- Email: [email protected]
Patient- and clinician-perceived need for orthognathic surgery
ORIGINAL ARTICLE
Patient- and clinician-perceived need for orthognathic surgery Karen J. Juggins,a Fiona Nixon,b and Susan J. Cunninghamc London and Oxford, United Kingdom Introduction: To better understand why patients believe they need orthognathic treatment, a study was undertaken to examine perceived need for treatment by patients and clinicians. This questionnaire-based study was undertaken at The John Radcliffe Hospital, Oxford, United Kingdom. Subjects and methods: Forty patients were recruited from combined orthodontic-surgical clinics. They were asked to rate their perceived need for treatment based on facial appearance, dental appearance, function, and overall need. Twenty orthodontists and 20 maxillofacial surgeons were asked to rate perceived need for treatment based on the same parameters, using study models and clinical photographs. Ratings were marked on visual analog scales. Results: Significant differences were found between patients and clinicians in perceived need for treatment based on facial appearance (orthodontists compared with patients, P ⫽ .023; surgeons compared with patients, P ⫽ .001). In addition, maxillofacial surgeons rated a significantly greater overall need for treatment than patients (P ⫽ .027), and they rated treatment need based on facial appearance (P ⫽ .005) and function (P ⬍ .001) significantly higher than orthodontists. Conclusions: Clinicians rated greater need for orthognathic treatment based on facial appearance than did patients. Surgeons also rated greater overall need for treatment than patients. In addition, surgeons rated treatment need based on facial appearance and function significantly higher than orthodontists, but large variations existed in both clinician groups. (Am J Orthod Dentofacial Orthop 2005;128:697-702)
I
t has been shown that facial and dental anomalies that are sufficient to affect a person’s appearance might put that person at a social disadvantage.1,2 When a dentofacial deformity is so severe that an acceptable improvement cannot be obtained by growth modification or orthodontic camouflage, a combination of orthodontic treatment and orthognathic surgery might be the only viable treatment option. As orthognathic treatment has become more widely available and more socially acceptable, the demand for it has increased enormously. What used to be considered a variant of normal appearance is now less acceptable in society. This has led to treatment for relatively moderate “defects” and also to more extensive surgical procedures. The importance of careful assessment to ascertain what the patient expects from the procedure a
Specialist registrar, Department of Orthodontics, The John Radcliffe Hospital, Oxford, and Eastman Dental Institute, University College London, London, United Kingdom. b Consultant, Department of Orthodontics, John Radcliffe Hospital, Oxford, United Kingdom. c Senior lecturer/honorary consultant in orthodontics, Eastman Dental Institute, University College London, London, United Kingdom. Reprint requests to: K. J. Juggins, Department of Orthodontics, John Radcliffe Hospital, Oxford OX3 9DU United Kingdom; e-mail, [email protected]. Presented in June 2004 at the European Orthodontic Congress in Denmark; received the WJB Houston Research Award. Submitted, March 2004; revised and accepted, September 2004. 0889-5406/$30.00 Copyright © 2005 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2004.09.022
cannot be overemphasized. It is easy to imagine that there is a linear relationship between the severity of psychological disturbance and the degree of deformity, but clinical experience has shown that a mild deformity can be harder to cope with than a more severe problem.3,4 Many patients who request orthognathic treatment fall into the milder-deformities group and might be more severely affected psychologically than is apparent at initial consultation.4 Combined orthodontic and surgical treatment aims to produce more harmonious facial skeletal and soft tissue relationships and to improve occlusal function. Research has shown that most patients who request orthognathic treatment do so because of a desire to improve their facial or dental appearance and not because of concerns about occlusal function.5,6 An appropriate plan of combined orthodontic and surgical treatment should be based not only on the clinician’s assessment of esthetics, function, and stability, but also on the patient’s objectives and perceptions of need. In addition, clinicians tend to be more critical than laypersons in their assessments, particularly when rating facial profiles.7-9 Planning orthognathic treatment requires clinical guidelines, or a generally agreed “ideal” for facial proportions. These esthetic standards exist in various forms including artistic views, cephalometric analyses, and anthropometric measurements. It therefore seems 697
698 Juggins, Nixon, and Cunningham
American Journal of Orthodontics and Dentofacial Orthopedics December 2005
Fig 1. Example of visual analog scale.
