummer P H 1988 The relationship of anterior overjet to plaque and gingivitis in children. American Journal of Orthodontics 93: 303-309 Draker H L I960 Handicapping labio-lingual deviations: a proposed index for public health purposes. American Journal of Orthodontics 46: 295-315 Evans M R and Shaw W C 1987 Preliminary evaluation of an illustrated scale for rating dental attractiveness. European Journal of Orthodontics 9: 314-318 Foster T D, Walpole Day A J 1973 A survey of malocclusion and the need for orthodontic treatment in a Shropshire school population. British Journal of Orthodontics 3: 7378 Gardiner J H 1956 A survey of malocclusion and some aetiological factors in 1000 schoolchildren. The Dental Practitioner 6: 187-201 Grainger R M 1967 Orthodontic Treatment Priority Index. Public Health Service Publication No 1000, Series 2, No. 25. Washington DC, U.S. Government Printing Office. Gravely J F, Johnson D B 1974 Angle's classification of malocclusion: an assessment of reliability. British Journal of Orthodontics 3: 79-86 Haynes S 1982 Discontinuation of orthodontic treatment in the General Dental Service in England and Wales 19721979. British Dental Journal 152: 127-129 Helm S 1977 Intra-examiner reliability of epidemiological registrations of malocclusion. Acta Odontologica Scandinavica 35: 161-165 Helm S, Kreiborg S, Barlebo J, Caspersen J, Eriksen J H, Hansen W, Hanusardottir B, Munck C, Perregaard J, Prydson U, Reumert C, Spedtsberg H 1975 Estimates of orthodontic treatment need in Danish schoolchildren. Community Dentistry and Oral Epidemiology 3: 136-142 Horup N, Melsen B, Terp S 1987 Relationship between malocclusion and maintenance of teeth. Community Dentistry and Oral Epidemiology 15: 74-78 Jarvinen S 1979 Traumatic injuries to upper permanent incisors related to age and incisal overjet. Acta Odontologica Scandinavica 37: 335-338 Landif. J R, Kock G G 1977 The measurement of observer agreement for categorical data. Biometrics 33: 159-174 Lau D, Griffiths G, Shaw W C 1984 Reproducibility of an index for recording the alignment of individual teeth. British Journal of Orthodontics 11: 80-84 Linder-Aronson S 1974 Orthodontics in the Swedish Public Dental Health System. Transactions of the European Orthodontic Society 233-240 Lundstrom A 1977 Need for treatment in cases of malocclusion. Transactions of the European Orthodontic Society 53: 111-123 Malmgren O 1980 Studies on the need and demand for orthodontic treatment. Swedish Dental Supplement 6 Mohlin B, Thilander B 1984 The importance of the relationship between malocclusion and mandibular dysfunction and some clinical applications in adults. European Journal of Orthodontics 16: 192-204 Roth R H 1973 Temporomandibular pain-dysfunction and occlusal relationships. Angle Orthodontist 43: 136-154 Salzmann J A 1968 Handicapping malocclusion assessment to establish treatment priority. American Journal of • Orthodontics 54: 749-765 Sandali T 1973 Irregularities of the teeth and their relation to the periodontal condition with particular reference to the lower labial segment. Transactions of the European Orthodontic Society 319-333 Shaw W C 1988 Risk benefit appraisal in orthodontics. In orthodontics: evaluation and future. Proceedings of the Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 The authors would like to thank Mr G. O. Taylor and his staff for providing access and assistance for the school visits, Mrs H. Worthington and Mrs C. Mitropoulos for advice with the study design and statistics, Mr J. Sinclair for computational assistance and Mrs C. Corkill for acting as a scribe. The study was supported in part by a grant from the DHSS and BLG. 319 320 international conference on the occasion of the 25th Anniversary of the Orthodontic Department of the University of Nymegen, The Netherlands. October 22-24, 1987. Editors: C F A Moorrees and F P G M van der Linden. Pp 63-81 Shaw W C, Addy M, Ray C 1980 Dental and social effects of malocclusion and effectiveness of orthodontic treatment: a review. Community Dentistry and Oral Epidemiology 8: 36-45 Summers C J 1971 The occlusal index: a system for identifying and scoring occlusal disorders. American Journal of Orthodontics 57: 552-567 Downloaded from http://ejo.oxfordjournals.org/ at Aarhus Universitets Biblioteker / Aarhus University Libraries on May 21, 2012 Shaw W C, Addy M, Dummer P M H, Ray C, Frude N 1986 Dental and social effects of malocclusion and effectiveness of orthodontic treatment: a strategy for investigation. Community Dentistry and Oral Epidemiology 14: 60-64 PETER H. BROOK AND WILLIAM C. SHAW