The influence of using 2D cephalometry on orthodontic treatment outcome

i ii iii “The influence of using 2D cephalometry on orthodontic treatment outcome” Thesis submitted in partial fulfilment of the requirements for the degree of “DOCTOR IN MEDICAL SCIENCES” presented to the Faculty of Dental Medicine of the University of Porto. Ana Paula Oliveira dos Reis Durão Porto, 2014 iv v “The influence of using 2D cephalometry on orthodontic treatment outcome.” Ana Paula Oliveira dos Reis Durão Promotor: Prof. Dr. Reinhilde Jacobs Full Professor, Head of the Oral Imaging Centre, Oral Imaging Center, OMFSIMPATH research group, Department Imaging & Pathology, Faculty of Medicine, University of Leuven, Belgium. Co-promotor: Prof. Dr. Afonso Pinhão Ferreira Full Professor, Orthodontics Department, Director of the Faculty of Dental Medicine of the University of Porto, Portugal. vi vii Faculty of Dental Medicine of the University of Porto Members of the Scientific Committee Prof. Doutor Afonso Manuel Pinhão Ferreira (Full Professor) – President Prof. Doutor Américo dos Santos Afonso (Associate Professor with aggregation) Prof. Doutor António Cabral Campos Felino (Full Professor) – Vice-President Prof. Doutor César Fernando Coelho Leal Silva (Associate Professor with aggregation) Prof. Doutor Germano Neves Pinto Rocha (Associate Professor) Prof. Doutora Irene Graça Azevedo Pina Vaz (Associate Professor) Prof. Doutora Inês Alexandra Costa Morais Caldas (Assistant Professor) Prof. Doutor João Carlos Antunes Sampaio Fernandes (Full Professor) Prof. Doutor João Carlos Gonçalves Ferreira de Pinho (Associate Professor with aggregation) Prof. Doutor João Fernando Costa Carvalho (Full Professor) Prof. Doutor Jorge Manuel Carvalho Dias Lopes (Full Professor) Prof. Doutor José António Macedo Carvalho Capelas (Associate Professor with aggregation) Prof. Doutor José Carlos Reis Campos (Assistant Professor with aggregation) Prof. Doutor José Mário Castro Rocha (Assistant Professor) Prof. Douto Manuel José Fontes de Carvalho (Associate Professor) Prof. Doutora Maria Cristina P. C. M. Figueiredo Pollmann (Associate Professor) Prof. Doutora Maria Helena Guimarães Figueiral da Silva (Full Professor) Prof. Doutora Maria Helena Raposo Fernandes (Full Professor) Prof. Doutora Maria Lurdes Ferreira Lobo Pereira (Assistant Professor) Prof. Doutor Mário Augusto Pires Vaz (Associate Professor - invited) Prof. Doutor Mário Jorge Rebolho Fernandes Silva (Full Professor) Prof. Doutor Mário Ramalho Vasconcelos (Associate Professor with aggregation) Prof. Doutor Miguel Fernando Silva Gonçalves Pinto (Full Professor) Prof. Doutor Paulo Rui Galrão Ribeiro Melo (Associate Professor with aggregation) Prof. Doutor Ricardo Manuel Lobo Faria Almeida (Associate Professor with aggregation) viii Emeritus Professors Prof. Doutor Adão Fernando Pereira (Full Professor) Prof. Doutor Amílcar Almeida Oliveira (Associate Professor) Prof. Doutor António Manuel Machado Capelas (Associate Professor✝) Dr. António Ulisses Matos dos Santos (Invited Assistant) Prof. Doutor Durval Manuel Belo Moreira (Associate Professor with aggregation) Prof. Doutor Francisco António Rebelo Morais Caldas (Full Professor) Dr. José Maria Vaz Osório (Invited Assistant) Prof. Doutor José Serra Silva Campos Neves (Full Professor) Prof. Doutor Manuel Desport Marques (Invited Associate Professor✝) Prof. Doutor Manuel Guedes de Figueiredo (Associate Professor) Retired Professors Prof. Doutor António Manuel Guerra Capelas (Assistant Professor) Prof. Dr. Artur Manuel Osório de Araújo (Invited Associate Professor) Prof. Doutor Fernando Jorge Morais Branco (Full Professor) Prof. Doutor Fernando José Brandão Martins Peres (Full Professor ✝) Prof. Doutor José Albertino Cruz Lordelo (Associate Professor with aggregation) Prof. Doutor José Carlos Pina Almeida Rebelo (Full Professor) Prof. Doutor Manuel Pedro da Fonseca Paulo (Full Professor) Prof. Doutora Maria Adelaide Macedo Carvalho Capelas (Associate Professor ✝) Prof. Doutora Maria Purificação Valenzuela Sampaio Tavares (Full Professor) Prof. Doutor Rogério Serapião Martins Aguiar Branco (Full Professor) ix Dedications: To Hugo, Pedro and Miguel. To my mother and father. To the memory of my grandparents. xvi 1.7. Synthesis of evidence .................................................................................. 46 1.8. Results ......................................................................................................... 46 1.8.1. Role of cephalometry on the orthodontic treatment planning ........... 49 1.8.2. Cephalometric measurements and landmark identification .............. 55 1.8.3. Cephalometric analysis ..................................................................... 60 1.9. Discussion ................................................................................................... 62 1.10. Conclusions ................................................................................................. 66 Chapter 2. Accuracy and reliability of 2D cephalometric analysis in orthodontics as compared to the gold standard measurement on skull ......... 67 2.1. Introduction ................................................................................................. 69 2.2. Materials and methods ................................................................................ 71 2.3. Results ......................................................................................................... 75 2.4. Discussion ................................................................................................... 79 2.5. Conclusions ................................................................................................. 84 Chapter 3. Reproducibility of 2D cephalometric landmark identification by orthodontists and dentomaxillofacial radiologists ............................................ 85 3.1. Introduction ................................................................................................. 87 3.2. Materials and methods ................................................................................. 88 3.3. Results .......................................................................................................... 93 3.4. Discussion ................................................................................................. 102 3.5. Conclusions ............................................................................................... 106 Chapter 4. Variations in Sella landmark identification and its effect in angles SNA and SNB in lateral cephalometric radiographs ..................................... 107 4.1. Introduction ............................................................................................... 109 4.2. Materials and methods .............................................................................. 112 4.3. Results ....................................................................................................... 114 4.4. Discussion ................................................................................................. 119 4.5. Conclusions ............................................................................................... 121 xvii Chapter 5. Influence of using lateral cephalometric radiography in orthodontic diagnosis and treatment planning .............................................. 123 5.1. Introduction ................................................................................................ 125 5.2. Materials and methods ............................................................................... 126 5.3. Results ....................................................................................................... 129 5.4. Discussion .................................................................................................. 133 5.5. Conclusions ............................................................................................... 136 General discussion and conclusions ................................................................. 137 Summary ............................................................................................................ 143 Resumo ............................................................................................................... 145 References .......................................................................................................... 147 Appendices ......................................................................................................... 159 Appendix 1. Letter written to the Ethics Committee of the FMDUP ................. 161 Appendix 2. Letter of Approval from the Ethical Committee of FMDUP .......... 165 Appendix 3. Questionnaire ................................................................................. 169 xviii xix LIST OF ABREVIATIONS 2D Two-dimensional 3D Three-dimensional LCR Lateral cephalometric radiograph DMFR Dentomaxillofacial radiologists mAs Product of tube current (mA) and exposure time (s) mGy milligray kV kilovoltage m meters mm millimetres PA postero-anterior CCD Charged Couple Device ALARA As low as reasonably achievable N Nasion Or Orbital S Sella Po Porion Co Condylion Go Gonion Me Menton Pog Pogonion Gn Gnathion B B point xx A A point ANS Anterior Nasal Spine PNS Posterior Nasal Spine LIA Lower incisor apex LIB Lower incisor border UIB Upper incisor border UIA Upper incisor apex xxi LIST OF FIGURES Figure 1 – Drawing of a human’s head with measurements, by Leonardo da Vinci 1488-9. Figure 2 – Representation of measurements of two human heads, by Albrecht Dürer. Figure 3 – The picture shows a superior and a lateral view of two skulls, one is brachycephalic (B) and the other is dolicocephalic (A). Figure 4 – Method of determining the facial angle by Petrus Camper. Figure 5 – Illustration of the Frankfort Plane on a skull, on a patient and on a lateral cephalogram. Figure 6 – Illustration of the SN Plane on a lateral cephalogram. Figure 7 – Cephalostat used by Hofrath in Germany. Figure 8 – The Broadbent-Bolton cephalostat. Figure 9 – Two X–ray sources were positioned at mutually perpendicular locations, the patient head position did not have to be moved or changed to take a lateral view or a postero-anterior view. Figure 10 – Images performed without (A) and with (B) a triangular shapedcollimation. Figure 11 – Patient in cephalostat and radiography with anatomically shaped cranial collimator (ACC) attached, proposed by Hoogeveen et al., 2014. Figure 12 – Flow chart showing clinical decision making, regarding lateral cephalograms. Figure 13 – Cephalometric tracing of a lateral cephalometric radiography showing the main cephalometric landmarks. xxii Figure 1.1 – Methodology followed in the article selection process (adapted from: Moher et al., 2009). Figure 2.1 – Cephalometric landmarks used in the study. N – Nasion; Me – Menton; ANS – Anterior Nasal Spine; Co – Condylion; Gn – Gnathion; A – Point A; B - Point B; Pog – Pogonion; Po – Porion; Or – Orbitale; Go – Gonion. Figure 3.1 – Cephalometric landmarks used in the study. N – Nasion; Or – Orbitale; S – Sella; Co – Condylion; Po – Porion; PNSPosterior Nasal Spine; ANS – Anterior Nasal Spine; A – Point A; UIA – Upper incisor apex, UIB – Upper incisor border; LIB – lower incisor border; LIA – Lower incisor apex; B - Point B; Pog – Pogonion; Gn –Gnathion; Me – Menton; Go – Gonion. Figure 3.2 – Example of a lateral cephalometric radiography with identification of landmarks by two observers. Figure 4.1 – Lateral cephalometric radiograph showing the identified landmarks and the measured angles. Figure 5.1 – Example of the information given to the orthodontists. xxiii LIST OF TABLES Table 1.1. Protocol 1, Selection for inclusion of publications. Table 1.2. Protocol 2, based on the QUADAS-2 tool for evaluation of methodology of included studies. Table 1.3. Publications related to the importance and contribution of cephalometry on the orthodontic treatment planning. Table 1.4. Publications concerning landmark identification. Table 1.5. Publications on cephalometric analysis. Table 2.1. Linear measurements evaluated on human skulls and lateral cephalometric radiographs in this study. Table 2.2. Mean differences between the first and second observations with regards to intra-observer agreement (mm). ( 1 standard deviation; 2 intraclass correlation coefficient; 3 confidence interval; 4 limits of agreement). Table 2.3. Inter-observer agreement ( 1 standard deviation; 2 intraclass correlation coefficient; 3 confidence interval; 4 limits of agreement). Table 2.4. Mean of differences and level of agreement between measurements performed on the skull and radiography. Table 3.1. ICC for the interand intra-observer evaluation. Table 3.2. Minimum and maximum euclidean distances for orthodontists, defined as absolute differences in millimetres between the mean values and standard deviations of each landmark and the averaged for all observers (mm). Table 3.3. Minimum and maximum euclidean distances for dentomaxillofacial radiologists, defined as absolute differences in millimetres between the mean values and standard deviations of each landmark and the averaged for all observers (mm). xxiv Table 3.4. Standard deviation for each linear and angular measurement identified by two observers on 20 radiographs. Table 4.1. Intraand inter-observer differences in S landmark identification (mm). Table 4.2. Intra and inter-observer differences in SNA and SNB angles (in °). Table 4.3. Correlation between point S identification of and its effect on SNA and SNB angles. Table 4.4. Number of cases (n) in which diagnosis was changed, regarding the SNA and SNB standard values, according to each observer. Minimum and maximum degree variations are indicated. Table 5.1. Mean percentage of agreement between the first and second sessions for all observers. Table 5.2. Mean differences in the proposed treatment plan duration (months) between the two sessions. Table 5.3. Number of additional information required for each observer in the 1 st and 2 nd sessions. 