reasonable to assume that orthodontists and surgeons would share common clinical objectives, but it must be questioned whether this collective judgment of attractiveness is always valid. This study aimed to investigate which aspects of appearance and function patients consider when deciding on orthognathic treatment. This is important to both orthodontists and maxillofacial surgeons, who need to know whether their views of the expected improvement will coincide with those of the patients. The following questions were posed: 1. Does patient-perceived need for orthognathic treatment correlate with clinician-perceived need for treatment? 2. Do orthodontists and maxillofacial surgeons share similar views on need for treatment? 3. Is there consistency within the clinician groups? SUBJECTS AND METHODS
This was a questionnaire-based study undertaken at John Radcliffe Hospital, Oxford, United Kingdom. Ethical approval was obtained, and all patients consented in writing. Forty consecutive new patients (16 men, 24 women) referred by general dental practitioners, fulfilling the following criteria, were recruited from orthognathic clinics: over 16 years of age, about to undergo orthognathic treatment, no history of orthodontic treatment or facial surgery, and no congenital deformities (eg, clefts) or traumatically acquired deformities. Forty senior clinicians (20 orthodontists, 20 maxillofacial surgeons), all experienced in orthognathic treatment, were also recruited. They rated each patient, using study models and standardized orthodontic extraoral and intraoral photographs. No clinicians were actively involved in the care of these patients. The questionnaire required responses to be marked on 10-cm visual analog scales (VAS) that were anchored at
each end, with 0 representing little need for treatment and 10 representing great need for treatment. Patients and clinicians were asked to mark the perceived need for treatment based on: facial appearance, dental appearance, function/anticipated functional problems (clinicians), and overall treatment need (Fig 1). To ensure accurate completion of the VAS and to prevent raters from placing a mark randomly on the same part of the scale for all scales, the anchors for the final question, regarding overall need for treatment, were reversed (0 represented great need for treatment, and 10 represented little need for treatment). To ensure repeatability of the study, both patient and clinician repeatability studies were undertaken. Ten patients were randomly selected to repeat the questionnaire 4 to 6 weeks later but before starting treatment. In addition, 10 clinicians (5 maxillofacial surgeons, 5 orthodontists) were asked to rate 10 patients a second time 2 to 4 weeks later. Statistical analyses
The Bland-Altman method,10 including the paired t test, was used to compare perceived need for treatment between patients and clinicians, and between orthodontists and surgeons. Variation within the clinician groups was studied by using box plots and calculating standard deviations. Patient repeatability was also assessed with the Bland-Altman method and the intraclass correlation coefficient (ICC).11 Clinician repeatability was measured with the ICC within observers.12 This test considers variation within each observer, for each subject, as a proportion of all types of variation. RESULTS
In general, there was good repeatability among patients. The ICC results were all within the acceptable range, with all values above 0.70. However, the Bland-
Juggins, Nixon, and Cunningham 699
American Journal of Orthodontics and Dentofacial Orthopedics Volume 128, Number 6
Table I.
Agreement between patients and orthodontists (values in cm)
Variable
Mean difference
SD
⫺1.00 0.00 0.13 0.68
2.67 3.62 3.24 2.29
Facial appearance Dental appearance Function Overall need for treatment
95% limits of agreement ⫺6.2 ⫺7.1 ⫺6.2 ⫺3.8
to to to to
4.2 7.1 6.5 5.2
P value from t test .023 NS NS NS
NS, Not significant. Table II.
Agreement between patients and maxillofacial surgeons (values in cm)
Variable Facial appearance Dental appearance Function Overall need for treatment
Mean difference
SD
⫺1.41 ⫺0.21 ⫺0.42 0.83
2.53 3.48 3.22 2.28
95% limits of agreement ⫺6.4 ⫺7.0 ⫺6.7 ⫺3.6
to to to to
3.6 6.6 5.9 5.3
P value from t test .001 NS NS .027
NS, Not significant. Table III.
Agreement between orthodontists and maxillofacial surgeons (values in cm)
Facial appearance Dental appearance Function Overall need for treatment
Mean difference
SD
⫺0.40 ⫺0.21 ⫺0.54 0.15
0.85 0.69 0.54 0.75
95% limits of agreement ⫺2.1 ⫺1.6 ⫺1.6 ⫺1.3
to to to to
1.3 1.1 0.5 1.6
P value from t test .005 NS ⬍.001 NS
NS, Not significant.
Altman method highlighted a slight bias in the scoring of dental appearance and overall need for treatment. In general, clinicians exhibited good repeatability when rating perceived need for treatment based on facial appearance, dental appearance, and function. Repeatability for overall need for treatment was below 0.70 but was still considered acceptable at r ⫽ 0.63. Of the 40 patients recruited for this study, 1 had a Class I malocclusion (with an anterior open bite), 22 had Class II malocclusions, and 17 had Class III malocclusions. Each patient was given a score based on the index of orthodontic treatment need13 dental health component. Two patients scored 3 on the index of orthodontic treatment need, 19 scored 4, and 19 scored 5. A comparison was made between the patient score for each variable (facial appearance, dental appearance, function, and overall need for treatment) and the mean scores of the orthodontists and the surgeons for that patient. The mean difference was then calculated by subtracting the mean clinician score from the patient score. The results showed that, when rating facial appearance, there was significant bias, with orthodon-
tists consistently rating a greater need for treatment than the patients (Table I). Similarly, there was significant bias when comparing patient scores and mean scores for the maxillofacial surgeons, with the surgeons rating a greater need for treatment based on facial appearance and overall need for treatment (Table II). A comparison was made between the mean orthodontist score and the mean surgeon score for each variable (Table III). The mean difference was calculated by subtracting the surgeon score from the orthodontist score. The results showed that there was significant bias between maxillofacial surgeon and orthodontist scores, with the maxillofacial surgeons rating greater need for treatment based on facial appearance and function. The negative values for mean difference (⫺0.40 and ⫺0.54) indicate that surgeons scored treatment need higher than orthodontists. Although not statistically significant, the negative score for dental appearance and the positive score for overall need both indicate that surgeons, in general, perceived a greater need for treatment than did orthodontists. Although the values for facial appearance and function were significant, the values for the mean differences
700 Juggins, Nixon, and Cunningham
American Journal of Orthodontics and Dentofacial Orthopedics December 2005
Fig 2. Box plot showing variation among orthodontists rating treatment need based on facial appearance. Box represents range within which 50% of values lie. Black horizontal line represents median. Whiskers mark extremes.
