INTRODUCTION

The author discovered an enlarged sella turcica (Figure 1) on a lateral cephalogram of a 29-year-old general dentist. A thyroid-stimulating hormone-secreting pituitary macroadenoma was eventually removed. If this pathology had not been discovered, it would have potentially had a severe deleterious impact on the life and career of this young general dentist. The discovery prompted me to question whether other orthodontists had identified any significant pathologies on lateral cephalograms. A review of the literature did not reveal the frequency of practicing orthodontists who had discovered pathologies on cephalograms.1–12

View Figure

• Download as Powerpoint
• Download Figure

Dentists and orthodontists have been concerned about the radiation exposure from radiographs.13–20 In addition, there have been questions about orthodontists' level of responsibility in discovering incidental abnormalities on radiographs.21–28 Thus, the frequency of pathologies on lateral cephalograms that have been discovered by practicing orthodontists could be a factor in the decision to obtain those images. Therefore, the purpose of the present study was to determine the frequency of significant pathologies on lateral cephalograms discovered by a sample of orthodontists during their careers.

MATERIALS AND METHODS

A brief survey was sent to a sample population of orthodontists. The data collected from the respondents were analyzed and interpreted.

Survey Design

The survey consisted of five statements on the back of a self-addressed and stamped postcard (Figure 2). The survey was designed to elucidate the number of years the orthodontist had been in private practice; the number of patients diagnosed or the number of cephalograms reviewed; whether significant pathologies had been observed; and, if so, how many and what types. The survey was kept succinct to encourage a response, yet provide the needed information. The study relied on simple recall by the orthodontists of significant and meaningful pathologies discovered in their careers.

View Figure

• Download as Powerpoint
• Download Figure

The term “significant pathology” was purposefully not defined. The decision was made to not constrain the orthodontists but to allow their decision on reporting pathologies based on the specific circumstances. The cover memorandum stated, “The survey concerns your career experience of discovery of head/neck pathologies on routine lateral cephalometric radiographs (ie neural, optic, glandular, osseous, vascular and etc.).” This was intended to provide guidance for the respondents. Therefore, “significant pathology” was one that could affect a patient's life and well-being as determined by the respondent.

RESULTS

Of the 417 surveys that were mailed, 201 were returned for a 48% response rate. The respondents represented a total of 5548 years of clinical practice, and they had observed lateral cephalograms on 1,325,338 patients. Of the 201 surveys received, 102 respondents had not observed pathologies, and 99 respondents (49.3%) had observed significant pathologies during their orthodontic careers.

Table 1 summarizes the responses; the mean number of years in practice for the sample was 27.6 years and the mean number of patients diagnosed was 6594. The shortest time in practice was 5 years, and the longest was 54 years. The lowest number of patients diagnosed was 250 and the highest was 25,000. The respondent with 25,000 patients voluntarily identified himself as an educator and included the patients of residents whom he had supervised. Even though the survey did not differentiate between private practitioners and full-time educators, it could be assumed that large numbers of observed patients were probably in an educational institution.

Table 1 Summary of Responses

View Fullscreen

The data may be used to estimate the discovery rate for significant pathologies on lateral cephalograms. Within the average career of 6600 patients observed, it would be anticipated that an orthodontist would observe 1.33 patients with significant pathologies. The 95% confidence interval (CI) would be between 1.18 to 1.50 patients per 6600 total patients observed.

If the sample were arbitrarily segregated by years in practice, the discovery rate for each group is as follows: 41% for 20 years or less, 48% for 21 to 30 years, and 56% for 31 years or more (Table 2). To account for the influence of the increased numbers of patients observed by the older orthodontists, the Poisson model1 was used, and the number of patients diagnosed by an orthodontist was the exposure variable. The number of expected patients with significant pathologies observed actually decreased as the number of years in practice increased: 20 years or less was 1.83 (95% CI  =  1.41–2.37), 21 to 30 years was 1.40 (95% CI  =  1.15–1.69), and 31 years or more was 1.12 (95% CI  =  0.93–1.36) (Table 2).

Table 2 Discovered Pathologies by Grouped Years of Practice

View Fullscreen

There were 268 significant pathologies listed, and these had been observed by 99 respondents. The pathologies were of many different specific diagnoses. Consequently, the specific pathologies were organized into generalized tissue groups. Table 3 displays the seven generalized tissue groups with the number of pathologies and the percentage of each within the sample.