1 INTRODUCTION 8 also diverge, in the same patient within time. Natural head position (NHP) can also be used as a reference. It provides an extracranial reference line, defined as a physiologic position and it is relatively constant over time. The concept of NHP was introduced in orthodontics in the 1950s by Downs (1956), Bjerin (1957), and Moorrees and Kean (1958). NHP has been found to be highly reproducible in adults and children, males and females, Caucasians and non-Caucasians, with a variance of only about 4°. Some authors believe that the analysis based on NHP should have a greater clinical application than traditional methods in describing morphology (Bansal et al., 2012). Figure 6. Illustration of the SN Plane on a lateral cephalometric radiography. 9 Cephalometric Radiography Pacini immobilized the patient’s head with bandages or gauze, taking radiographs with the sagittal plane parallel to the radiographic film. The equipment had an arm with a distance of two meters between the X-ray source and the film (Wahl, 2006; Athanasios and Athanasiou, 1995; Moyers, 1988). He used craniometric points available for anthropology studies and evaluated the development and deviations of the normality in structures of the skull. In 1922, he was the first to use skull radiographs for craniometrical measurements, and demonstrated that cranioskeletal measurements could be made from skull radiographs more easily than from the skull itself. Hofrath (Figure 7) in Germany used a cephalostat of Korkhaus. He described in detail its radiographic technique and cephalometric analysis which was published in Germany in 1931. Figure 7. Cephalostat used by Hofrath in Germany. 10 Broadbent-Bolton cephalostat Broadbent designed a head-holder, of excellent accuracy. The basic principles of this cephalostat are still in use today (Athanasios and Athanasiou, 1995). The cephalostat design was named Broadbent-Bolton (Figure 8), due to the financial support that he was given by the Bolton foundation. It was first used in children. This cephalostat used two X-ray sources separately and two film receptors, to take one posterior-anterior (PA) radiograph and one lateral radiograph. By using two X-ray sources in different locations, the patient head position did not have to be moved or changed between the two exposures (Athanasios and Athanasiou, 1995; Moyers, 1988) (Figure 9). From this moment, the method of performing measurements from radiographs of the skull, as a scientific assessment for orthodontic problems, has become possible. The serial x-rays, which previously were taken with imprecise cephalostat and therefore of questionable value, were modified after the Broadbent invention. These radiographs are now routinely used in the observation of skull growth and in the evaluation of orthodontic treatment. Figure 8. The Broadbent-Bolton cephalostat (Athanasios and Athanasiou, 1995). 11 Figure 9. Two X–ray sources were positioned at mutually perpendicular locations, the patient head position did not have to be moved or changed to take a lateral view or a posterior-anterior view (Raju et al., 2010). Teleradiography In 1940, Higley presented a cephalostat with only one x-ray source and a movable head fixer. In the same year, Margolis developed a cephalostat with the same principles but with less distortion, which solved some problems of the existent technique. At the First Congress of Cephalometric Radiography in 1957, the teleradiography technique was standardized and a distance of 1.524 meters from the focal spot to the plane of the image receptor was determined to be the standard, as well as the positioning of the left (as opposed to the right) side of the patient’s head near to the image receptor. The distance of the head to the image receptor is standardized being of 20 cm from the sagittal plane of the patient to the image receptor. 12 By having a relatively large distance between the X-ray source and the head, it helps to minimize magnification errors (Sánchez and Filho, 2009). The introduction of the head positioning device and the technique of radiographic cephalometry were pioneered by Broadbent in the United States and by Hofrath in Germany in 1931, simultaneously but independently (AlBarakati. et al., 2012; Devereux. et al., 2011; Nijkamp. et al., 2008). Until 1931, diagnosis was performed with clinical examination. After 1931, possibilities emerged for orthodontists, with LCR providing invaluable help in treatment planning, analysis of growth, mid-treatment monitoring and prediction of possible treatment outcomes. The main clinical indications of this radiographic technique can be considered in two major areas: orthodontics and orthognatic surgery (Whaites, 2007). Lateral cephalometry and orthodontics Since the introduction of lateral cephalometric radiograph (LCR) (also denoted as “lateral cephalogram”, “lateral cephalometry” or “lateral teleradiograph”) in 1931, this radiograph and its related analysis has become a standard tool in orthodontic assessment and treatment planning (AlBarakati et al., 2012; Devereux et al., 2011; Nijkamp et al., 2008). Lateral cephalogram is different from a lateral skull view by the standardized projection geometry using a cephalostat, to enable standardized measurements of jaw bones, teeth and skeletal relationships. Apart from lateral cephalometry, posterior-anterior (PA) projections can also be carried out using standardized projection geometry, particularly when skull asymmetry does apply. However, the indication for these PA cephalograms is far below that 13 of the lateral ones, and so not very much used in Orthodontics. The present review, will therefore only focus on lateral cephalograms. Indeed, nowadays, orthodontic treatment is performed in many children in Europe, with many of them receiving a lateral cephalogram during the initial diagnostic phase and many also later on, at the end of the treatment period. Notwithstanding the fact that it is widely used, the real value of lateral cephalometry for the diagnosis and planning of the orthodontic treatment remains uncertain (Bourriau et al., 2012; Devereux et al., 2011; Nijkamp et al., 2008, Pae et al., 2001; Bruks et al., 1999; Atchison et al., 1991). Some authors stated that in many instances an adequate orthodontic diagnosis and treatment plan cannot be done without comparing cephalograms before and after orthodontic treatment. For that reason a lateral cephalogram is needed. They reinforced by stating that to treat skeletal malocclusions without a cephalometric radiograph is a serious error (Graber and Vanarsdall, 1994). However, Atchison et al. in 1991, reported that many radiographic techniques used in orthodontics are often not useful or are ineffective. According to Atchison et al., approximately three quarters of the radiographs exposed for orthodontic treatment purposes did not provide unexpected information which might lead to a change in the orthodontic diagnosis or treatment planning. In 1992, the same authors stated that the decision of taking a cephalogram prior to orthodontic treatment may be influenced by several factors, such as the suspicion by the clinician of underlying disease or medico legal reasons. According to the European Commission guidelines on radiation protection in dental radiology in 2004, only a small percentage of diagnosis and treatment plan changed after evaluating radiographs, alternating from 16% to 37% and 4% to 20% respectively (European Commission, 2004). 14 Dose reduction in lateral cephalometric radiography The International Commission on Radiological Protection (ICRP) recommends that any practice involving ionizing radiation, or irradiation of patients with ionizing radiation, should be justified in relation to other diagnostic methods and produces a positive benefit to the patient (ICRP, 2007). The benefit should overcome any possible risk of damage that may occur associated with the use of ionizing radiation, taking into account social and economic factors, among others. The appropriate justification and imaging technique selection is also crucial in orthodontics. That is due to the fact that the patients are usually children and because the treatment period is usually 18 months or more. Radiographs are often taken at different time intervals during treatment, and young children are more vulnerable to radiation exposure (Tsuji et al., 2006). Therefore, it is a basic premise of radiological practice that patient exposure should be kept “As Low As Reasonably Achievable” (ALARA principle), while at the same time producing images of sufficient diagnostic quality. Dose reduction in lateral cephalometric radiography may be achieved by several means, which include:  reduction of the field by collimating the beam to shield the thyroid gland and/or the brain tissue;  use of collar shielding for the thyroid gland;  using a more sensitive detector than conventional film, such as a photostimulable phosphor plate or a direct-digital scanning system;  remove the anti-scatter grid;  introduction of the air-gap technique;  lowering the mAs yields the lowest effective dose and is therefore preferred; 15  use of distance: For a point source of radiation, the dose rate falls off as the inverse of the square of the distance from the source. A true teleradiographic cephalostat would introduce a gap of 4 meters between the head of the patient and the x-ray source, because the radiation dose is reduced exponentially with increased distance. An equivalent dose of 1 intra-oral radiograph can be reached at 4 meters (ICRP, 2007; Tsiji et al., 2006; European Commission, 2004; Gijbels et al.,2003). Kaeppler et al., 2007, referred that the most frequently used kilovoltage is of 70kV. The use of a digital imaging receptor (phosphor-stimulated computed plates) can also substantially reduce radiation exposure, when compared to conventional film radiography (Chen et al., 2004; Lim and Foong, 1997; Seki and Okano, 1993). At some important organs of the head and neck region, the absorbed dose from conventional radiography was approximately 2-fold higher than for the digital radiography. On the side of the head closer to the tube, Visser et al. in 2001, measured 81 mGy versus 34 mGy at the level of the lens of the eye, 103 mGy versus 45 mGy at the parotid gland, 53 mGy versus 34 mGy at the level of the submandibular gland, and 3 mGy versus 2 mGy at the level of the thyroid gland. The absorbed dose was about 9 times less on the side of the head nearer to the film than on the other side. Digital image receptors can be classed as indirect (using phosphorstimulated computed plates) and direct (using charged couple device-CCD), according to whether the receptor is physically linked to the computer and is capable of converting the ionising radiation into electrical signals directly. CCD sensors are relatively small. Large ones, as large as a patient’s head would be very expensive and difficult to make. Therefore, in the direct digital (CCD) 16 imaging method the head is “scanned” rather than imaged using a one-shot approach (Gijbels et al., 2001). By contrast, the one shot approach can be applied with the indirect digital (phosphor plate) technique, giving it the advantages of reducing exposure time and therefore also minimising movement artefact (Chen et al., 2004). Beam Collimation Beam collimation is recommended by the European guidelines on radiation protection in dental radiology in order to restrict the irradiated field to the minimum area required for diagnosis. In the past, this was performed with true teleradiographic machines, but cephalometric arms in modern multimodal units often have little collimation potential. Wedge collimation is possible but not available on any kind of digital cephalometric equipments. Gijbels et al., 2003, suggested that the use of a wedge-shaped collimator mounted on the X-ray tube could reduce the dose to more than 40%. Tsuji et al. in 2006, suggested a triangleshaped collimation to reduce the effective dose to the thyroid gland and also avoid scatter radiation (Figure 10). Later, in 2012, Lee et al, advocated a dose reduction of approximately 60%. Radiation protection is especially important for children. Some authors, state that since the brain and thyroid receive high radiation doses, wedge-shaped collimation should be considered (Gijbels et al., 2001). Radiation hazards from cephalometry examinations have been reported since the fifties (Tyndall et al., 1988). In orthodontics the area of interest is the facial skeleton, which is situated below the level of the base of the skull 17 (European Commission, 2004). Imaging structures superior to the superior orbital rim, posterior to the occipital condyles, and inferior to the hyoid bone are clinically unnecessary (Mupparapu, 2005). However, some authors believe that beam collimators do not ensure complete protection and also involve a major change with high costs in cephalometric equipment (Sansare et al., 2011). Moreover in some machines this modification is not possible. Besides the known advantages of using beam collimation, its use in orthodontics is not a current practice. Hoogeveen et al., in 2014, suggested two reasons for that, one is due to anatomical variability of the area below the mandible and the fact that the use of wedge collimation covers the cervical vertebrae, disabling the determination of bone maturation. Another reason is because these collimators were not designed for today’s combination panoramic–cephalometric imaging systems. Figure 10. Images performed without (A) and with (B) a triangular shaped-collimation (Tsuji et al., 2006). A B 24 Technique and equipment Cephalometry produces standardized images of the entire head and a portion of the cervical spine. It is used to identify skeletal and dental landmarks for orthodontic and craniofacial analysis (Chien et al., 2009). It is a standardized and reproducible lateral skull radiograph used to assess the relationship of teeth to the jaws and the jaws to the facial skeleton (Whaites, 2007). The fact that this is a standardized technique is of extreme importance. It is sometimes necessary to perform these radiographs at different periods of time during the orthodontic treatment. A comparison is possible