were small compared with those in Tables I and II. The 95% limits of agreement were also relatively narrow compared with Tables I and II. The VAS scores for each clinician, rating each patient, were used to create box plots, which were constructed for each of the 4 VAS for both orthodontists and surgeons. The box plots show the distribution of the observations, within a group (eg, orthodontists or surgeons) for each patient. Figure 2 shows a box plot of the orthodontist group rating need for treatment based on facial appearance. This plot shows large variation; similar plots were seen for all variables in both clinician groups. Standard deviations were calculated and used to measure variation within a group. In general, the standard deviation varied between 1.5 and 2.5 cm, and the range within which the 95% limits of agreement fell was wide. DISCUSSION
Although the ICC scores for patient repeatability were all high (⬎.70), the Bland-Altman method highlighted a slight bias for dental appearance and overall need for treatment. This might be a reflection of the small sample size for the repeatability study. In addition, there could be a tendency for patients to modify their scores after considering the initial rating. How-
ever, the mean differences in the repeatability study were small compared with those in the main study, and this, considered in conjunction with the high ICC values, suggested adequate repeatability. The ICC within observers was used to evaluate the extent to which clinicians obtained similar scores when rating the same patient for a second time. It was expected that, with the same information, clinicians would make a similar decision on treatment need on both occasions. In this study, clinicians were reliable when rating facial appearance, dental appearance, and function, although slightly less so when rating overall need for treatment (ICC ⫽ 0.63). However, this was still deemed to be acceptable and might reflect the relatively small sample size of this repeatability study. Significant differences were found between patients and clinicians (both orthodontists and surgeons) for perceived need for treatment based on facial appearance (orthodontists P ⫽ .023, maxillofacial surgeons P ⫽ .001). The results suggest that clinicians perceive a greater need for alteration of the facial appearance than do patients. However, the 95% limits of agreement were wide (orthodontists compared with patients, ⫺6.2 to 4.2 cm; maxillofacial surgeons compared with patients, ⫺6.4 to 3.6 cm), suggesting large variability. It has been widely reported that clinicians are more critical of facial appearance than laypeople,7-9 and
Juggins, Nixon, and Cunningham 701
American Journal of Orthodontics and Dentofacial Orthopedics Volume 128, Number 6
Table IV.
Measuring agreement between groups scoring perceived need for treatment based on facial appearance
Group Orthodontist and patient Surgeon and patient Orthodontist and surgeon
Mean difference
SD
95% limits of agreement
P value from t test
⫺1.00 ⫺1.41 ⫺0.40
2.67 2.53 0.85
⫺6.2 to 4.2 ⫺6.4 to 3.6 ⫺2.1 to 1.3
.023 .001 .005
significant differences in perceived need for treatment between clinicians and patients were identified in other studies.14,15 In addition, although clinical assessment requires examination in 3 dimensions, quantitative measurement of a dentofacial deformity is still predominantly carried out in the lateral or profile view, with cephalometric analysis playing an important part in assessment, planning of clinical objectives, and measurement of treatment outcome. This could explain, in part, the differences between patient and clinician assessments, with patients placing less weight on the profile view than clinicians. The clinicians used patient photographs to rate perceived need for treatment. Ideally, the study would use patients, rather than records, for all ratings, but this was not logistically possible. There is, of course, the possibility that patients might have rated themselves differently from photographs, but it was thought that their perceived need for treatment based on the “reallife” situation would be more accurate than looking at standard clinical photographs. In addition to differences for facial appearance, there was also significant bias between patient and maxillofacial surgeon scores for overall treatment need (P ⫽ .027). This differs from the orthodontist group where there was no significant difference between the scores. This supports the work of Bell et al,7 who showed that, although both groups recognized need for treatment, the orthodontists as a group perceived significantly less need for orthognathic treatment than did surgeons. There were no other significant perceived-need differences between patients, orthodontists, and maxillofacial surgeons for any other variable. According to the mean scores, orthodontists and patients rated a similar need for treatment based on dental appearance (mean difference in scores ⫽ 0.00 cm). However, if the data are studied more closely, the standard deviation was high (3.62 cm) and the 95% limits of agreement wide (⫺7.1 to 7.1 cm). Improvement of the patient’s facial and dental appearance is the main objective of orthognathic treatment. This study showed that maxillofacial surgeons rated treatment need based on facial appearance signif-