Table 3 General Tissue Types of the 268 Pathologies Reported

View Fullscreen

DISCUSSION

Of the 201 responding orthodontists, 49.3% reported finding a significant pathology on a lateral cephalogram during their orthodontic careers. When the literature was reviewed, no specific studies could be found concerning the frequency of practicing orthodontists observing radiographic pathologies. Respondents were asked to recall significant pathologies that they had observed. The assumption was that a significant pathology, which might affect the life of a patient, would likely be remembered by an orthodontist. The finding of this study suggests that about 50% of orthodontists would likely discover a significant, potentially life-affecting pathology on a lateral cephalogram during their career.

A total of 268 patients of an estimated 1,325,338 had significant pathologies discovered on lateral cephalograms. This is a 0.02% incidence of significant pathologies in this particular orthodontic population. Nanda et al.1 reported on four patients who had unsuspected abnormalities with major medical significance that had been discovered on lateral cephalograms. Bick and Lee2 reviewed lateral cephalograms of 513 consecutive patients and found that 3.5% had abnormalities or pathologies. During the 1982 Pacific Coast Society of Orthodontists meeting, Buck3 reported on 465 consecutive University of Oregon patients, 22% of whom had some radiographic pathology. Kantor and Norton4 published an extensive review of clinically healthy anatomy seen on cephalometric films and common anomalies that could be seen. They encouraged orthodontists to consider radiographs as skull films first and to look for pathosis before any morphometric analysis.

Kuhlberg and Norton5 published an article where oral radiologists reviewed radiographic images taken by the University of Connecticut Orthodontic Department on 396 consecutive patients. They reported remarkable findings in 6.2% of those patients. An earlier study from the University of Connecticut by Tetradis and Kantor6 studied various radiographic images of 325 consecutive orthodontic patients. They discovered 431 incidental pathologies, but only 6 pathologies were significant enough to report to a physician for care. This produced a 1% incidence of significant pathologies in that study. In 2008, Abdel-Kader7 found 22 (1.6%) serious incidental findings in the pretreatment radiographs of 1354 patients.

A recent study using three-dimensional cone-beam volumetric scans made on 500 consecutive patients at the University of Southern California School of Dentistry revealed that 25% of the patients had some incidental pathological findings on radiographic images of the head and neck.8 Thus, the recent orthodontic literature contains reports on the frequency of pathologies on consecutively observed patients in academic environments. Fortunately, the 0.02% frequency of discovery in this study is low compared to previous studies. Factors that might have influenced the low discovery rate in this17 study include the younger orthodontic patient population, the use of oral and maxillofacial radiologists to analyze cephalograms in most studies, the reporting of any deviations from normal rather than those that have a potential life-affecting significance, and the fact that some orthodontists could have been more focused on the morphometric aspects of the lateral cephalograms than on observing incidental findings.

The average orthodontic career length of the sample was 27.6 years in practice. The average total number of patients seen (observed cephalograms) was 6593. Therefore, the survey sampled practitioners who had observed a large number of patient cephalograms. The sample was arbitrarily grouped by years in practice, and those groups reflected the logical assumption that as the years and numbers of patients increases, the number of orthodontists who discovered significant pathologies would increase (Table 2). Those who had been in practice 20 years or less had a 41% discovery rate, those who had been in practice 21 to 30 years had a 48% discovery rate, and those who had been in practice 31 or more years had a 56% discovery rate. If the number of patients observed in a career is evened between the groups using the statistical Poisson model,29 the youngest group of orthodontists had a greater likelihood to discover significant pathology on a cephalogram in an average career patient population of 6600 than their older colleagues (Table 2.) Respondents who had been in practice 20 years or less had a 1.83 anticipated mean discovery, those who had been in practice 21 to 30 years had a 1.40 anticipated mean discovery, and those who had been in practice 31 years or more had a 1.12 anticipated mean discovery of significant pathologies.

The reason for this finding may be due to one of the following: (1) younger orthodontists with fewer patients might spend more time analyzing the cephalograms; (2) perhaps younger orthodontists do not delegate the tracing of cephalometric radiographs to staff; (3) older orthodontists may use computerized imaging systems that use anatomic points, which could affect the analysis of the image; or (4) mature orthodontists may not recall all discoveries in their career. Whatever the explanation, further study might reveal an answer.