by superimposing the cephalometry tracings. This technique requires three components: 1) a fixed X-ray point source, 2) a cephalostat where the patient’s head is fixed at three points (external auditory meatus bilaterally and bridge of the nose), and 3) an image receptor (Athanasios and Athanasiou, 1995; Graber and Vanarsdall, 1994). The sagittal plane of the patient should be perpendicular to the central ray of the beam and parallel to the plane of the image receptor. The Frankfort plane should be horizontal. The patient is positioned with one side toward the image receptor, conventionally it is the left side which should be nearest to the image receptor. Patient should bite in centric occlusion position and the lips should be relaxed (Albarakati et al., 2012; Athanasios and Athanasiou, 1995; Moyers, 1988). Exact superimposition of the right and left sides is impossible due to magnification of the structures further away from image receptor and the slightly lesser magnification of the structures nearer to the image receptor. Structures close to the midsagittal plane should be nearly exactly superimposed. Bilateral structures near to the midsagittal plane show less discrepancy in size compared with bilateral structures further away from the midsagittal plane (Bourriau et al., 25 2012; Duarte et al., 2009; White and Paroah, 2009; Whaites, 2007; European Commission, 2004; National Radiological Protection Board, 2001; Ahlqvist et al., 1986). When lateral cephalometry was used for the very first time, the distance from the x-ray source to the film was much greater, and conversely the magnification was smaller being of 3% at a distance of 5 meters, 3.5% at a distance of 4 meters and 11.5% at a distance of 1.5 meters (Bourriau et al., 2012; Ahlqvist et al., 1986). Nowadays, only the equipments with a focus-to-film distance of 1.5 to 1.8 meters are in use. Although to minimise the magnification effects, the focus-to-film distance should be greater than 1 meter and ideally within the range 1.5 to 1.8 meters. There is always a minimal enlargement that still creates discrepancies between left/middle/right sides of the skull. The equipment should provide a perfect alignment between patient, X-ray source and image receptor, to reduce errors on the radiography. A light beam diaphragm, or other suitable means, should be used to help collimate the x-ray beam to include only the area that would be used for orthodontic proposes (National Radiological Protection Board, 2001). Visualisation of the soft tissue profile is necessary, therefore, an aluminium wedge filter should be provided at the anterior part of the x-ray tube head between patient and the X-ray tube, to absorb some radiation (White and Paroah, 2009; Whaites, 2007). The aluminium wedge filter attenuates the X-ray beam in the region of the facial soft tissues (Whaites, 2007). In the beginning of cephalometry, two images were taken at different kilovoltages, first one to visualize soft tissue the second one to visualize hard tissues. Nowadays a soft tissue filter is used to overcome this double irradiation to the patient. 26 Analysis of the cephalograms After obtaining a good quality lateral cephalogram, it is possible to perform a cephalometric analysis, which allows angle and linear measurements to be made, including: • the outline and inclination of the anterior teeth; • the positional relationship of the mandibular and maxillary dental bases to the cranial base; • the positional relationship between maxillary and mandibular dental bases; • the relationship between the bones of the skull and the soft tissue profile of the face (Bourriau et al., 2012; Deveraux et al., 2011; Sánchez and Filho, 2009; Arpoen et al., 2008; Whaites, 2007; McIntyre and Mossey, 2003). In 1951, Downs published the first article on cephalometric analysis. Until recently, cephalometric analysis could only be done manually and laboriously. A sheet of tracing paper or transparent acetate was placed directly over the radiograph on top of a lightbox, and the anatomical landmarks are identified using pencil or pen onto the paper or acetate, producing the “orthodontic tracing”. After this step, the various angles and all the measurements and other calculations are performed manually from the tracing. Nowadays, there are numerous computer software programmes available that allow a faster identification of the anatomical landmarks, calculating the data and indicating the most suitable treatment plan. The software requires a digital image, which may be digitally acquired radiographic image or obtained after digitizing a conventional film radiograph on an optical scanner (Lim and Foong, 1997). There are many analyses available and the choice may be based on clinician's preference or patients’ conditions. Some authors compared the accuracy of digital cephalometric measurements with the 27 hand-tracing method (Bruntz et al., 2006; Santoro et al.,2006; Chen et al., 2004). Computerized cephalometric measurement using direct digital imaging is better than digitized conventional radiographs. However the principle of the digital cephalometric analysis is the same. The observer needs to identify each landmark. All the values are then compared with reference values. In 1982, De Abreu found a lack of agreement in the four cephalometric analyses he studied. Despite his observation, few authors have afterwards investigated the importance and usefulness of the different existing landmarks (Chen et al., 2004). Definitions of anatomical landmarks used in 2D lateral cephalometry Anatomical points or landmarks identified on lateral cephalometric radiographs to allow precise linear and angular measurements. The points are recorder either on an overlying sheet of paper or acetate or digitally. The definition of the main cephalometric landmarks is listed below (Figure 13): • Porion (Po): Most superior point of left external auditory meatus. • Sella (S): Geometric centre of the sella turcica. • Orbitale (Or): Most inferior point of the infraorbital margin. • Nasion (N): Most anterior point on frontonasal suture. • Basion (B): Lowest point on anterior rim of foramen magnum. • Pogonion (Pog): Most anterior midpoint of the bony chin. • Gnathion (Gn): Most anterior and inferior point on the bony outline of the chin, situated equidistant from pogonion and menton. • Menton (Me): Lowest point on the bony outline of the mandibular symphysis. 28 • Gonion (Go): Point on curvature of the angle of the mandible located by bisecting the angle formed by lines tangent to the posterior ramus and the inferior border of the mandible. • Anterior Nasal Spine (ANS): The tip of the anterior nasal spine. • Posterior Nasal Spine (PNS): The tip of the posterior spine of the palatine bone in the hard palate. • Point A (A): Deepest midline point between the anterior nasal spine and prosthion. • Prostion (Pr): Most anterior point of the alveolar crest in the premaxilla, usually between the upper central incisors. • Point B (B): Deepest point in the bony outline between the infradental and the Pogonion. • Infradental (Id): Most anterior point of the alveolar crest, situated below the lower central incisors. 29 Figure 13. Cephalometric tracing of a lateral cephalometric radiography showing the main cephalometric landmarks. Accuracy of cephalometric measurements The accuracy of cephalometric measurements is of great interest. Many studies have been published on the errors associated with landmark identification, errors arising from the registration of landmarks, and errors due to measurement procedures (Chen et al., 2004). Errors due to the projection of a three-dimensional object on a two-dimensional film have been studied less extensively (Albarakati et al., 2012; Bruks et al., 1999; Ahlqvist et al., 1986). Few studies, however, have attempted to assess the accuracy of cephalometric measurements as applied threedimensionally (3D) because of known intrinsic limitations of these images, such as distortion and magnification. Lateral cephalograms have intrinsic limitations that result in distorted images, enlarged in some areas and reduced in others. 30 When doing the tracing, precise landmark identification is important for the diagnosis and treatment plan (Sánchez and Filho, 2009). A trained person should do the tracing, it can be done by orthodontists or dentomaxillofacial radiologists. Measurements based on cephalometry may involve errors, which are classified by Baumrind and Frantz as “errors of projection” and “errors of identification” (Baumrind and Frantz, 1971). • Projection errors Projection errors result from imaging 3D structures in a two dimensional (2D) radiographic image. Projection magnification of objects is the result of varying the distance between individual structures and the film or imaging receptor, resulting in variable enlargement of some structures depending on proximity to the image receptor. The positioning of the patient’s head is also of extreme importance, since a slight rotation of the head may lead to distortion and errors in linear and angulation measurements. Ahlqvist et al. (1986), reported that a +/- 5º of head rotation from the ideal position resulted in an insignificant error, however if the head rotation increased the probability of an error occurring was greater and may become significant even at rotations of a few degrees more than +/-5º. • Identification errors Errors of identification are those that can occur in the landmark identification process, such as the porion, condylion, orbitale, basion, gonion, anterior and posterior nasal spine, and lower incisor apex. Adenwalla et al. in 1988, studied the reliability of the Po and Co identification on lateral cephalogram, and concluded 31 that these two anatomical landmarks could not be accurately located on lateral cephalograms taken with the patient in the mouth closed position. Therefore, they suggested an open-mouth cephalogram should be taken and superimposed on the respective cephalogram in the centric occlusion position to obtain the most accurate and reliable measurements. The main problem with these two landmarks is that the ear rods are superimposed on the patient skull region of interest. These errors are due to overlapping structures that are superimposed on landmarks of interest, as well as the resolution and quality of the acquired images. Inherent cephalometric errors can lead to variations in orthodontic and surgical treatment planning (Chien et al., 2009). The errors in cephalometric analysis are composed of systematic errors and random errors. The latter involves tracing, landmark identification, and measurements errors (Chen et al., 2004). Previously, landmark identification and measurements were done by tracing outlines on the radiograph and measuring by hand. Nowadays, many cephalometric analysis software programmes are available and only landmark identification has to be done by hand whilst the analysis is done automatically. This means that identification errors may still occur. Computer-aided cephalometric analysis can totally eliminate the mechanical errors in drawing lines between landmarks and in measurements with a protractor, although it does not introduce more measurement errors than hand tracing, as long as the landmarks are identified manually (Chen et al., 2004). Digitally acquired cephalometric imaging presents numerous advantages, as the possibility of enhancement imaging techniques that allow improved landmark identification, faster cephalometric data acquisition and analysis, more efficient storage and archiving and easier transfer of the image to distant sites. 32 Recently, automatic cephalometric landmark identification is possible using cephalometric software can be used directly on a digitally acquired image or after digitizing a conventional film with a scanner or a digital camera (AlBarakati et al., 2012). For this modality the mean success rate for identifying landmark positions was 88% with a range of 77% to 100% (Tanikawa et al., 2009). 33 OBJECTIVES AND HYPOTHESIS The overall aim of this thesis was to validate the accuracy and reliability of 2D cephalometric radiograph in orthodontic diagnosis and treatment planning. The outcome of this study is mandatory to further judge any potential and additional role of 3D cephalometric analysis. The various chapters and topics address the following hypotheses: 1. 2D cephalometrics suffers a poor accuracy when compared to real skull analysis (Chapter 2). 2. 2D cephalometrics has a poor intraand inter-observer variability, thus influencing planning and treatment decisions (Chapter 3). 3. Landmark identification on the point Sella as a reduced variability, and does interfere with the angles SNA and SNB (Chapter 4). 4. The availability of a 2D lateral cephalometric radiograph influences the orthodontic treatment plan and decision in some but not all cases. (Chapter 5). 40 Therapeutic efficacy was defined as follows: 1. Percentage of times the image was judged helpful in planning management of the patients in a case series 2. Percentage of times therapy-planned pre-visualization of a lateral cephalogram needed to be changed after the image information was obtained 3. Percentage of times clinicians prospectively stated therapeutic choices needed to be changed after evaluating a cephalogram 4. Whether different analyses lead to different decisions on treatment planning 5. Intraand inter-observer identification errors 6. Reliability of landmark identification The analysis had to be based on primary materials or comprise a review on efficacy. When an abstract was considered by at least one author to be relevant, it was read in full text. At the second stage, the full texts were retrieved and critically examined. Reference lists of publications that had been found to be relevant in the first stage were hand-searched, and articles containing the words ‘cephalometry’, ‘lateral cephalometric radiography’, together with ‘treatment planning’, ‘orthodontic radiographs’, ‘landmark identification’ and ‘error’ were selected. Book chapters and reviews were excluded since the aim of this systematic review was to evaluate primary studies. 