icantly higher than orthodontists (P ⫽ .005); this is consistent with Bell et al,7 who noted similar findings. They reported that, when orthodontists and surgeons evaluated the profile views of patients, the surgeon group perceived significantly greater need for orthognathic treatment than orthodontists. Surgeons also rated significantly higher treatment need based on function (P ⬍ .001). This might be because orthodontists see patients with significant skeletal and dental discrepancies more frequently and do not perceive such malocclusions to be functionally handicapping. There was no significant difference between surgeon and orthodontist ratings in terms of perceived need for treatment based on dental appearance or overall need for treatment, although the surgeons showed a tendency to rate a greater need for treatment than did the orthodontists. Although there were differences between the 2 clinician groups, Table IV shows an example of how much smaller the mean differences, standard deviations, and 95% limits of agreement were for surgeons and orthodontists compared with the clinician-patient comparison. In addition, although these differences were statistically significant, the differences between surgeons and orthodontists as groups were small (0.15-0.40 cm) and perhaps not clinically significant. Ideally, it would be expected that clinicians prescribing orthognathic treatment would share similar views of the need for treatment, with minimal variation expected within the groups. However, this study showed that this was not the case. Although mean clinician scores were used for comparison with patient data, significant variation existed within the groups of clinicians. Box plots were therefore constructed to show the distribution of the observations (ie, variation) within the groups. In general, the standard deviations and 95% ranges were high, suggesting large variation, and this variation between clinicians might be of concern. There appeared to be no relationship between sex, malocclusion type, or index of orthodontic treatment need and the extent of clinician variation. However, relationships might be seen in a larger study. Such large variation could be due to a number of factors. A true variation might exist, the sample size might have
702 Juggins, Nixon, and Cunningham
been inadequate, or the variation within the groups might be due to different levels of experience in treating orthognathic cases. In addition, some clinicians commented that it was difficult to accurately assess function from study models and photographs alone. Phillips et al16 looked at the effects of dental training on assessment of dentofacial attractiveness and found that there were significant differences between the panels of judges (orthodontic residents, dental students, undergraduate students). Such wide variation among clinicians is clearly of concern. The impact on the quality of patient treatment and care was not investigated in this study, but, given the extent of the variation, this issue could certainly be addressed in future studies. However, it is a difficult problem to both investigate and manage. Perhaps, after a similar larger scale study, it will be necessary to consider ways of standardizing the way clinicians rate need for treatment. This study showed that patients’ perceived need for treatment might not always be the same as that of clinicians. Perhaps even more important is the variability in the clinician groups. This has important implications for the treatment planning of orthognathic patients. We must be able to understand these aspects of patient health care to provide and evaluate treatment. CONCLUSIONS
The results of this study suggest that: 1. Orthodontists and maxillofacial surgeons rated a greater need for orthognathic treatment based on facial appearance than patients. 2. Maxillofacial surgeons rated a greater need for orthognathic treatment based on overall treatment need than patients. 3. Large variation existed within both the orthodontist and maxillofacial surgeon groups, but neither group appeared to exhibit more variability than the other. 4. Maxillofacial surgeons rated treatment need based on facial appearance and function significantly higher than orthodontists. 5. There was no significant difference between the clinician groups rating perceived treatment need
American Journal of Orthodontics and Dentofacial Orthopedics December 2005
based on dental appearance or overall need for treatment, but surgeons generally tended to rate a greater need than orthodontists. We thank Dr Aviva Petrie for her statistical advice; Mary McKnight, Steve Watt-Smith, and Heather Nevard for their contributions to this study; and the clinicians and patients who participated. REFERENCES 1. Shaw WC, Meek SC, Jones DS. Nicknames, teasing, harassment and the salience of dental features among school children. Br J Orthod 1980;7:75-80. 2. Heldt L, Haffke EA, Davis LF. The psychological and social aspects of orthognathic treatment. Am J Orthod 1982;82:318-28. 3. Macgregor FC. Social and psychological implications of dentofacial disfigurement. Angle Orthod 1970;40:231-3. 4. Lansdown R, Lloyd J, Hunter J. Facial deformity in childhood: severity and psychological adjustment. Child Care Health Dev 1991;17:165-71. 5. Laufer D, Glick D, Gutman D, Sharon A. Patient motivation and response to surgical correction of prognathism. J Oral Surg 1976;41:309-13. 6. Kiyak HA, Hohl T, Sherrick P, West RA, McNeill RW, Bucher F. Sex differences in motives for and outcomes of orthognathic surgery. J Oral Surg 1981;39:757-64. 7. Bell R, Kiyak HA, Joondeph DR, McNeill RW, Wallen TR. Perceptions of facial profile and their influence on the decision to undergo orthognathic surgery. Am J Orthod 1985;88:323-32. 8. Lines PA, Lines RR, Lines CA. Profilemetrics and facial esthetics. Am J Orthod 1978;73:648-57. 9. Cochrane SM, Cunningham SJ, Hunt NP. A comparison of the perception of facial profile by the general public and 3 groups of clinicians. Int J Adult Orthod Orthognath Surg 1999;14:291-5. 10. Altman DG. Practical statistics for medical research. London: Chapman and Hall/CRC; 1991. 11. Altman DG. Personal communication to A. Petrie about repeatability studies. May 15, 2002. 12. Streiner DL, Norman GR. Health measurement scales. A practical guide to their development and use. 2nd ed. Oxford: Oxford University Press; 1995. 13. Brook PH, Shaw WC. The development of an index of orthodontic treatment priority. Eur J Orthod 1989;11:309-20. 14. Prahl-Andersen B, Boersma H, van der Linden FP, Moore AW. Perceptions of dentofacial morphology by laypersons, general dentists and orthodontists. J Am Dent Assoc 1979;98:209-12. 15. Kerr WJ, O’Donnell JM. Panel perception of facial attractiveness. Br J Orthod 1990;17:299-304. 16. Phillips C, Tulloch C, Dann C. Rating of facial attractiveness. Community Dent Oral Epidemiol 1992;20:214-20.
