An algorithm for ordering pretreatment orthodontic radiographs Kathryn A. Atchison, DDS, MPH, ~ Larry S. Luke, DDS, MS, b and Stuart C. White, DDS, PhD o
Los Angeles, Ca~if. A study was conducted to identify selection criteria for ordering pretreatment orthodontic radiographs. Thirty-nine orthodontists evaluated six test cases. They provided information on the rationale for ordering each specific radiograph and the impact of the radiograph on the diagnosis and treatment plan. Skeletal relationshi p of the jaws was the most common indication for a radiograph request, followed by root formation/length and molar position or development. Of the radiographs, 16% produced a change in diagnosis, and 20% produced a change in treatment plans. The criteria specified for the radiographs, their impact, and relevant information in the literature were used to develop a n algorithm or set of decision rules that, when tested on the six test cases, resulted in a 36% reduction in the total number of radiographs. (AM J ORTHOD DENTOFACORTHOP 1992;102: 29-44.)
I d e a l l y , each time an orthodontist examines a new patient, he or she decides which radiographs are needed to complement the clinical information. The decision to order a particular radiographic examination may be influenced by several factors. A specific diagnostic examination may be ordered because a clinician wishes to perform a very thorough examination or believes there is a possibility of disease. This may be done to meet the standards of care for the community or to protect the orthodontist medicolegally. These factors are tempered, however, by the desire to expose the patient to the least amount of radiation necessary. Competing factors such as these interject a degree of difficulty to the decision of which or how many radiographs are appropriate as part of a diagnostic examination. In an effort to assist clinicians in making these difficult decisions regarding the use of diagnostic imaging, selection criteria have been developed for a variety of medical and dental radiographic examinations. 1"1~Selection criteria, or high-yield criteria, are guidelines or decision rules that guide the clinician in the choice of radiographs. Specifically, selection criteria are those signs, symptoms, or historic findings, which, if present, indicate there is a reasonable probability that a radiograph will be of value to the clinician in managing the care of a patient. The goal of developing selection criteria is to identify the optimal clinical circumstances for ordering radiographs necessary for diagnosis and
From the School of Dentistry, University of California, Los Angeles. *Assistant Professor, Section of Public Health Dentistry. ~'Clinieal Professor, Sections of Pediatric Dentistry and Orthodontics. *Professor, Section of Oral Radiol~y. 8/1/26113
treatment planning and to balance the clinicians diagnostic needs and desire to expose the patient to as little unnecessary radiation as possible. The value of selection criteria have been well documented. AS an example, in a classic study of skull radiographs obtained to evaluate a traumatic head injury, Bell and Loop t found that the use of clinical criteria for ordering radiographs significantly improved the percent of positive examinations. Of the adults admitted to the emergency room for head trauma, 58% had skull radiographs. However, evidence of fracture was seen in only 1.7%. When patients were assigned to high- or low-risk groups on the basis of clinical criteria, the detection of positive findings became 1 in 11.5 patients for the high-risk group and 1 in 435 patients for the low-risk group. A subsequent study demonstrated a 40% reduction in the number of skull radiographs ordered after adoption of ordering by clinical criteria. 2.3 Two studies have been conducted concerning the use of clinical criteria for ordering periapical radiographs for dental diagnosis. 9"t~Both demonstrated significant reductions in the number of radiographs ordered when clinical criteria were followed. A set of guidelines adopted by the American Dental Association tt.tz were recently developed that specify selection criteria for ordering general dental radiographs for new and recall patients. The American Association of Orthodontics (AAO) prepared recommendations that the appropriate radiograph examination for all patients initiating orihodontic treatment consist of a minimum of a lateral cephalogram, a panogram, and a complete-mouth set of periapicals and bitewings, t3 Atchison et al.~4 in their study of the ordering pat29
30
Atchison, Luke, and White .
terns of orthodontists on simulated cases, reported that about three-fourths of the radiographs ordered did not contribute original information to the orthodontist's certainty or content of his diagnosis and treatment plan. Further, it was found that the number of radiographs ordered by the orthodontists showed a stronger relationship to the orthodontist ordering the films than to the patient under consideration. These findings suggest the need for a detailed examination of the individual radiographic examinations ordered during the course of a comprehensive orthodontic patient examination. The purpose of this article is to identify the selection criteria for ordering orthodontic radiographs and then to organize these criteria into a set of easily followed decision rules or an algorithm. Because there are a number of radiographic examinations used for an orthodontic examination, one must consider not only the information influencing the decision to order a radiograph but also which radiograph best provides this information.