There were a total of 268 significant pathologies reported with many different specific diagnoses. Therefore, pathologies were divided into generalized groups based on tissue types and phenomena. The musculoskeletal group was the largest at 38%. Various types of bone cysts predominated in this group. The dental group represented 12% of pathologies, of which 11 patients had an ameloblastoma. The most interesting group was the foreign-body group, which represented 5% of the total. Examples of foreign-body observations included a sewing needle in the ear, an earring embedded in the nasal cavity, and a BB pellet lodged behind the eyeball. The vascular group was the most surprising at only 1% of the total observed pathologies. One would anticipate that more carotid calcifications would be observed on a cephalogram (only four were reported). However, the younger orthodontic patient population would likely reduce that observation compared with the entire population. The glandular group made up 28% of the pathologies and contained 67 patients with an enlarged sella turcica and pituitary adenoma. The recent literature did contain publications of case reports9–12 and studies of sella turcica morphology.30,31

The impact of this study's findings, along with previous findings in the literature,1–12,22 certainly supports the position that all orthodontists who take lateral cephalograms should approach each image as a skull film and observe the entire head and neck. Then the film may be used as a source of morphometric measurements to assist in orthodontic diagnosis.

The efficacy of obtaining routine cephalometric and dental radiographs has been a concern of the orthodontic specialty and the dental profession. The U.S. Food and Drug Administration's guidelines to lessen the dental radiation exposure to the public was confirmed and supported by a 1995 study published in the Journal of the American Dental Association.13 The British Orthodontic Society published guidelines for radiographs used in orthodontic diagnosis,14 and they were discussed by the editor of the American Journal of Orthodontics and Dentofacial Orthodedics (AJODO)15 and later criticized by a letter to the editor.16 A study published in 2008 concerning radiation organ dosage from routine orthodontic care concluded that it does increase the radiation burden of those patients.17

An article that used a phantom-head to measure radiation dosage to glands in the head and neck region, concluded that lateral and PA (posterior-anterior) head films delivered levels of radiation much lower than standard dental radiographs.18 A follow-up study by the same authors stated that the responsible use of cephalometric radiographics adds very little to the patient's carcinogenic burden.19 Another phantom-head study by Freeman and Brand20 found that the radiation exposure from a lateral cephalogram to the eye and the parotid and thyroid glands was far less than from a panoramic and bitewing series. A radiation burden from the routine lateral cephalometric head film is obviously present. The ultimate degree of consequence to the patient is still ambiguous within the literature. Therefore, it is important that the orthodontist obtain radiographic images when there is a need for those images to help achieve an appropriate diagnosis or to evaluate posttreatment changes.

Recently in the orthodontic literature there has been a dialogue concerning the responsibility of the observer of a radiographic image to detect incidental findings. The topic of observer responsibility has come to the forefront with the advent of cone-beam computed tomography (CBCT) with editorials24,25 and letters to the AJODO editor.26,27 Also, there have been several recent articles that address the issue of the radiographic observer's responsibility to detect incidental findings.7,8,26–28

The intention of this study is not to debate the efficacy of cephalometric radiographs in orthodontics. It is to provide information on the frequency of orthodontists discovering significant, potentially life-affecting head and neck pathologies on lateral cephalograms. The point can be made that if a cephalogram had not been taken, any existing pathology could have been undetected. This study should contribute additional information to the debate by the orthodontic specialty on the issue of appropriate use of radiographs in orthodontics. Additionally, this study reinforces the responsibility of each orthodontist to personally and thoroughly evaluate each radiographic image that they obtain on a patient.

As with so many studies, many questions arise that require further study. Examples of questions that could be explored are as follows: (1) What percentage of orthodontists delegate tracings? (2) What percentage of orthodontists use computer-generated tracings from anatomic points placed on the digital head film? (3) How often do orthodontists personally review the radiographic images they prescribe? (4) At what stage within a practice career did an orthodontist discover a pathology?

CONCLUSIONS

• Of the responding orthodontists in this sample, 49.3% had discovered one or more potentially life-affecting pathologies on lateral cephalograms of their patients during their careers to date. This would imply that about 50% of all orthodontists would likely discover a significant pathology at some point in their career.

• Only a 0.02% frequency of significant pathologies was discovered in the more than 1.3 million orthodontic patient population of this study's sample.

• Even though the frequency of discovering a radiographic pathology becomes greater with increasing numbers of patients, statistical allowance for the increased numbers reveals that a younger group of orthodontists would have a higher rate of discovery than an older group of orthodontists.

• Musculoskeletal pathologies were the most common grouping of pathologies observed. The glandular group was second; in particular, an enlarged sella turcica from pituitary adenoma was the most common glandular tissue pathology.