Patient- and clinician-perceived need for orthognathic surgery Karen J. Juggins,a Fiona Nixon,b and Susan J. Cunninghamc London and Oxford, United Kingdom Introduction: To better understand why patients believe they need orthognathic treatment, a study was undertaken to examine perceived need for treatment by patients and clinicians. This questionnaire-based study was undertaken at The John Radcliffe Hospital, Oxford, United Kingdom. Subjects and methods: Forty patients were recruited from combined orthodontic-surgical clinics. They were asked to rate their perceived need for treatment based on facial appearance, dental appearance, function, and overall need. Twenty orthodontists and 20 maxillofacial surgeons were asked to rate perceived need for treatment based on the same parameters, using study models and clinical photographs. Ratings were marked on visual analog scales. Results: Significant differences were found between patients and clinicians in perceived need for treatment based on facial appearance (orthodontists compared with patients, P ⫽ .023; surgeons compared with patients, P ⫽ .001). In addition, maxillofacial surgeons rated a significantly greater overall need for treatment than patients (P ⫽ .027), and they rated treatment need based on facial appearance (P ⫽ .005) and function (P ⬍ .001) significantly higher than orthodontists. Conclusions: Clinicians rated greater need for orthognathic treatment based on facial appearance than did patients. Surgeons also rated greater overall need for treatment than patients. In addition, surgeons rated treatment need based on facial appearance and function significantly higher than orthodontists, but large variations existed in both clinician groups. (Am J Orthod Dentofacial Orthop 2005;128:697-702)
I
t has been shown that facial and dental anomalies that are sufficient to affect a person’s appearance might put that person at a social disadvantage.1,2 When a dentofacial deformity is so severe that an acceptable improvement cannot be obtained by growth modification or orthodontic camouflage, a combination of orthodontic treatment and orthognathic surgery might be the only viable treatment option. As orthognathic treatment has become more widely available and more socially acceptable, the demand for it has increased enormously. What used to be considered a variant of normal appearance is now less acceptable in society. This has led to treatment for relatively moderate “defects” and also to more extensive surgical procedures. The importance of careful assessment to ascertain what the patient expects from the procedure a
Specialist registrar, Department of Orthodontics, The John Radcliffe Hospital, Oxford, and Eastman Dental Institute, University College London, London, United Kingdom. b Consultant, Department of Orthodontics, John Radcliffe Hospital, Oxford, United Kingdom. c Senior lecturer/honorary consultant in orthodontics, Eastman Dental Institute, University College London, London, United Kingdom. Reprint requests to: K. J. Juggins, Department of Orthodontics, John Radcliffe Hospital, Oxford OX3 9DU United Kingdom; e-mail, [email protected]. Presented in June 2004 at the European Orthodontic Congress in Denmark; received the WJB Houston Research Award. Submitted, March 2004; revised and accepted, September 2004. 0889-5406/$30.00 Copyright © 2005 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2004.09.022
cannot be overemphasized. It is easy to imagine that there is a linear relationship between the severity of psychological disturbance and the degree of deformity, but clinical experience has shown that a mild deformity can be harder to cope with than a more severe problem.3,4 Many patients who request orthognathic treatment fall into the milder-deformities group and might be more severely affected psychologically than is apparent at initial consultation.4 Combined orthodontic and surgical treatment aims to produce more harmonious facial skeletal and soft tissue relationships and to improve occlusal function. Research has shown that most patients who request orthognathic treatment do so because of a desire to improve their facial or dental appearance and not because of concerns about occlusal function.5,6 An appropriate plan of combined orthodontic and surgical treatment should be based not only on the clinician’s assessment of esthetics, function, and stability, but also on the patient’s objectives and perceptions of need. In addition, clinicians tend to be more critical than laypersons in their assessments, particularly when rating facial profiles.7-9 Planning orthognathic treatment requires clinical guidelines, or a generally agreed “ideal” for facial proportions. These esthetic standards exist in various forms including artistic views, cephalometric analyses, and anthropometric measurements. It therefore seems 697
698 Juggins, Nixon, and Cunningham
American Journal of Orthodontics and Dentofacial Orthopedics December 2005
Fig 1. Example of visual analog scale.