41 1.6 Data extraction Data was extracted with the aid of protocol 1 (Table 1.1). It was established by reading the relevant literature on how to critically evaluate studies about diagnostic methods. To minimise bias, two observers independently evaluated the quality and validity of original studies according to the quality assessment of diagnostic accuracy studies tool using protocol 2 (quality assessment of studies of diagnostic accuracy included in systematic reviews - QUADAS) (Table 1.2) (Whiting et al, 2003). When there was any disagreement concerning the relevance of an article, it was resolved by a discussion between the two reviewers. Each observer presented their arguments, and further discussion was held until a consensus was reached. Before the assessment, the protocols were tested for ten publications. A further five publications were read to calibrate the two reviewers regarding the criteria in protocol 2. Only publications that were found to be relevant to the reviewer in both protocols 1 (diagnostic efficacy) and 2 (level of evidence) were ultimately included. The quality and internal validity (level of evidence) of each publication was judged to be high, moderate or low according to the criteria in the following subsection. 42 Table 1.1. Protocol 1, Selection for inclusion of publications. First author: Title: Journal; Year; Volume; Pages: Yes No 1. Is there a well-defined hypothesis? 2. Are the accuracy, reliability, validity of cephalometry studied? 3. Is the contribution of cephalometry in determining the treatment plan evaluated? 4. Reliability of landmark identification in cephalometry? 5. Errors that occur in cephalometry? 6. What is the level according to Fryback and Thornbury? 7. Is the publication relevant for the review? 43 Table 1.2. Protocol 2, based on the QUADAS-2 tool for evaluation of methodology of included studies. Observer initials ______________ Date ________ Paper nº ⁄ ______ First author; Title; Journal; Year; Volume; Pages _______________________________________________________________________ 1. Are the results of the study valid? Yes No Unclear 2. Was the spectrum of patient’s representative of the patients who perform orthodontic treatment? Yes No Unclear 3. Were selection criteria clearly described? Yes No Unclear 4. Is the reference standard likely to correctly classify the target condition? Yes No Unclear 5. Were the methods for performing the radiographic examination described in sufficient detail to permit replication? Yes No Unclear 6. Was the execution of the reference standard described in sufficient detail to permit its replication? Yes No Unclear 7. Were the index test results interpreted without knowledge of the results of the reference standard? Yes No Unclear 8. Were the reference standard results interpreted without knowledge of the results of the index test? Yes No Unclear 9. Were the same clinical data available when test results were interpreted as would be available when the test is used in practice? Yes No Unclear 10. Were uninterpreTable ⁄ intermediate test results reported? Yes No Unclear 11. Were withdrawals from the study explained? Yes No Unclear 12. Was the number of observers sufficient to evaluate the influence of observer reproducibility and diagnostic efficacy? Yes No Unclear 13. Was observer reproducibility described? Yes No Unclear 14. Were appropriate results presented (percentage of correct diagnosis, sensitivity, specificity, predictive values, measurements of ROC, likelihood ratios, or other relevant measurements) and were these calculated appropriately? Yes No Unclear Comments 44 Levels of evidence and criteria for evidence synthesis: • High level of evidence A study was classified with high level of evidence if it fulfilled all of the following criteria: • There was an independent blind comparison between test and reference methods. • The population was described so that the status, prevalence and severity of the condition were clear. The spectrum of patients was similar to the spectrum of patients on whom the test method will be applied in clinical practice. • The results of the test method being evaluated did not influence the decision to perform the reference method(s). • Test and reference methods were well described concerning technique and implementation. • The judgments (observations and measurements) were well described considering diagnostic criteria applied and information and instructions to the observers. • The reproducibility of the test method was described for one observer (intraobserver performance) as well as for several (minimum 3) observers (interobserver performance). • The results were presented in terms of relevant data needed for necessary calculations. 45 • Moderate level of evidence A study was assessed to have a moderate level of evidence if any of the above criteria were not met. On the other hand, the study was assessed not to have deficits that are described below for studies with a low level of evidence. • Low level of evidence A study was assessed to have a low level of evidence if it met any of the following criteria: • The evaluation of the test and reference methods was non-independent. • The population was not clearly described, and the spectrum of patients was distorted. • The results of the test method influenced the decision to perform the reference method. • The test or the reference method or both were not satisfactorily described. • The judgments were not well described. • The reproducibility of the test method was not described or was described for only one observer. • The results could have a systematic bias. • The results were not presented in a way that allowed efficacy calculations to be made. 46 Rating conclusions according to evidence grade The scientific evidence of a conclusion on diagnostic efficacy was judged to be strong, moderately strong, limited or insufficient depending on the quality and internal validity (level of evidence) of the publications assessed (CBEM, Jaeschke et al., 1994). • Strong research-based evidence: at least two of the publications or a systematic review must have a high-level of evidence. • Moderately strong research-based evidence: one of the publications must have a high level of evidence and two more of the publications must have a moderate level of evidence. • Limited research-based evidence: at least two of the publications must have a moderate level of evidence. • Insufficient research-based evidence: scientific evidence is insufficient or lacking according to the criteria defined in the present study. 1.7 Synthesis of evidence The results of this review were described narratively. No meta-analyses were attempted because of lack of original studies. 1.8 Results The number of articles reviewed in each phase to perform this systematic review is presented in the PRISMA flow diagram (Figure 1.1) (Moher et al., 2009). The initial search revealed 784 articles listed in Medline (Ovid), 1,034 in Scopus and 47 264 articles in the Web of Science. The second stage of the search protocol was to retrieve the reference lists of the selected articles, which yielded 14 additional articles of interest. After excluding 1,128 duplicates, 968 articles remained for review. In the first phase selection, the observers screened the articles by reading titles and abstracts. Articles that were not eligible because of irrelevant aims and were not directly related to this systematic review were excluded, thus 203 articles remained for further reading. Thirty-five articles were assessed for eligibility. After screening all the articles using protocols 1 and 2, 17 articles met the inclusion criteria and were selected for qualitative synthesis and appraised to present some level of evidence. All articles that remained after screening passed the qualitative synthesis. These 17 articles were categorised by topics as follows: 7 studies on the role of cephalometry on the orthodontic treatment planning, 8 studies on cephalometric measurements and landmark identification and 2 studies on cephalometric analysis. 48 Figure 1.1. Methodology followed in the article selection process (adapted from: Moher et al., 2009). Records identified through database searching (n =2082) Additional records identified through other sources (n =15) Records after duplicates removed (n = 968) Records excluded (n =765) Records screened (n = 968) Full-text articles assessed for eligibility (n =35) Full-text articles excluded, with reasons (n =19) Studies included in qualitative synthesis (n =17) Screening Eligibility Included Identification 49 1.8.1 Role of cephalometry on the orthodontic treatment planning Seven articles related to the importance and contribution of cephalometry to orthodontic treatment planning was found (Table 1.3). Six of the publications were found to have low levels of evidence (Deveraux et al., 2011; Nijkamp et al., 2008; Bruks et al., 1999; Atchinson et al., 1992; Atchinson et al., 1991; Silling et al., 1979) and one classified as moderate level of evidence (Pae et al., 2001). 56 Table 1.4. Publications concerning landmark identification. Authors (year) Aim of the study Observers Subjects Design of the study Statistical method Results according to authors Level of evidence Baumrind and Frantz, 1971 Quantification of errors in landmark identification 5 observers 20 lateral skull radiographs Observer identified 16 cephalometric landmarks on a transparent plastic template Mean, standard deviation and standard errors Least reliable landmarks: Gonion and lower incisor apex Moderate Effects of errors on angular and linear measurements Kvam and Krogstad, 1969 Evaluation of measurements in lateral cephalograms. 18 observers 3 lateral skull radiographs Hand cephalometric analysis made by each participant, 8 angles measured Mean and standard deviation 16 out of 24 angular measurements: less variability in post-graduates than students Low Assess influence of knowledge and impact of angular errors In 7 measurements, no difference was observed Post-graduates' tracings used for diagnostic purposes Standard deviation of students greater than post-graduates Haynes and Chau, 1993 Evaluation of landmark identification on Delaire analysis 2 observers 28 lateral skull radiographs Establish a coordinate system for measurement on tracings Mean deviation Intra-observer: NS differences between values of T1 and T2 tracings Moderate 57 Comparison with data of conventional cephalometry Radiographs were traced twice by each observer (3 to 4 weeks) Inter-observer: differences between the averaged mean values on tracings were NS for either x or y co-ordinates Ahlqvist et al., 1986 Study the magnitude of projection errors on measurements in cephalometry 1 observer A patient was modelled Computer software designed to allow movement of model on the 3 axes. The magnitude of errors was studied by a diagram Measurement errors studied by a diagram with the relative length of distances between modelled landmarks Less than 1% error on length measurements if head is rotated up to 5° Low Study the effects of incorrect patient position on linear measurements Head rotated more than 5° the error is increased Houston et al., 1986 Evaluate errors at various stages of measurements in cephalometric radiograph 4 observers 24 lateral cephalograms 2 radiographs of the same patient Analysis of variance Error variance is small (radiograph and tracing) when compared with the variance among groups Moderate Radiographs traced on acetate sheet by each observer at T1/T2 (1-week interval) SNA has a higher tracing variance than SNB due to the difficulty to identify point A 58 Kamoen et al., 2001 Determine errors involved in landmark identification and its consequence to treatment results 4 observers 50 lateral cephalograms Items studied: (1) accuracy of digitiser, (2) intraand inter-observer digitising errors and (3) intraand interobserver tracing errors (1) Levene's test for homogeneity of variances, (2) one-way ANOVA and (3) Levene's test for homogeneity (1) NS variances of coordinates for landmark at different positions on the digitiser. (2) NS intraand inter-observer differences in digitisation. (3) S differences in landmarks and in the same landmark on different cephalograms and between observers Moderate Tng et al., 1994 Evaluate the validity of dental and skeletal landmarks. Effect on angles and distances. 1 observer 2 lateral cephalograms of 30 dry skulls Steel balls placed in 15 dental and skeletal landmarks Mean and standard deviation 7 out of 10 skeletal and 5 dental landmarks were NS (p < 0.05) Moderate Two radiographs taken with and without the markers and digitised. Measurements compared 4 angles (SNA-SN/MnP, MxP/MnP and LI/MnP) and 3 distances (N-Me, MxP-Me and Lie to APg) were invalid (p < 0.05) Major errors in angles with dental landmarks Bourriau et al., 2012 Analyse the influence of film-object distance and type of receptor on landmark identification 53 orthodontists 4 lateral cephalograms of the same patient 19 cephalometric landmarks on each film Mean NS difference between 2 imaging receptors neither between 2 cephalograms achieved by 2 equipments (p > 0.99) Low 59 2 radiographs performed at an equipment with a 4m arm and 2 in a 1.50-m arm equipment with 2 different imaging receptors (digital and indirect digital) Results obtained by cephalometric analysis was judged: ‘very important’ for 20.5%, ‘important’ for 70%, ‘less important’ for 8% and ‘accessory’ for 1 participant NS, non-significant; S, significant. 60 1.8.3 Cephalometric analysis Two publications with low-level evidence were found (Abdullah et al., 2006; De Abreu, 1982). The studies did not use any reference standards, and the number of observers was not stated. The study designs were also not clearly explained (Table 1.5). 