MATERIALS AND METHODS The design for this study has been previously described. ~4Briefly, 39 orthodontists, members of the Pacific Coast Society of Orthodontists and/or faculty at dental schools, were each interviewed for approximately 2 hours during which the participant evaluated six patient cases selected from the postdoctoral orthodontic program at the UCLA School of Dentistry. The records were duplicated maintaining patient confidentiality. The records included extraoral and intraoral slides, trimmed study models, and a patient history and examination summary. Composites of the extraoral photographs, study models, and a cephalometric tracing for each of the six cases are included in Figs. 1 through 6. The orthodontist was shown the case records and asked to formulate a diagnosis (enumerate the problems presented), which was recorded. The process was repeated for the treatment plan. The orthodontist was then told that he could request a radiograph if needed to complete his diagnosis or treatment plan. Each radiograph was requested individually and evaluated before another radiograph was requested. The complete-mouth series was given whenever a portion of it was requested. The Ricketts, Downs, Steiner, Wits, Harvold, and Sassouni analyses were presented along with a traced lateral cephalogram. The information sought from the radiograph was recorded. Once the respondent had examined the radiograph, he was asked whether the radiograph caused a change or a confirmation of the diagnosis. Any reSulting changes were noted. Again, the same questions were asked about the treatment plan. Radiographs were
Am. J. Orthod. Dentofac. Orthop. July 1992
selected until the orthodontist was confident he had all the diagnostic information he needed. Case A (Fig. 1), because of its relative simplicity, was used as the first case for each participant. The order of the remaining cases was randomized to prevent differences among the cases resulting from fatigue. Information was obtained regarding the number and type of radiographs ordered for the patients. Data was also collected regarding the information sought by the orthodontist from each radiograph for diagnosis and treatment planning, and the impact of the radiograph on the diagnosis and treatment plan. The impact of the radiograph was the self-reported assessment of the information provided by the radiograph on the diagnosis and treatment plan. Thus, a radiograph could provide information that changed the diagnosis or treatment plan, confirmed information provided by the clinical records or another radiograph, or had no effect on the diagnosis or treatment plan. Descriptive statistics were computed for each variable. Responses were grouped by orthodontist, test case, and radiograph. On the basis of the information requested by the participants and the impacts of each radiograph, a set of decision rules (an algorithm) was developed and tested for each case.
RESULTS The information sought by participants Thirty-nine participants completed the interview process, for a total of 234 case evaluations. Table I shows the information that participants reported needing in an open-ended format to complete a diagnosis and treatment plan for the six test cases, as well as the radiographs they requested to provide it. Skeletal relationship of the jaws was the most common indication for a radiograph, cited 216 times or 14.4% of all reasons for ordering a radiograph. Eight other indications each accounted for more than 5% of the requests for information: root formation or length, molar position and development, incisor angulation, general pathosis, presence of teeth, tooth shape, size and position, mandibular plane angle, and periodontal problems. Three of the indications, incisor angulation, skeletal relationship, and mandibular plane angle, involve information provided primarily by a lateral cephalogram. The lateral cephalogram alone accounts for 34% of the information sought by the orthodontists. The other commonly requested information was provided by the panogram or complete-mouth series examinations. The complete-mouth series and the panogram together provided about 51% of the information, but critical examination of the indications for them demonstrate con-
Volume 102 Number 1
An algorithm for ordering pretreatment radiographs
Fig. 1. Facial photographs, cephalometric tracing, and orthodontic models of patient A.