reasonable to assume that orthodontists and surgeons would share common clinical objectives, but it must be questioned whether this collective judgment of attractiveness is always valid. This study aimed to investigate which aspects of appearance and function patients consider when deciding on orthognathic treatment. This is important to both orthodontists and maxillofacial surgeons, who need to know whether their views of the expected improvement will coincide with those of the patients. The following questions were posed: 1. Does patient-perceived need for orthognathic treatment correlate with clinician-perceived need for treatment? 2. Do orthodontists and maxillofacial surgeons share similar views on need for treatment? 3. Is there consistency within the clinician groups? SUBJECTS AND METHODS
This was a questionnaire-based study undertaken at John Radcliffe Hospital, Oxford, United Kingdom. Ethical approval was obtained, and all patients consented in writing. Forty consecutive new patients (16 men, 24 women) referred by general dental practitioners, fulfilling the following criteria, were recruited from orthognathic clinics: over 16 years of age, about to undergo orthognathic treatment, no history of orthodontic treatment or facial surgery, and no congenital deformities (eg, clefts) or traumatically acquired deformities. Forty senior clinicians (20 orthodontists, 20 maxillofacial surgeons), all experienced in orthognathic treatment, were also recruited. They rated each patient, using study models and standardized orthodontic extraoral and intraoral photographs. No clinicians were actively involved in the care of these patients. The questionnaire required responses to be marked on 10-cm visual analog scales (VAS) that were anchored at
each end, with 0 representing little need for treatment and 10 representing great need for treatment. Patients and clinicians were asked to mark the perceived need for treatment based on: facial appearance, dental appearance, function/anticipated functional problems (clinicians), and overall treatment need (Fig 1). To ensure accurate completion of the VAS and to prevent raters from placing a mark randomly on the same part of the scale for all scales, the anchors for the final question, regarding overall need for treatment, were reversed (0 represented great need for treatment, and 10 represented little need for treatment). To ensure repeatability of the study, both patient and clinician repeatability studies were undertaken. Ten patients were randomly selected to repeat the questionnaire 4 to 6 weeks later but before starting treatment. In addition, 10 clinicians (5 maxillofacial surgeons, 5 orthodontists) were asked to rate 10 patients a second time 2 to 4 weeks later. Statistical analyses
The Bland-Altman method,10 including the paired t test, was used to compare perceived need for treatment between patients and clinicians, and between orthodontists and surgeons. Variation within the clinician groups was studied by using box plots and calculating standard deviations. Patient repeatability was also assessed with the Bland-Altman method and the intraclass correlation coefficient (ICC).11 Clinician repeatability was measured with the ICC within observers.12 This test considers variation within each observer, for each subject, as a proportion of all types of variation. RESULTS
In general, there was good repeatability among patients. The ICC results were all within the acceptable range, with all values above 0.70. However, the Bland-
Juggins, Nixon, and Cunningham 699
American Journal of Orthodontics and Dentofacial Orthopedics Volume 128, Number 6
Table I.
Agreement between patients and orthodontists (values in cm)
Variable
Mean difference
SD
⫺1.00 0.00 0.13 0.68
2.67 3.62 3.24 2.29
Facial appearance Dental appearance Function Overall need for treatment
95% limits of agreement ⫺6.2 ⫺7.1 ⫺6.2 ⫺3.8
to to to to
4.2 7.1 6.5 5.2
P value from t test .023 NS NS NS
NS, Not significant. Table II.
Agreement between patients and maxillofacial surgeons (values in cm)
Variable Facial appearance Dental appearance Function Overall need for treatment
Mean difference
SD
⫺1.41 ⫺0.21 ⫺0.42 0.83
2.53 3.48 3.22 2.28
95% limits of agreement ⫺6.4 ⫺7.0 ⫺6.7 ⫺3.6
to to to to
3.6 6.6 5.9 5.3
P value from t test .001 NS NS .027
NS, Not significant. Table III.
Agreement between orthodontists and maxillofacial surgeons (values in cm)
Facial appearance Dental appearance Function Overall need for treatment
Mean difference
SD
⫺0.40 ⫺0.21 ⫺0.54 0.15
0.85 0.69 0.54 0.75
95% limits of agreement ⫺2.1 ⫺1.6 ⫺1.6 ⫺1.3
to to to to
1.3 1.1 0.5 1.6
P value from t test .005 NS ⬍.001 NS
NS, Not significant.
Altman method highlighted a slight bias in the scoring of dental appearance and overall need for treatment. In general, clinicians exhibited good repeatability when rating perceived need for treatment based on facial appearance, dental appearance, and function. Repeatability for overall need for treatment was below 0.70 but was still considered acceptable at r ⫽ 0.63. Of the 40 patients recruited for this study, 1 had a Class I malocclusion (with an anterior open bite), 22 had Class II malocclusions, and 17 had Class III malocclusions. Each patient was given a score based on the index of orthodontic treatment need13 dental health component. Two patients scored 3 on the index of orthodontic treatment need, 19 scored 4, and 19 scored 5. A comparison was made between the patient score for each variable (facial appearance, dental appearance, function, and overall need for treatment) and the mean scores of the orthodontists and the surgeons for that patient. The mean difference was then calculated by subtracting the mean clinician score from the patient score. The results showed that, when rating facial appearance, there was significant bias, with orthodon-
tists consistently rating a greater need for treatment than the patients (Table I). Similarly, there was significant bias when comparing patient scores and mean scores for the maxillofacial surgeons, with the surgeons rating a greater need for treatment based on facial appearance and overall need for treatment (Table II). A comparison was made between the mean orthodontist score and the mean surgeon score for each variable (Table III). The mean difference was calculated by subtracting the surgeon score from the orthodontist score. The results showed that there was significant bias between maxillofacial surgeon and orthodontist scores, with the maxillofacial surgeons rating greater need for treatment based on facial appearance and function. The negative values for mean difference (⫺0.40 and ⫺0.54) indicate that surgeons scored treatment need higher than orthodontists. Although not statistically significant, the negative score for dental appearance and the positive score for overall need both indicate that surgeons, in general, perceived a greater need for treatment than did orthodontists. Although the values for facial appearance and function were significant, the values for the mean differences
700 Juggins, Nixon, and Cunningham
American Journal of Orthodontics and Dentofacial Orthopedics December 2005
Fig 2. Box plot showing variation among orthodontists rating treatment need based on facial appearance. Box represents range within which 50% of values lie. Black horizontal line represents median. Whiskers mark extremes.