61 Table 1.5. Publications on cephalometric analysis. Authors (year) Aim of the study Observers Subjects Design of the study Statistical method Results according to authors Level of evidence De Abreu, 1982 Assessment criteria of unanimity for different cephalometric analyses Not referred 129 patients Diagnosis performed based on Ricketts, Steiner, Cervera and Coutand cephalometric analyses Not referred 3 out of 61 cases with similar diagnosis. In 23 cases, 4 analyses achieved similar diagnosis. In 13 cases, 3 different diagnoses were obtained. In 8 cases, the diagnosis was different for class II and class III Low Abdullah et al., 2006 Examine accuracy and precision of Steiner analysis for changes on ANB angle, the PgNB distance and upper and lower incisor positions Different orthodontists (not reference to the number) 275 patients Radiographs traced and analysed by orthodontists according to the Steiner analysis Paired t test, mean and standard deviation The predicted change in L1 (lower incisor) to NB was underestimated by 0.8 mm. Only the prediction for Pogonion and NB showed improvement of the precision (30%) Low Radiographs at the end of treatment (T2) were traced by one observer 62 1.9 Discussion The validity, efficacy and contribution of cephalometry in orthodontic treatment planning remain questionable (Deveraux et al., 2011). In 2002, 90% of orthodontists in the USA routinely performed cephalometric radiographs (Nijkamp et al., 2008). This systematic review was performed to assess the validity and reliability of 2D lateral cephalometry used for orthodontic treatment planning as well as the errors that can occur on 2D tracing. Despite the abundant amount of articles found on lateral cephalometry (n = 968), it is surprising that the present systematic review could only identify very few studies (n = 17, 1.6%) on its validity and reliability. This finding underlines the need for the present study and is an important cross point, considering the fact that we are flooding into 3D cephalometric studies nowadays. Apart from our findings, 2D cephalometry has other specific limitations, such as orthognatic surgery, airway and growth assessment and skeletal maturation. In order to be included in this systematic review, publications had to satisfy pre-defined methodological criteria. Two protocols were used regarding the search strategy, one based on diagnostic methods and the second based on the QUADAS tool (Whiting et al., 2003). The ‘levels of evidence’ for assessing the quality and internal quality of each publication included in this review - how well the study was designed, how reliable its results appeared to be and the extent to which it addressed the questions posed - were modified according to the Oxford Centre for EvidenceBased Medicine levels of evidence for diagnostic methods (CBEM, 2012). Only publications assessed to present a high or moderate level of evidence can form the 63 basis for any scientific conclusions. Ten articles were identified as low level of evidence, five had moderate level and only one showed high level of evidence. All retrieved articles, assessing the importance and contribution of lateral cephalometric radiograph in orthodontic treatment, concluded that there is no significant difference on treatment planning decision with or without the evaluation of the lateral cephalogram. However, it should be considered that the suitable studies in this review were based on small samples rather than large cohorts representing the entire population. In one study, the sample used was restricted (six patients) (Deveraux et al., 2011). Furthermore, the short time lapse between observations in some studies did not allow a full washout effect, which could lead to the repetition of the results (Pae et al., 2001; Atchison et al., 1992; Atchison et al., 1991). The latter bias is further strengthened by the fact that recognition factors were often included, e.g. the possibility of identifying patient by photographic visualisation as part of the examination. On the other hand, in one paper, only dental casts were presented to the observers, which might also lead to error since it does not mimic the clinical situation. Sample bias is also suspected based on the fact that selection of subjects is often poorly described or unclear (Deveraux et al., 2011; Bruks et al., 1999; Silling et al., 1979), like the questions made to the observers that were not stated by any questionnaire (Bruks et al., 1999), and in one article, observers were forced to choose yes/no answers, which again do not perfectly simulate the reality (Nijkamp et al., 2008). In the two articles by Atchison et al., there was the possibility to identify patients as well as sample size was very restricted (six patients). There was no repetition of the questionnaire to test the variability between answers (Atchison et 64 al., 1992; Atchison et al. 1991). When it comes to the validity and reliability of cephalometric analysis, several errors should be considered: landmark identification, tracing and measuring, and magnification of certain anatomical structures. Landmarks placed in anatomically formed edges are easier to identify, while some landmarks placed on curves are more prone to error. The gonion and lower incisor apex are the least consistent landmarks (Baumrind and Frantz, 1971). Furthermore, landmarks such as point A have a higher variance than others like point B because of wider variation and anatomical localisation of point A (Houston et al., 1986). Dental landmarks tend to have poorer validity than skeletal landmarks. Also, when landmarks are located on a curve like point A, point B or Pogonion, the error is larger (Tng et al., 1994). The evidence shows that landmark identification is a great source of error in 2D lateral cephalometry (Kamoen et al., 2001). Major errors in angles with dental landmarks may occur (Tng et al., 1994). In addition, different levels of knowledge and experiences between the observers also lead to varying results on landmark identification. In a study using 18 observers, in which 13 were dental students and 5 were post-graduate's, the lasts revealed lower intra-observer tracing variance than dental students (Kvam and Krogstad, 1969). Patient positioning during the procedure is also very important to avoid errors on measurements and landmark identification (Houston et al., 1986; Ahlqvist et al., 1986). The publication of Ahlqvist et al., 1986 was assessed with a low level of evidence because there was only one observer. A similar classification occurred for Bourriau et al., 2012, intra-observer agreement could not be evaluated and the number of radiographs (n = 4) used was very low. Kvam 65 and Krogstad's (1969), publication also used a limited number of subjects (n = 3). The choice of the observers also plays an important role on the results. Eighteen observers, in which 13 were dental and 5 were post-graduate students, participated in their study (Kvam and Krogstad, 1969). The latter can also bias results because of the distinct level of education and expertise due to the lack of experience of the observers. Regarding the influence of magnification, Bourriau et al., 2012 could not identify significant differences between equipment with a 4-m distant cephalometric machine and a 1.5-m distant cephalometric arm. Despite that, it should be considered that distance varying between the X-ray source and the image receptor will always cause a degree of magnification, the larger the distance, the lower the magnification. A focus object distance of 4 m in 2D cephalometric equipment is usually favoured for the reduced radiation burden and lack of enlargement, while equipment with 1.5-m arm has a direct advantage of being compact and integrated in a multimodal system as well as having an increased resolution. On the other hand, panoramic equipment with a cephalometric arm at a 1.5-m distance may present shortcomings in enlargement factors and superimposition of the bilateral structures more distant from the midsagittal plane, considering the less magnified structures on the side nearby the image receptor (White and Paroah, 2009). We were not able to identify studies correlating landmark identification errors in lateral cephalograms and their influence on the outcome of patient treatment. Finally, in 1982, De Abreu showed that different 2D cephalometric analysis may lead to different diagnosis of the same patient, varying the diagnosis between 72 radiographic evaluation, the skull position was adjusted to allow the Frankfort horizontal plane to be parallel with the horizontal plane for further measurements. Analysis Two experienced observers (dentomaxillofacial radiologists) performed this study with a session of calibration prior to the analysis. Ten commonly used skeletal landmarks were identified on twenty skulls and radiographs according to figure 2.1 (Proffit et al., 2006). Both observers had been informed about all the anatomical landmarks, identification methods used on radiographs, and also craniometric measurement of the skulls. Five skulls and its radiographs were used for calibration. At the end of the calibration, both observers were in agreement and any remaining doubt was clarified. In case of any uncertainty between the two observers, an additional advice from a third observer was essential to reach agreement. 73 Figure 2.1. Cephalometric landmarks used in the study. N – Nasion; Me – Menton; ANS – Anterior Nasal Spine; Co – Condylion; Gn – Gnathion; A – Point A; B - Point B; Pog – Pogonion; Po – Porion; Or – Orbitale; Go – Gonion. Craniometric measurements considered to be the gold standard were done on 20 dry dentate skulls by using a digital caliper (Absolute Digimatic Caliper No. 500-161U; Mitutoyo America Corp., Aurora, IL). The same measurements were performed by digital determining the landmarks on the viewing monitor in a dimlighted room without any interruption. All measurements were repeated one month later, both on skulls and radiographs. The results of the intraand interobserver reliability were analysed. The linear measurements were chosen according to the vertical and anteroposterior dimensions of the craniofacial form (Table 2.1). The landmarks on which these measurements were based represented both midsagittal and bilateral anatomical structures. 74 Table 2.1. Linear measurements evaluated on human skulls and lateral cephalometric radiographs in this study (mm). Linear Measurements Total anterior face height: N-Me Upper face height: ANS-N Lower face height: ANS-Me Mandibular unit length: Co-Gn Maxillary unit length: Co-ANS AN: A to N with respect to true vertical BN: B to N with respect to true vertical PogN: Pog to N with respect to true vertical Po-Or (Frankfort plane) Go-Me (mandibular plane) Statistical analysis Variables were described through its mean, standard deviation and measurements of dispersion. Intraand inter-observer variation was studied using the intraclass correlation coefficient (ICC) with a confidence interval of 95%. General guidelines for this measure rate an ICC > 0.90 as excellent, an ICC of 0.75–0.90 as good, and an ICC < 0.75 as representing poor to moderate reliability (Shrout and Fleiss, 1979). Differences between the measurements performed on skulls and on radiographs were evaluated by the Bland-Altman limits of agreement (Bland and Altman, 1986). One sample t-test was used to evaluate if the mean of the differences between the two measurements was different from 0 (Moore and McCabe, 2006). The Statistical Package for Social Sciences 20.0 for Windows (SPSS Inc., Chicago, Illinois, USA) was used for statistical analysis. The level of statistical significance for all tests was set at α = 0.05. 75 2.3 Results Intra-observer consistency is shown on Table 2.2. On Table 2.3 the inter-observer reliability is presented. Craniometric measurement revealed ICC values in general, above 0.90, for the intra-observer reliability, with exception of the A-N measurement for observer 2, which showed an ICC of 0.76 (Table 2.2). For the inter-observer reliability seen in craniometric measurement, the ICC was also, in general, above 0.90, with exception of ANS-N for the second observation; A-N and Po-Or for both observations (Table 2.3). Intra-observer reliability for the linear measurement on radiographs revealed ICC values above 0.90, except for ANS-N and Co-ANS for the second observer, and A-N for both observers (Table 2.2). There was an overall good agreement regarding inter-observer reliability for linear measurement performed on radiographs, when comparing between linear measurements, with the exception of ANS-N, Co-ANS, A-N and Po-Or for both observations (Table 2.3). With regards to accuracy of 2D cephalometric radiographs, the mean differences between linear measurements (mm) when performed by both observers on skulls and radiographs were investigated and the results are shown in Table 2.4. Radiograph and craniometric measurements presented statistically significant differences between them, with p < 0.05, implying that there was a difference in landmark identification between these two modalities. 76 We found that seven of the ten linear measurements on radiographs were on average significantly higher. Only three of the linear measurements were on average significantly higher when performed directly on the skulls (Co-Gn, CoANS, and Go-Me). It was seen that these three measurements had at least one bilateral landmark. The widest deviation between the two methods was seen on the measurement N-Me, with a difference of 0.96 mm. The lowest value was detected on the measurements between Co-Gn (0.14) and Po-Or (0.14). BlandAltman limits of agreement showed the 95% differences between measurements performed on the skulls and on radiographs. All the differences found between the two methods were inferior to two units of measurement (mm), which is, generally, within one standard deviation of the norm values in cephalometric analysis (Chen et al., 2004). 