31
32 Atchison, Luke, and White
Am. J. Orthod. Dentofac. Orthop. July 1992
Table I. Information needs reported by orthodontists (11 = 39) when ordering radiographs to complete
diagnostic process for six test cases, by radiograph Itformation sought Incisor angulation Soft tissue profile Mandibular plane angle Facial type (pattern) Skeletal relationship of jaws Airway Skeletal asymmetry Midline deviation Condylar shape and position Osteoarthritis Molar position and development Tooth size, shape, position Presence of teeth Eruption sequence General pathosis Periodontal problems Root formation/ lengdl Caries Sutures Growth/skeletal age Other
tero, Cephalogram
Cephalogranl
AFI FMX
Panogrant
T~U
SMV
Occlusal
I Corpo,I index
127 34 91
127 34 91 17 202 17 2
17 216
14
4
7 36 14 1
3
5
I 4 19
25 44 14 67
1
43
14 144
14
3
I
2
2
17
119
38
59
4
104
26 7 42 63
79 14 60 24
I
106 22 121 87
93
54
59
9
13
I
147
13 11
4
5
Total
8
2
siderable overlap. The complete-mouth series was generally ordered for tasks requiring high image resolution, such as evaluation of periodontal problems, root formation or length, and dental caries. Participants requested a panogram for tasks where a broad view of the jaws was desirable, such as evaluation of molar position and development, tooth size, shape and position, and the presence of teeth or eruption problems. Impact of the radiograph on the diagnosis and treatment plan The participant determined the contribution of each radiograph toward the development of the diagnosis and treatment plan. Each radiograph could influence a diagnosis in one of three ways. A positive yield was defined as a radiograph that the participant stated caused a change to the diagnosis. A neutral yield was one that the participant stated confirmed diagnostic information already available, either as a result of the clinical information or from a previous radiograph. A radiograph was rated as having no effect if the participant reported it contributed no information to the diagnosis. The effect of the radiographs on diagnosis is shown in Table II. There were 120 radiographs (16%) that produced positive yields. These radiographs provided
2
I
2
68 13 15 20
359 information changes to the diagnosis (one radiograph could provide more than one piece of information). In terms of formulating a diagnosis, the radiographs were found to be most useful in the assessment of skeletal problems (138 information changes), and in the assessment of molar problems, including crowding, missing teeth, or impactions (85 information changes). Radiographs were also demonstrated to have beneficial effects on the development of the treatment plan. Table II shows that 148 radiographs, 20% of all radiographs ordered, produced changes to the treatment plan. The largest single category of changes dealt with molar, premolar, or other extraction decisions (123 treatment changes or 41%). This is not surprising since molar crowding was evident in three test cases. Ordering patterns observed in the study To provide the information needed, participants requested 741 radiographs, of which 192 (26%) produced a positive yield on diagnosis or treatment plan. To comprehensively examine the influence of practitioner versus patient characteristics on the number and type of radiographs ordered for the examination of a patient, the ordering patterns demonstrated by the 39 participants for the six test cases were examined. Table III
An algorithm for ordering pretreatnlent radiographs 33
Volume 102 Number 1
Fig. 2. Facial photographs, cephalometric tracing, and orthodontic models of patient B.
compares the distribution of the radiographs by test case, as well as the number of radiographs that produced a positive yield. It shows that the lateral cephalogram, panogram, and the complete-mouth series were ordered
by a majority of the practitioners (the denominator) for each test case. Less than half of the participants for any specific case ordered any of the other radiographs. Variability across case is noted, particularly in the number
34
Atchison, Luke, and White
Am. J. Orthod. Dcnt~zc'. Orthop. July 1992
t \. !
9
.
Fig. 3. Facial photographs, cephalometric tracing, and orthodontic models of patient C.
Volume 102
Number l
of times the posteroanterior cephalometric and TMJ radiographs were ordered for one case compared with another. Variability is also noted in the impact of the radiographs from one case to another. Although the number of participants who ordered a lateral cephalogram did not vary substantially across case, the number of lateral cephalograms that produced a positive yield did, ranging from six productive lateral cephalograms for case A or E to a high of 22 productive lateral cephalograms for case F. A similar range of productivity was found for the complete-mouth series examination.