were small compared with those in Tables I and II. The 95% limits of agreement were also relatively narrow compared with Tables I and II. The VAS scores for each clinician, rating each patient, were used to create box plots, which were constructed for each of the 4 VAS for both orthodontists and surgeons. The box plots show the distribution of the observations, within a group (eg, orthodontists or surgeons) for each patient. Figure 2 shows a box plot of the orthodontist group rating need for treatment based on facial appearance. This plot shows large variation; similar plots were seen for all variables in both clinician groups. Standard deviations were calculated and used to measure variation within a group. In general, the standard deviation varied between 1.5 and 2.5 cm, and the range within which the 95% limits of agreement fell was wide. DISCUSSION
Although the ICC scores for patient repeatability were all high (⬎.70), the Bland-Altman method highlighted a slight bias for dental appearance and overall need for treatment. This might be a reflection of the small sample size for the repeatability study. In addition, there could be a tendency for patients to modify their scores after considering the initial rating. How-
ever, the mean differences in the repeatability study were small compared with those in the main study, and this, considered in conjunction with the high ICC values, suggested adequate repeatability. The ICC within observers was used to evaluate the extent to which clinicians obtained similar scores when rating the same patient for a second time. It was expected that, with the same information, clinicians would make a similar decision on treatment need on both occasions. In this study, clinicians were reliable when rating facial appearance, dental appearance, and function, although slightly less so when rating overall need for treatment (ICC ⫽ 0.63). However, this was still deemed to be acceptable and might reflect the relatively small sample size of this repeatability study. Significant differences were found between patients and clinicians (both orthodontists and surgeons) for perceived need for treatment based on facial appearance (orthodontists P ⫽ .023, maxillofacial surgeons P ⫽ .001). The results suggest that clinicians perceive a greater need for alteration of the facial appearance than do patients. However, the 95% limits of agreement were wide (orthodontists compared with patients, ⫺6.2 to 4.2 cm; maxillofacial surgeons compared with patients, ⫺6.4 to 3.6 cm), suggesting large variability. It has been widely reported that clinicians are more critical of facial appearance than laypeople,7-9 and
Juggins, Nixon, and Cunningham 701
American Journal of Orthodontics and Dentofacial Orthopedics Volume 128, Number 6
Table IV.
Measuring agreement between groups scoring perceived need for treatment based on facial appearance
Group Orthodontist and patient Surgeon and patient Orthodontist and surgeon
Mean difference
SD
95% limits of agreement
P value from t test
⫺1.00 ⫺1.41 ⫺0.40
2.67 2.53 0.85
⫺6.2 to 4.2 ⫺6.4 to 3.6 ⫺2.1 to 1.3
.023 .001 .005
significant differences in perceived need for treatment between clinicians and patients were identified in other studies.14,15 In addition, although clinical assessment requires examination in 3 dimensions, quantitative measurement of a dentofacial deformity is still predominantly carried out in the lateral or profile view, with cephalometric analysis playing an important part in assessment, planning of clinical objectives, and measurement of treatment outcome. This could explain, in part, the differences between patient and clinician assessments, with patients placing less weight on the profile view than clinicians. The clinicians used patient photographs to rate perceived need for treatment. Ideally, the study would use patients, rather than records, for all ratings, but this was not logistically possible. There is, of course, the possibility that patients might have rated themselves differently from photographs, but it was thought that their perceived need for treatment based on the “reallife” situation would be more accurate than looking at standard clinical photographs. In addition to differences for facial appearance, there was also significant bias between patient and maxillofacial surgeon scores for overall treatment need (P ⫽ .027). This differs from the orthodontist group where there was no significant difference between the scores. This supports the work of Bell et al,7 who showed that, although both groups recognized need for treatment, the orthodontists as a group perceived significantly less need for orthognathic treatment than did surgeons. There were no other significant perceived-need differences between patients, orthodontists, and maxillofacial surgeons for any other variable. According to the mean scores, orthodontists and patients rated a similar need for treatment based on dental appearance (mean difference in scores ⫽ 0.00 cm). However, if the data are studied more closely, the standard deviation was high (3.62 cm) and the 95% limits of agreement wide (⫺7.1 to 7.1 cm). Improvement of the patient’s facial and dental appearance is the main objective of orthognathic treatment. This study showed that maxillofacial surgeons rated treatment need based on facial appearance signif-