77 Table 2.2. Mean differences between the first and second observations with regards to intra-observer agreement (mm). Observation 1 Observation 2 Mean (SD) ICC CI 95% LA Mean (SD) ICC CI 95% LA N-Me Skull 10.08 (0.96) 0.999 0.997-0.999 -0.10;0.09 10.08 (0.96) 0.998 0.995-0.999 -0.11;0.12 Radiograph 11.02 (1.01) 0.978 0.948-0.991 -0.47;0.36 11.03 (1.02) 0.999 0.998-1.000 -0.06;0.09 ANS-N Skull 4.41 (0.32) 0.949 0.810-0.978 -0.19;0.21 4.43 (0.34) 0.926 0.832-0.969 -0.26:0.26 Radiograph 4.79 (0.35) 0.905 0.786-0.960 -0.36;0.25 4.82 (0.32) 0.831 0.636-0.926 -0.49;0.39 ANS-Me Skull 5.87 (0.72) 0.997 0.94-0.999 -0.14;0.06 5.84 (0.76) 0.980 0.952-0.991 -0.34;0.26 Radiograph 6.38 (0.82) 0.984 0.961-0.993 -0.34;0.24 6.43 (0.83) 0.973 0.937-0.989 -0.49;0.26 Co-Gn Skull 10.87 (0.89) 0.989 0.974-0.996 -0.31;0.20 10.85 (0.87) 0.994 0.985-0.997 -0.25;0.13 Radiograph 10.72 (0.93) 0.989 0.973-0.995 -0.28;0.27 10.71 (0.90) 0.982 0.957-0.992 -0.36;0.32 Co-ANS Skull 9.19 (0.60) 0.981 0.954-0.992 -0.24;0.22 9.22 (0.60) 0.972 0.934-0.988 -0.34;0.22 Radiograph 8.54 (0.57) 0.935 0.851-0.973 -0.40;0.42 8.61 (0.50) 0.845 0.663-0.933 -0.72;0.43 A-N Skull 4.97 (0.35) 0.911 0.798-0.962 -0.27;0.32 4.90 (0.35) 0.763 0.512-0.895 -0.43;0.59 Radiograph 5.31 (0.36) 0.797 0.573-0.911 -0.51;0.45 5.39 (0.35) 0.619 0.276-0.822 -0.58;0.76 B-N Skull 8.49 (0.74) 0.982 0.957-0.993 -0.26;0.29 8.57 (0.75) 0.959 0.905-0.983 -0.53;0.32 Radiograph 9.25 (0.76) 0.991 0.979-0.996 -0.20;0.20 9.39 (0.82) 0.984 0.962-0.993 -0.27;0.31 Pog-N Skull 9.39 (0.88) 0.991 0.979-0.996 -0.24;0.22 9.45 (0.87) 0.982 0.958-0.993 0.24;0.41 Radiograph 10.29 (0.95) 0.982 0.956-0.992 -0.34;0.38 10.29 (0.97) 0.991 0.978-0.996 -0.26;0.25 Po-Or Skull 7.24 (0.38) 0.957 0.901-0.982 -0.13;0.32 7.40 (0.41) 0.910 0.082-0.745 -0.78;1.14 Radiograph 7.42 (0.40) 0.957 0.900-0.982 -0.28;0.18 7.50 (0.38) 0.906 0.789-0.960 -0.36;0.30 Go-Me Skull 7.43 (0.57) 0.955 0.895-0.981 -0.30;0.37 7.55 (0.65) 0.931 0.841-0.971 -0.44;0.53 Radiograph 7.05 (0.55) 0.936 0.853-0.973 -0.42;0.36 7.03 (0.54) 0.952 0.889-0.980 -0.23;0.44 SD – standard deviation; ICCIntraclass correlation; CI (5% - 95%) confidence interval; LALimits of agreement 78 Table 2.3. Inter-observer agreement (mm). Observer 1 Observer 2 Mean (SD) ICC CI 95% LA Mean (SD) ICC CI 95% LA N-Me Skull 10.08 (0.96) 0.997 0.993-0.999 -0.14;0.14 10.07 (0.95) 0.999 0.998-1.000 -0.07;0.08 Radiograph 11.02 (1.00) 0.972 0.934-0.988 -0.52;0.43 11.04 (1.01) 0.996 0.900-0.998 -0.16;0.20 ANS-N Skull 4.42 (0.32) 0.954 0.893-0.981 -0.20;0.19 4.42 (0.32) 0.852 0.677-0.936 -0.38;0.34 Radiograph 4.78 (0.32) 0.855 0.684-0.937 -0.40;0.32 4.83 (0.39) 0.861 0.694-0.940 -0.45;0.38 ANS-Me Skull 5.83 (0.74) 0.992 0.982-0.997 -0.15;0.21 5.88 (0.74) 0.980 0.951-0.991 -0.27;0.32 Radiograph 6.36 (0.82) 0.953 0.890-0.980 -0.51;0.49 6.44 (0.82) 0.985 0.965-0.994 -0.36;0.20 Co-Gn Skull 10.83 (0.87) 0.982 0.957-0.992 -0.27;0.29 10.89 (0.89) 0.994 0.986-0.998 -0.18;0.20 Radiograph 10.71 (0.90) 0.978 0.947-0.991 -0.36;0.39 10.72 (0.92) 0.990 0.977-0.996 -0.25;0.26 Co-ANS Skull 9.18 (0.61) 0.982 0.957-0.992 -0.23;0.23 9.22 (0.60) 0.990 0.976-0.996 -0.22;0.12 Radiograph 8.55 (0.55) 0.857 0.688-0.938 -0.59;0.60 8.61 (0.52) 0.866 0.706-0.942 -0.72;0.43 A-N Skull 4.96 (0.34) 0.857 0.687-0.938 -0.33;0.41 4.91 (0.37) 0.867 0.707-0.943 -0.29;0.48 Radiograph 5.36 (0.36) 0.673 0.361-0.850 -0.77;0.49 5.33 (0.35) 0.740 0.470-0.883 -0.55;0.51 B-N Skull 8.51 (0.74) 0.954 0.892-0.980 -0.47;0.42 8.55 (0.73) 0.947 0.877-0.978 -0.62;0.33 Radiograph 9.33 (0.79) 0.977 0.945-0.990 -0.49;0.19 9.32 (0.79) 0.984 0.962-0.993 -0.41;0.14 Pog-N Skull 9.44 (0.88) 0.991 0.980-0.996 -0.34;0.11 9.40 (0.87) 0.980 0.952-0.992 -0.36;0.33 Radiograph 10.29 (0.96) 0.972 0.933-0.988 -0.44;0.46 10.29 (0.95) 0.989 0.973-0.995 -0.26;0.26 Po-Or Skull 7.35 (0.38) 0.805 0.116-0.706 -1.07;0.66 7.25 (0.41) 0.804 0.586-0.914 -0.64;0.41 Radiograph 7.45 (0.39) 0.944 0.871-0.976 -0.35;0.16 7.48 (0.39) 0.873 0.720-0.945 -0.47;0.32 Go-Me Skull 7.51 (0.61) 0.919 0.816-0.966 -0.62;0.36 7.47 (0.63) 0.925 0.829-0.968 -0.50;0.42 Radiograph 7.06 (0.51) 0.901 0.778-0.958 -0.50;0.42 7.02 (0.57) 0.950 0.883-0.79 -0.27;0.45 SD – standard deviation; ICCIntraclass correlation; CI (5% - 95%) confidence interval; LALimits of agreement 79 Table 2.4. Mean of differences and level of agreement between measurements performed on the skull and radiography. Mean of differences (mm) p LA N-Me -0.96 <0.001 -1.710;-0.742 ANS-N -0.39 <0.001 -0.712;-0.067 ANS-Me -0.58 <0.001 -0.869;-0.294 Co-Gn 0.14 <0.001 -0.191;0.477 Co-ANS 0.62 <0.001 0.252;-0.986 A-N -0.41 <0.001 -0.753;-0.074 B-N -0.79 <0.001 -1.179;-0.409 Pog-N -0.87 <0.001 -1.148;-0.602 Po-Or -0.15 0.001 -0.860;0.566 Go-Me 0.45 <0.001 0.038;0.859 p - One-sample ttest; LALimits of agreement 2.4 Discussion Evidence shows that landmark identification is a great source of error in 2D cephalometric analysis because of the uncertainty in recognizing accurately where the landmark is located. Some landmarks also show a wider variation in localization than others (Tng et al., 1994; Baumrind and Frantz, 1971). Superimposition between bilateral anatomical structures and anatomical localization may hinder its identification, as for example of landmarks Co, Go, Po, Or, and lower incisor apex (Tng et al., 1994; Baumrind and Frantz, 1971). Therefore, it is essential to accurately determine anatomical landmarks in order to reduce linear measurement error in cephalometric analysis. Moreover, it is important to assess the quantitative differences between craniometric measurement and the corresponding radiographic measurements. 80 The observers’ agreement is another factor that influences the measurement error. Chen et al. (2004) found that in general the inter-observer error presents greater values than the intra-observer error. We confirmed that, on average, there was a higher rate of inter-observer error. We found that intraand inter-observer reliability of linear measurements performed on skulls were on average significantly lower than on radiographs (Table 2.2 and 2.3). Table 2 shows that intra-observer reliability for skull linear measurement AN was the least consistent for observer 2, with an ICC of 0.76. When comparing intra-observer reliability on radiographs, the lowest agreement was seen in A-N, Co-ANS and ANS-N, respectively, for both observers. Linear measurement A-N showed a lower agreement between observers both on skulls and on radiographs. This might be due to the localization of point A, Co and ANS (Tng et al., 1994; Baumrind and Frantz, 1971). The evidence shows that bilateral anatomical landmark identification, such as Co, is a great source of error in 2D lateral cephalometry (Tng et al., 1994). Relating to points A and ANS, they might appear more radiolucent on radiograph, which may lead to uncertain position of these landmarks. Intraand inter-observer SD for the skulls and radiographs were lower (value inferior to 0.5) for linear measurements ANS-N, A-N and Po-Or on observations 1 and 2. On average, in a 12-years old male, the Harvold linear measurement ANSMe presents a SD of approximately 3.7 mm (Proffit et al., 2006), which is a value higher than the ones we found (maximum 0.83). 81 The results revealed that, in general, craniometric measurements tended to be shorter than linear measurement on radiographs, except for Co-Gn (mandibular unit), Co-ANS (maxillary unit), and Go-Me (mandibular plane) (Table 2.4). This may be related with the fact that on these linear measurements, at least one of the landmarks is placed on bilateral structures (Co and Go), which may have increased this variability. Also, it is more difficult to establish a middle point directly on the skull than on the radiograph. Validity of cephalometric distances depended on the validity of individual landmarks involved. In the case of a linear measurement, it is known that the shorter the line segment measured, the greater the percentage of error produced by a given measurement error (Chen et al., 2004). Our results contrast with the study from Farkas et al. (2002), where they found that singular and paired cephalometric distances were significantly shorter than the craniometric distances on postero-anterior cephalometric radiographs. Our ten measurements were statistically significant (p<0.05), even though intervals oh the limits of agreement were on average low (see Table 3.4). The mean difference was significant and presented the highest variance for total anterior face height linear measurement (on average, N-Me at 0.956 mm). This means that there is a 95% chance that the value varies from -1.71 to -0.74, which is within the clinically acceptable limits, since it is inferior to 1 mm (Table 2.4). The McNamara cephalometric analysis, published in 1983, estimated an error of +/- 2 mm for linear measurement A-N (Proffit et al., 2006), while in the 88 believe that errors can be caused by different individual conceptions of landmark definition and its perception, rather than education and training (Kamoen et al., 2001; Lau et al., 1997). Inconsistency of landmark identification can increase the degree of error (Silveira and Silveira, 2006; Chen et al., 2004; Chen et al., 2000). Inter-observer reproducibility of landmark identifications was found to be very low among dentomaxillofacial radiologists (DMFR) (Silveira and Silveira, 2006). Some dentomaxillofacial radiologists, as well as orthodontists, are trained to perform 2D cephalometric analyses. There are no previous reports on the reliability of landmark identification that compare orthodontists and dentomaxillofacial radiologists. Therefore, the aims of the present study were to evaluate the reproducibility of 17 commonly used cephalometric landmarks by orthodontists and dentomaxillofacial radiologists, and to assess the impact of different landmark identifications on patient diagnosis. 3.2 Materials and Methods Twenty digital lateral cephalometric radiographs were selected from the database at the Oral Imaging Center, University of Leuven. Lateral cephalograms were acquired by positioning the patients in a standard digital cephalometric device and using a charged couple device sensor (Veraviewepocs 2D ® , J. Morita, Kyoto, Japan). Exposure values were set at 77 kV and 7.2 mA, with an exposure time of approximately 1.6 s, according to each patient. Inclusion criteria were: 89 • No evidence of current orthodontic treatment. • Digital cephalometric image were of good quality to allow the identification of landmarks, and the ruler on the radiograph was clearly visible, allowing calibration of the images in the cephalometric analysis software program. • There were no unerupted or partially erupted incisors that could have compromised landmarks identification. • No gross skeletal asymmetry. All selected images were then exported in TIFF format, and subsequently imported to the computerized program for cephalometric analysis (Radiocef Studio 2, Radio Memory Ltd., Belo Horizonte, Brazil). Analysis Seventeen commonly used cephalometric landmarks were included in this analysis; these are shown in Figure 3.1 (Proffit et al., 2006). Landmark identification was carried out on the digital image, using a mouse-driven cursor in a predetermined sequence. 90 Figure 3.1 - Cephalometric landmarks used in the study. N – Nasion; Or – Orbitale; S – Sella; Co – Condylion; Po – Porion; PNSPosterior Nasal Spine; ANS – Anterior Nasal Spine; A – Point A; UIA – Upper incisor apex, UIB – Upper incisor border; LIB – lower incisor border; LIA – Lower incisor apex; B - Point B; Pog – Pogonion; Gn –Gnathion; Me – Menton; Go – Gonion. Eight experienced observers (four orthodontists and four dentomaxillofacial radiologists) performed this study. Experience of the observers ranged from eight to 15 years. An initial training and calibration session was attended by all observers, including an explanation of the anatomical structures and required landmark identification. At the end of the session, the main author clarified any remaining doubt. Thus, all observers followed the same definitions of landmarks 91 in the identification process. For optimal visualization, landmark identification was performed in a dim-lighted room without any interruption. Intra-observer reliability was assessed by one dentomaxillofacial radiologist, repeating the same procedure 3 months after. After selecting a landmark with the mouse cursor, a dot on the monitordisplayed image indicated its position. Landmark position could be corrected until the operator was satisfied. Vertical and horizontal positions of each landmark were recorded in the format of x and y coordinates. Landmarks’ digitized coordinates were then imported into the Excel software (Version 2003; Microsoft, Redmond, Washington, USA). Statistical analysis was performed using Statistical Package for Social Sciences, version 20.0 for Windows (SPSS Inc., Chicago, Illinois, USA). The level of statistical significance for all tests was set at α = 0.05. We also compared which group was closer to the gold standard measurements. Some linear and angular measurements used in Ricketts and McNamara’s cephalometric analysis were performed by all observers on 20 cephalometric radiographs. Of these, three radiographs were classified as borderline cases, in between orthognathic surgery and orthodontics. An example, showing the differences on landmark identification by two observers is seen in Figure 3.2. In general, the differences ranged between 1 and 2 mm. 92 Figure 3.2 – Example of a lateral cephalometric radiography with identification of landmarks by two observers. The same computer software was used to access differences on angular and linear measurements. The angular and linear measurements used were the following: AN; Co-Gn (Mandibular unit length); Co-A; (Po-Or).(Go-Me); Pog-N (Facial plane); LIB (A-Pog); Convexity of Point A; Go-Me (Mandibular Plane) and S-Go. 93 Statistical analysis To analyse the precision in landmark identification, each landmark’s mean, standard deviation and measurements of dispersion were calculated. Intraand inter-observer reliability for each landmark in the x and y directions were studied, using intraclass correlation coefficients (ICC) with a confidence interval of 95%. General guidelines for this measure rate an ICC > 0.90 as excellent, an ICC of 0.75–0.90 as good, and an ICC < 0.75 as representing poor to moderate reliability (Shrout and Fleiss, 1979). Therefore, the “best estimate” of landmarks identification was obtained from the mean value of each landmark identified by the observers, and defined as the gold standard. Inter-observer reliability was assessed using the Euclidean distances. The average distance between the mean positions pointed by an observer was calculated and presented as the “intra-observer error”, which was an indicator of reliability. Differences in landmark location were analysed by Student’s t-test with the significance level of p < 0.05. 