icantly higher than orthodontists (P ⫽ .005); this is consistent with Bell et al,7 who noted similar findings. They reported that, when orthodontists and surgeons evaluated the profile views of patients, the surgeon group perceived significantly greater need for orthognathic treatment than orthodontists. Surgeons also rated significantly higher treatment need based on function (P ⬍ .001). This might be because orthodontists see patients with significant skeletal and dental discrepancies more frequently and do not perceive such malocclusions to be functionally handicapping. There was no significant difference between surgeon and orthodontist ratings in terms of perceived need for treatment based on dental appearance or overall need for treatment, although the surgeons showed a tendency to rate a greater need for treatment than did the orthodontists. Although there were differences between the 2 clinician groups, Table IV shows an example of how much smaller the mean differences, standard deviations, and 95% limits of agreement were for surgeons and orthodontists compared with the clinician-patient comparison. In addition, although these differences were statistically significant, the differences between surgeons and orthodontists as groups were small (0.15-0.40 cm) and perhaps not clinically significant. Ideally, it would be expected that clinicians prescribing orthognathic treatment would share similar views of the need for treatment, with minimal variation expected within the groups. However, this study showed that this was not the case. Although mean clinician scores were used for comparison with patient data, significant variation existed within the groups of clinicians. Box plots were therefore constructed to show the distribution of the observations (ie, variation) within the groups. In general, the standard deviations and 95% ranges were high, suggesting large variation, and this variation between clinicians might be of concern. There appeared to be no relationship between sex, malocclusion type, or index of orthodontic treatment need and the extent of clinician variation. However, relationships might be seen in a larger study. Such large variation could be due to a number of factors. A true variation might exist, the sample size might have
702 Juggins, Nixon, and Cunningham
been inadequate, or the variation within the groups might be due to different levels of experience in treating orthognathic cases. In addition, some clinicians commented that it was difficult to accurately assess function from study models and photographs alone. Phillips et al16 looked at the effects of dental training on assessment of dentofacial attractiveness and found that there were significant differences between the panels of judges (orthodontic residents, dental students, undergraduate students). Such wide variation among clinicians is clearly of concern. The impact on the quality of patient treatment and care was not investigated in this study, but, given the extent of the variation, this issue could certainly be addressed in future studies. However, it is a difficult problem to both investigate and manage. Perhaps, after a similar larger scale study, it will be necessary to consider ways of standardizing the way clinicians rate need for treatment. This study showed that patients’ perceived need for treatment might not always be the same as that of clinicians. Perhaps even more important is the variability in the clinician groups. This has important implications for the treatment planning of orthognathic patients. We must be able to understand these aspects of patient health care to provide and evaluate treatment. CONCLUSIONS
The results of this study suggest that: 1. Orthodontists and maxillofacial surgeons rated a greater need for orthognathic treatment based on facial appearance than patients. 2. Maxillofacial surgeons rated a greater need for orthognathic treatment based on overall treatment need than patients. 3. Large variation existed within both the orthodontist and maxillofacial surgeon groups, but neither group appeared to exhibit more variability than the other. 4. Maxillofacial surgeons rated treatment need based on facial appearance and function significantly higher than orthodontists. 5. There was no significant difference between the clinician groups rating perceived treatment need
American Journal of Orthodontics and Dentofacial Orthopedics December 2005
based on dental appearance or overall need for treatment, but surgeons generally tended to rate a greater need than orthodontists. We thank Dr Aviva Petrie for her statistical advice; Mary McKnight, Steve Watt-Smith, and Heather Nevard for their contributions to this study; and the clinicians and patients who participated. REFERENCES 1. Shaw WC, Meek SC, Jones DS. Nicknames, teasing, harassment and the salience of dental features among school children. Br J Orthod 1980;7:75-80. 2. Heldt L, Haffke EA, Davis LF. The psychological and social aspects of orthognathic treatment. Am J Orthod 1982;82:318-28. 3. Macgregor FC. Social and psychological implications of dentofacial disfigurement. Angle Orthod 1970;40:231-3. 4. Lansdown R, Lloyd J, Hunter J. Facial deformity in childhood: severity and psychological adjustment. Child Care Health Dev 1991;17:165-71. 5. Laufer D, Glick D, Gutman D, Sharon A. Patient motivation and response to surgical correction of prognathism. J Oral Surg 1976;41:309-13. 6. Kiyak HA, Hohl T, Sherrick P, West RA, McNeill RW, Bucher F. Sex differences in motives for and outcomes of orthognathic surgery. J Oral Surg 1981;39:757-64. 7. Bell R, Kiyak HA, Joondeph DR, McNeill RW, Wallen TR. Perceptions of facial profile and their influence on the decision to undergo orthognathic surgery. Am J Orthod 1985;88:323-32. 8. Lines PA, Lines RR, Lines CA. Profilemetrics and facial esthetics. Am J Orthod 1978;73:648-57. 9. Cochrane SM, Cunningham SJ, Hunt NP. A comparison of the perception of facial profile by the general public and 3 groups of clinicians. Int J Adult Orthod Orthognath Surg 1999;14:291-5. 10. Altman DG. Practical statistics for medical research. London: Chapman and Hall/CRC; 1991. 11. Altman DG. Personal communication to A. Petrie about repeatability studies. May 15, 2002. 12. Streiner DL, Norman GR. Health measurement scales. A practical guide to their development and use. 2nd ed. Oxford: Oxford University Press; 1995. 13. Brook PH, Shaw WC. The development of an index of orthodontic treatment priority. Eur J Orthod 1989;11:309-20. 14. Prahl-Andersen B, Boersma H, van der Linden FP, Moore AW. Perceptions of dentofacial morphology by laypersons, general dentists and orthodontists. J Am Dent Assoc 1979;98:209-12. 15. Kerr WJ, O’Donnell JM. Panel perception of facial attractiveness. Br J Orthod 1990;17:299-304. 16. Phillips C, Tulloch C, Dann C. Rating of facial attractiveness. Community Dent Oral Epidemiol 1992;20:214-20.