3.3 Results The ICC was calculated for the intraand also for inter-observer repeatability in the two groups (Table 3.1). In general, the ICC from the intra-observer ranged above 0.90, which implied an excellent agreement, with exception of the x direction of points Po, Me and point B. y component of landmarks N, Or and S, was considered good (ICC between 0.75 and 0.90). Furthermore, the vertical 94 components of landmarks Go and point B revealed a poor or moderate agreement (ICC <0.75) seen for the intra-observer reproducibility. The ICC for the inter-observer reliability was in general above 0.90 for all observers, with exception of the x component of points N and Or, which presented a good agreement (Table 3.1). Overall, in both groups, the highest variation found was associated with the vertical component of point Go (1.73 mm) and the lowest was seen in the vertical component of point Po (0.04 mm). For the orthodontists, the ICC showed an overall lower value when compared to DMFRs. The ICC varied from 0.75 to 0.90 regarding landmarks Or, Po, Gn, point B and UIA. Likewise, the x coordinates of landmarks N, Me, Pog, point A, PNS, LIA and LIB also showed a good agreement. Only the x coordinate of ANS was classified as having poor or moderate agreement. For dentomaxillofacial radiologists, the overall ICC was higher than 0.90. The exceptions were the x coordinates of landmarks Or and Po and the y coordinates of Go and point B, for which the agreement was good. Poor or moderate agreement was observed in the y component of Or. Overall in both groups, there was a high variation related to point Co, in the x direction. Between two DMFRs observers there was a difference of 5.05 mm; and between orthodontists, a difference of 3.56 mm was found. The horizontal component of point Or was less reproducible for DMFRs. Point Go in the x and y directions, points Me and PNS in the x direction and point B in the y direction were less precise among orthodontists. 95 Table 3.1. ICC for interand intra-observer evaluation. Inter-observer * Intra-observer ** ICC 95% CI ICC 95% CI N – Nasion x 0.78 0.310-0.848 0.99 0.985-0.998 y 0.99 0.985-0.998 0.85 0.301-0.845 Or – Orbitale x 0.88 0.726-0.951 0.91 0.797-0.965 y 0.96 0.923-0.987 0.78 0.202-0.812 S – Sella x 0.99 0.984-0.997 0.99 0.976-0.996 y 0.98 0.962-0.994 0.87 0.701-0.946 PoPorion x 0.97 0.916-0.986 0.80 0.569-0.917 y 0.95 0.889-0.982 0.94 0.845-0.974 Co – Condylion x 0.91 0.797-0.965 0.94 0.856-0.976 y 0.97 0.917-0.986 0.96 0.899-0.983 Go – Gonion x 0.98 0.956-0.993 0.98 0.964-0.994 y 0.93 0.830-0.971 0.71 0.394-0.872 Me – Menton x 0.97 0.918-0.987 0.84 0.631-0.932 y 0.99 0.970-0.995 0.97 0.932-0.989 Pog – Pogonion x 0.98 0.941-0.990 0.98 0.949-0.992 y 0.98 0.951-0.992 0.98 0.951-0.992 Gn – Gnathion x 0.97 0.924-0.988 0.97 0.936-0.990 y 0.99 0.963-0.994 0.99 0.954-0.979 B point x 0.97 0.930-0.989 0.89 0.754-0.957 y 0.96 0.894-0.983 0.72 0.418-0.879 A point x 0.97 0.916-0.986 0.92 0.804-0.966 y 0.98 0.960-0.994 0.95 0.890-0.982 ENA x 0.93 0.823-0.970 0.95 0.868-0.978 y 0.99 0.968-0.995 0.97 0.939-0.990 ENP x 0.98 0.952-0.992 0.94 0.852-0.975 y 0.99 0.972-0.996 0.99 0.969-0.995 LIA x 0.97 0.932-0.989 0.97 0.917-0.987 y 0.97 0.937-0.990 0.99 0.969-0.995 LIB x 0.98 0.961-0.994 0.98 0.960-0.994 y 0.99 0.968-0.995 0.99 0.981-0.997 UIB x 0.98 0.951-0.992 0.98 0.960-0.994 y 0.98 0.947-0.991 0.98 0.963-0.994 UIA x 0.97 0.925-0.988 0.98 0.956-0.993 y 0.97 0.924-0.988 0.98 0.944-0.991 * Between the 8 observers; ** a dentomaxillofacial radiologist 96 The Euclidean distance was used to test differences in landmark identification among observers and regarding the gold standard. The mean location differences of all landmarks for the orthodontists ranged from 5.92 mm to 0.99 mm. Generally, the landmark with least location differences was LIB (0.99 mm; SD 0.65 mm) and the one with most differences was point Gn (5.92 mm; SD 4.59 mm). The minimal and maximal variations on the horizontal component were associated with point ANS (0.53 mm; SD 3.74 mm and 2.97 mm; SD 2.02 mm). Regarding the reproducibility of the vertical component, point A presented the minimum variation (0.10 mm; SD 1.86 mm), while Gn was the most variable (4.60 mm; SD 3.67 mm). Table 3.2 shows the Euclidean distances between the “best estimate” of each landmark and orthodontist observers, defined as the interobserver error of landmark identification. In general, orthodontists revealed errors inferior to 1 mm in points S, Pog, LIB, and UIB in both horizontal and vertical directions. Table 3.2. Minimum and maximum euclidean distances (in mm) for orthodontists, defined as absolute differences in millimetres between the mean values and standard deviations of each landmark and the averaged for all observers. 97 Landmark Horizontal component (x) Vertical component (y) Mean SD p Mean SD p N – Nasion Minimum 0.19 3.30 0.804 -0.34 0.74 0.054 Maximum 1.43 2.74 0.031 0.21 1.89 0.631 Or – Orbitale Minimum 0.24 1.46 0.650 0.20 1.74 0.902 Maximum 2.04 1.85 <0.001 -1.29 2.13 0.014 S – Sella Minimum 0.24 0.92 0.263 -0.32 0.67 0.048 Maximum -0.42 0.80 0.029 -0.76 1.09 0.006 PoPorion Minimum 0.53 1.59 0.155 0.44 1.47 0.194 Maximum -2.60 2.08 <0.001 -1.42 2.28 0.012 Co – Condylion Minimum 0.23 1.69 0.549 -2.61 2.11 <0.001 Maximum -2.66 1.13 <0.001 1.15 2.68 0.069 Go - Gonion Minimum 0.50 1.34 0.110 -1.09 2.47 0.063 Maximum 2.20 1.83 <0.001 2.14 2.36 0.001 Me - Menton Minimum 0.73 3.77 0.395 -0.42 1.84 0.323 Maximum -1.57 2.38 0.008 1.44 2.64 0.025 Pog - Pogonion Minimum 0.27 2.30 0.430 -0.16 1.99 0.984 Maximum 0.42 3.31 0.576 0.61 2.23 0.238 Gn - Gnathion Minimum -0.31 2.22 0.025 0.21 2.00 0.847 Maximum 2.27 4.11 0.023 4.60 3.67 0.000 B point Minimum 0.20 1.84 0.725 0.21 2.87 0.759 Maximum 0.87 3.79 0.319 3.38 2.73 <0.001 A point Minimum 0.17 2.38 0.754 0.10 1.32 <0.001 Maximum 1.27 2.40 0.029 2.08 2.25 0.001 ANS Minimum 0.16 3.06 0.897 0.22 1.73 0.599 Maximum 2.97 3.74 0.002 -0.81 1.10 0.004 PNS Minimum 0.54 1.44 0.113 0.34 1.26 0.240 Maximum -2.66 1.43 <0.001 -0.80 1.23 0.009 LIA Minimum -0.38 2.38 0.495 -0.37 2.51 0.516 Maximum -1.44 2.01 0.005 2.61 2.53 0.010 LIB Minimum -0.10 2.01 0.781 0.24 2.19 0.631 Maximum 0.40 2.39 0.461 1.00 2.59 0.048 UIB Minimum -0.23 1.71 0.556 -0.21 2.13 0.659 Maximum -0.68 1.59 0.069 1.39 2.16 0.010 UIA Minimum 0.33 2.50 0.566 0.21 2.41 0.896 Maximum -1.20 1.87 0.010 -2.41 1.61 0.000 SD – standard deviation; ICCIntraclass correlation; CI (5% - 95%) confidence interval 104 point Go. Comparing to our findings, the variation of this landmark was smaller than 1 mm for one orthodontist and one DMFR. This could have happened due to difficulty in establishing the landmark along broadly curved structures, such as the mandible. Some cephalometric landmarks are more reliable in either the horizontal or vertical plane, making the distribution of errors follow a pattern of a non-circular envelope (Baumrind and Frantz, 1971). Differences in identification of landmarks were found in both groups along both axes. Although greater differences, were seen on the horizontal axis. Orthodontists and DMFRs revealed more variations when identifying the horizontal component of some landmarks. Despite that the majority of landmarks revealed low reproducibility, the x and y components of points Or, Go, Gn and LIA, the x coordinates of points Po, ANS, Co, PNS and the y component of point B showed a mean value, for at least one observer, higher than 2 mm. Landmark identification errors inferior to 1 mm are considered accurate (Richardson, 1981; Chen et al, 2000 ). Other authors showed that a difference of 2 mm is considered acceptable and does not have any influence on orthodontic diagnosis and treatment plan (Kamoen et al., 2001; Lau et al, 1997). We found differences in the SNA angle on three patients. Apart from that, we did not see differences between diagnoses associated with a variation of 1 to 2 mm in landmark identification. A greater difference regarding all landmarks would probably have an impact on diagnosis and possibly, in treatment planning. The latter is important, since variations that might exceed the SD proposed for a predetermined linear or angular measurement performed by one observer could demonstrate lack of knowledge and/or experience. This is particularly important 105 as we considered that variations higher than 2 mm can relate to lack of knowledge and/or experience of the observers. Although inter-observer variations occurred heavily on landmark identifications, they may not have an impact in patient diagnosis. This is one of the reasons why landmark identification reproducibility is quite low. Thus, reliability of cephalometric analysis should be questioned. Depending on the observer and on the type error, different results with no impact on diagnosis and treatment planning may appear. The existing literature suggests that lateral cephalometric radiographs have been used without adequate scientific evidence of its utility, and that it is often used prior to treatment. The evidence to agree or disagree with the usefulness of this radiographic technique in orthodontics is limited (Durão et al., 2013). Many variables contribute to the final diagnosis and treatment plan in orthodontics, such as face-bow recording, clinical examination, intraand extraoral photographs. Therefore, it is difficult to predict if a single error on landmark identification will have an impact on clinical practice. A combined error on dental casts and cephalometric analyses may lead to erroneous decisions about teeth extraction (Silveira and Silveira, 2006). The patient should be treated with maximum accuracy in every steps of diagnosis and treatment. 106 3.5 Conclusions We verified that some landmarks were not as reproducible as others, either on the horizontal or vertical component. The most consistent landmark identified in both groups was the LIB, while the least reliable points were Co, Gn, Or and ANS. Furthermore, the greatest variation was found in Co-Gn plane. Our results suggest a low reliability in the identification of cephalometric landmarks and lower agreement between orthodontists. In the presence of a range of variation from 1 to 2 mm on landmark identification, the patient’s diagnosis was altered. Moreover, we found changes in the SNA angle. Further studies focusing on the impact of deviating cephalometric analysis on a larger sample and in borderline cases may be needed to determine the real clinical impact. 107 CHAPTER 4. Variations in Sella landmark identification and its effect in angles SNA and SNB in lateral cephalometric radiographs 108 109 VARIATIONS IN SELLA LANDMARK IDENTIFICATION AND ITS EFFECT IN ANGLES SNA AND SNB IN LATERAL CEPHALOMETRIC RADIOGRAPHS 4.1 Introduction The Sella turcica is routinely traced for cephalometric analysis. It is defined as a depression in the skull base, where the pituitary gland is situated. Since the introduction of lateral cephalometric radiography by Broadbent in 1931, this radiographic technique has been widely used in orthodontics to evaluate cranial and dentofacial growth (Broadbent, 1931). Cephalometric analyses are based on angular and linear measurements, which might present some errors (Hussels et al., 1984). Nevertheless, it is widely used in orthodontics. One of the major causes of error in cephalometric analyses occurs in the identification of landmarks, moreover certain cephalometric points are more difficult to identify. The Sella (S) point, which is located at the midpoint cavity of the sella turcica, is an example (Proffit et al., 2006). This point is considered to be a floating landmark because it is identified by visual criteria and is not situated on a specific structure. In 1953, Steiner developed a cephalometric analysis, known nowadays as the first of the modern cephalometric analyses. He indicated some craniofacial norms “which expressed the concept of a normal average American child of average age” (Steiner, 1953). An analysis based on dentoalveolar compensatory mechanism was proposed, in order to determine the nature of malocclusion. Steiner created his cephalometric analysis based on the analyses of Downs, Margolis, Riedel, Thompson and Wylie, combined with some of his own 110 cephalometric values. The major influence to his work was from Riedel. He studied several patients’ cephalograms for the relationship of the maxilla to the cranium and mandible. In 1952, Riedel defined the ANB angle, which is based on A and B points - the deepest bony outline points of the maxilla and mandible, respectively. He established his analysis considering the SN line, which refers to the anterior cranium base, and used as reference, angles SNA and SNB to provide information on the upper and lower facial prognatism. He indicated that the arithmetic difference between SNA and SNB would result in the ANB angle. It indicates the magnitude of skeletal-jaw discrepancies, and was the major reference for Steiner, since it is an expression of the dental apical base relationships. ANB angle may vary according to the vertical distance between landmarks N and points A and B (vertical height of the face). If this distance increases, the ANB angle decreases. Incorrect identification or growth may create different positions of point N, and will also affect angle ANB (Proffit et al., 2006). Anteroposterior jaw relation can be determined either by angle ANB or Wits appraisal. Studies have shown that these two methods present some limitations. Angle ANB can vary according to cranial base length and/or jaws rotation, and Wits appraisal can change with the occlusal plane. Therefore, some authors suggested that both methods should be used (Ishikawa et al., 1998). Other factors can affect angle ANB, including: patient’s age (ANB decreases with age), position of point N, SN plane rotation, occlusal plane, and maxillary or facial prognatism (Oktay, 1991). Steiner indicated that, besides knowing where the discrepancy was, the most important factor was to know its magnitude (Proffit et al., 2006). The normal value for the angle ANB in a Caucasian should be of 2º (Proffit et al., 2006). 111 These angles evaluation plays an important role in the diagnosis and treatment of malocclusions. Anteroposterior position of the maxilla in relation to the anterior cranial base is determined by angle SNA. Its standard value is 82 o ± 3º. The SNB defines the anteroposterior position of the mandible, for which the standard value is 79º ± 3º. If the SNA or the SNB is greater or lower than the standard value, this indicates that the mandible or maxilla is either positioned anteriorly or posteriorly to the cranial base. If SNA is greater than 85º, it indicates a maxillary protrusion, and if it is lower than 79º, it reveals a maxillary retrusion. Likewise, if SNB is lower than 76º, it suggests a mandibular retrusion; and if it is greater than 82º, it indicates a mandibular protrusion. This interpretation is only valid if the SN plane is normally inclined to the true horizontal (Po-Or) and the N point position is normal (Proffit et al., 2006). SN plane represents the anterior cranial base. Variability in S landmark identification may modify angles SNA and SNB. On our previous study, we found that landmark S had low intraand inter-observer variability, which was consistent with other studies (Miloro et al., 2013, Oz et al., 2011; Chen et al., 2004). Errors in cephalometric analyses may occur by numerous reasons. One of the most important errors happens due to inconsistent and imprecise landmark identification. Inaccurate landmark identification may lead to erroneous diagnoses and treatment plans for orthodontic cases (Chen et al., 2004; Tng et al., 1994). Moreover, some authors stated that different levels of knowledge and observers background play an important role in landmark identification (Miloro et al., 2013; Kamoen et al., 2001; Gravely and Benzies, 1974; Kvam and Krogstad, 1969). Other authors believe that errors can be caused by different individual conceptions of landmarks’ definitions and its perception, 112 rather than education and training (Chen et al., 2004; Kamoen et al., 2001; Lau et al., 1997). Skeletal landmarks, like points A, B and N, play an important role in patient’s skeletal diagnosis. The aims of this study were to determine intraand inter-observer precision in identification of the landmarks Sella (S), Nasion (N), point A and B, as well as to determine how it can interfere with angular measurements of SNA and SNB by orthodontists and dentomaxillofacial radiologists. 4.2 Materials and Methods Twenty digital lateral cephalometric radiographs were selected from the database at the Oral Imaging Center, University of Leuven. Lateral cephalograms were acquired by positioning the patients in a standard cephalometric device (Veraviewepocs 2D ® , J. Morita, Kyoto, Japan). The exposure values were set at 77 kV and 7.2 mA, with an exposure time of approximately 1.6 s, according to each patient. The radiographs were considered to have good quality. Inclusion criteria were: • No evidence of current orthodontic treatment. • Digital cephalometric image were of good quality to allow the identification of landmarks, and the ruler on the radiograph was clearly visible, allowing calibration of the images in the cephalometric analysis software program. • There were no unerupted or partially erupted incisors that could have compromised landmarks identification. • No gross skeletal asymmetry. 113 All of images selected were then exported in TIFF format and introduced in the PowerPoint software (Version 2010; Microsoft, Redmond, Washington, USA). The Sella-Nasion horizontal plane was used as a reference and all images were orientated accordingly. Analysis A PowerPoint file with 20 lateral cephalometric radiographs was sent by e-mail to ten experienced observers (five orthodontists and five dentomaxillofacial radiologists [DMFR]). Each observer identified the following landmarks on each radiograph: Sella (S), Point A (A), Point B (B) and Nasion (N), by placing a predefined red dot (Figure 4.1). A detailed explanation of the procedure and definitions of the 4 landmarks were given (orally and on paper) to all observers. Thus, observers followed the same landmarks definitions during identification process. The same procedure was repeated 8 weeks after to test the intra-observer variance. 120 landmark’s definition was given by the main author. Furthermore, a low variability in the identification of the point S was found. Depending on the magnitude of the error landmarks identification, patients diagnosis can change. We studied how an imprecise identification of point S could lead to different SNA and SNB angles. According to some authors, landmark identification errors inferior to 1 mm are considered accurate (Chen et al., 2000; Richardson A., 1981). Other authors believe that a difference of 2 mm is considered acceptable and does not have any influence in orthodontic diagnosis and treatment plan (Kamoen et al., 2001; Lau et al., 1997). The variation in identification of the S landmark was relatively low, presenting a deviation of -0.22 mm in the x direction and of 0.28 mm in the y direction. Some cephalometric landmarks are more reliable in either the horizontal or vertical plane (Baumrind and Frantz, 1971). We had previously revealed a low variability for the S landmark. In that previous study, we had suggested that, with a small range of variation (1 to 2 mm) in landmark identification, patient diagnosis could change. Nevertheless, in general, in the present study, intraand inter-observer agreement for the SNA and SNB angles was good (ICC 0.75-0.90). Steiner used angles SNA and SNB for patients’ diagnoses. We revealed that minor changes in S landmark identification could change both SNA and SNB classifications. Dentomaxillofacial radiologists, showed differences in patients diagnosis in 17 cases out of 100 observations. Regarding SNB, mandible position diagnosis was changed in 5 cases. Overall, a higher variability was found amongst orthodontists. Between orthodontists, maxillary position diagnosis was changed in 25 cases, while the mandible’s position diagnosis was changed in 23 cases. Larger 121 variations were found in both groups on the SNA angle. This could happen due to point A identification, which is more difficult to identify than point AB. We should remember that “standard” values for these two angles were defined by a small number of individuals that were supposed to be representative of a population. Due to the inflexible interval given for these angles, patients who deviate just slightly from the standard value may have an erroneous diagnosis. We suggested that changes of 0.50º in SNA and of 0.31º in SNB could alter patients diagnosis concerning mandible and the maxilla positions. An incorrect diagnosis may lead to erroneous orthodontic treatment. To perform a diagnosis and treatment plan in orthodontics, many variables are taken into account; therefore, each step of the process should be performed with maximum accuracy. The results of this study question the validity of cephalometric analysis in orthodontics, since a small variation on these landmarks’ identification can lead to different diagnosis and, thus, different treatment plans. 4.5 Conclusions In conclusion, identification of the Sella landmark revealed a better agreement amongst dentomaxillofacial radiologists. Orthodontists, however, showed a larger variability in S identification and, consequently, the SNA and SNB angles drifted significantly. Small modifications in identifications of the S, N, A and B points may lead to differences in angles SNA and SNB. Therefore, patient diagnosis and treatment can vary. More differences existed regarding SNA than SNB. Further studies on a larger patient sample with inclusion of more borderline cases may be needed to determine the real clinical impact on treatment planning. 122 123 CHAPTER 5. Influence of using lateral cephalometric radiography in orthodontic diagnosis and treatment planning Ana R Durão, Ali Alqerban, Afonso P Ferreira, Reinhilde Jacobs. Influence of Lateral cephalometric radiography in orthodontic diagnosis and treatment planning. Accepted for publication in “Angle Orthodontist” on the 5th of April 2014. Angle Orthodontist, Vol 86, No 4, 2014. DOI.10.2319/011214-41.1. 124 125 INFLUENCE OF USING LATERAL CEPHALOMETRIC RADIOGRAPHY IN ORTHODONTIC DIAGNOSIS AND TREATMENT PLANNING 5.1 Introduction Lateral cephalometric radiography (LCR) is widely used in orthodontic assessment and treatment planning. Despite that, its usefulness in orthodontics remains questionable. Silling et al., in 1979, stressed that lateral cephalometric radiography was only needed for Class II division 1 patients. Later, in 1991, Han et al. stated that patient’s examination together with dental casts provided sufficient information to perform a diagnosis. According to them, only 55% of treatment plans were changed after the LCR evaluation. In the same vein, Bruks et al., in 1999, suggested that in 93% of the cases, treatment plans remained unachanged after the LCR evaluation. They evaluated the patient, dental casts, and extraoral photographs. In contrast Pae et al., in 2001, revealed that in patients with Class II division 2 malocclusion and bimaxillary protrusion, this radiography could change the decision with regard to teeth extraction. In 2008, Nijkamp et al. reinforced that LCR does not seem to have any impact on orthodontic treatment plan for Class II division 1 patients. Recently, in 2011, Deveraux et al. concluded that only in one out of six patients’ orthodontists decide to change their treatment decisions regarding with regard to tooth extraction. In contrast with the previous study, they suggested that LCR may be justified for orthodontic treatment. Considering the controversy in the literature, the present aim was to further explore the impact of additional LCR in orthodontic diagnosis and the treatment planning. 126 5.2 Materials and Methods Forty-three patients with pretreatment diagnostic records were randomly selected. All patients were seeking orthodontic treatment at the Faculty of Dental Medicine of the University of Porto. The study was approved by the Ethics Committee of the Faculty of Dental Medicine of the University of Porto (900079). The patients’ ages ranged from ten to 42 years-old (24 female and 19 male). Orthodontic diagnostic records included: three photographs of the angle of trimmed dental casts, digital lateral cephalometric and panoramic radiographs, as well as standard clinical photographs comprising seven intraand four extraoral pictures (Figure 5.1). The patients’ identification was blurred to avoid recognition. All the blinded information was saved in a PDF file and recorded in a compact disk and given to each observer. Ten qualified orthodontists were involved in this study, with experiences ranging from five to 24 years. Patients’ records were evaluated during two sessions. The time interval between observations was at least eight weeks. At the first session orthodontists evaluated records without LCR. In the second session the same information was presented, but this time the LCR was added. Between the two sessions the order in which the cases were presented was altered to avoid bias. The evaluation process for the two sessions involved the use of a questionnaire concerning the diagnosis and treatment plan; the questionnaire contained the following questions: 127 1. Skeletal relationship: neutro-, disto-, or mesio-relation? 2. Angle classification of the occlusion based on molar relationships: Class I, Class II or Class III 3. Detection of any abnormality? 4. The treatment plan will be: orthopedic growth modification; orthognathic surgery; dentoalveolar compensation? 5. Is there enough space for all teeth to erupt? 6. Would you extract any teeth in this patient? If yes, which one? 7. Would you expand the upper arch? 8. Would you use anchorage in the maxilla, mandible, or both? 9. Do you expect any complications during the treatment? 10. How long do you expect the treatment to last? 11. Would you need any additional information to make a decision? Which information? 12. How long has it been since you were qualified as an orthodontist? 128 Figure 5.1. Example of the information given to orthodontists. Statistical analysis The percentage of agreement of the answers between the two sessions (ratio of agreement between cases and total cases used) was evaluated. This was carried 129 out for each patient to test for differences in the percentages of changed decisions regarding diagnosis and treatment planning. 5.3 Results The percentage of agreement between sessions was lower with regard to diagnosis than it was with regard to treatment planning (Table 5.1). Treatment planning seemed to be changed, on average, in 36% of the cases by adding a LCR. In addition, the skeletal classification diagnosis was changed, on average, in 56% of the cases, and, in general, in 52% of the cases the malocclusion classification seemed to be altered. The most frequent changes appeared in Class II malocclusion patients. With regard to skeletal classification, the least experienced observer was the least consistent (28%), while the more experienced observer was the more reliable (67%). On average, ten cases were classified in the first session as Class II, and after evaluating the LCR the diagnosis of the skeletal classification changed to Class I. In nine cases skeletal classification was altered from Class I to Class II. Overall only in a single case did the orthodontists change from Class III to Class I. The presence of abnormality revealed a very good agreement between the two sessions (overall 87%). With regard to treatment modalities, in general there was an agreement of 64%. The most experienced observer revealed 80% of agreement between sessions, changing the treatment plan in only 8 cases, while the lower percentage was of 37%, seen in an observer with ten years of practice. In 26 cases the treatment modality was changed in the majority of cases, being altered from