Development of an algorithm for ordering orthodontic radiographs On the basis of the information needs reported by the participants, as well as the impact ofeach radiograph on the diagnosis and treatment plan, an algorithm was developed for ordering orthodontic radiographs (Fig. 7). Certain basic assumptions were made in developing the algorithm. 1. A radiograph is indicated only if the patient exhibits a clinical characteristic suggestive of a problem requiring radiographic examination or if there is a treatment decision that requires a radiograph. ~' The probability of occult disease being detected in the absence of any clinical sign or symptom is so low as to contraindicate radiographic examination solely on this basis. ~ 2. The algorithm described has been considered exclusively for prctreatment orthodontic radiographic examinations. 3. The algorithm is based on evaluation of patient cases that exhibit orthodontic problems and that would eventually require comprehensive (usually fully banded and/or bonded appliance) treatment. Simple tooth movement cases were not considered and may not require a lateral cephalogram radiograph. We propose in this algorithm that a patient should first be examined and a decision made whether the patient is a candidate for comprehensive orthodontic treatment. If so, the patient will require a lateral cephalogram. If the patient is heavily restored or exhibits periodontal problems, then a complete-mouth series is indicated to allow for examination of changes in alveolar bone and teeth. If not, a panogram, anterior periapicals, and posterior bitewings are indicated. If the patient displays a severe facial asymmetry, a posteroanterior cephalogram should be requested to assess whether the asymmetry is dental or skeletal, as well as the extent of asymmetry. Significant signs or symptoms of TMJ problems indicate the need for laterally cor-
An algorithm for ordering pretreatment radiographs
35
Table IIA. Impact of the 741 radiographs on orthodontist diagnoses and treatment plans Impact on diagnosis Radi~raphs with: Positive yield Neutral yield No effect
hnpaet on treatment plan 120 544 77
Radiographs with: Positive yield Neutral yield No effect
148 471 122
Table liB. Information changes
No. of changes
Diagnosis Skeletal problem Molar problem Caries or periodontal problem Incisor problem TMJ, condyle problem Root formation Profile considerations Eruption Midline problem Nonanterior crowding Other TOTAL
138 85 40 28 25 9 11 5 7 28 50 359
Treatment plan Molar extraction Premolar extraction Other extraction Functional appliance or headgear Treatment timing Ortbognathie surgery Treat nonextraction Expand arch Anterior bite correction Refer Other TOTAL
64 39 20 31 16 32 9 3 3 13 71 301
rected TMJ tomograms to evaluate for osseous or positional abnormalities in the joints. Finally, if the patient is thought to be at the end stages of growth potential and significant growth is needed for treatment success, or if orthognathic surgery is contemplated and the patient may still have growth potential, a carpal index should be requested. Table IV shows that if this algorithm had been followed there would have been a 36% reduction in the number of radiographs ordered for the six test cases. The extent of missed information had the algorithm been used was then determined for each test case. If no radiograph was indicated by the patient's characteristics, but one was ordered, the impact of the radiograph
36
Atchison, Luke, and White
Am. J. Orthod. Dentofac. Orthop. July 1992
Table Ill. Comparison of the number of radiographs with a positive yield* and the number and type of
radiographs ordered for each test case Number of changeslmtmber of radiographs ordered by patient test case Radiograph
a
Lateral cephal~ram Panogram Full-mouth series Posterioanterior cephalogram TMJ r a d i ~ m p h s Maxillary and mandibular occlusal Carpal index Submental vertex
6/39 8/32 2/30 1/4 014 0/l 1/4 0/1
I
I 13138 17/32 6/26 3/15 4115 0/3 0/3 0/0
C I 1/39 9/33 2/28 0/6 1/8 1/4 0/3 0/0
I
O 13139 9/29 10/29 4/11 l / 15 0/6 0/1 !/2
I
I 6/39 7/21 12134 4/12 4/10 0/7 0/0 0/0
TOTAL
r 22136 9/31 3/23 2/11 0/6 0/7 0/2 0/0
I
Total 71/230 59/178 35/170 14159 10158 1/28 1 / 13 1/3 1921741
*A positive yield was defined as a change produced in the diagnosis or treatment plan resulting from a radi~raph.
Table IV. Number and type of radiographs eliminated when ordering pretreatment orthodontic radiographs on the basis of the algorithm
Radiograph
Reduction in number
Reduction (%)
Lateral cephal~ram Pan.ram Complete-mouth series* P-A cephalogram TMJ radiographs Occlusals Carpal index TOTAL
0 - 22 -92 - 59 -58 - 28 - 13 -- 265
0 12 54 100 1130 100 1013 35.8
*When reducing the FMX, only posterior periapicals were removed because the algorithm calls for anterior periapicals and posterior bitewings to accompany all p a n . r a m s .
was checked to determine whether the orthodontist thought it provided original information that changed either the diagnosis or the treatment plan. The elimination of films not indicated by the algorithm would have failed to account for 21 changes or 8% of the 268 changes which were made to either the diagnosis.or the treatment plan. The changes not accounted for by using the algorithm are as follows: Patient A. One participant ordered a carpal index for a 12-year-old boy and added a functional appliance to speed up treatment. Although the carpal index showed considerable growth remaining, this could have been predicted by the child's age. Another participant ordered a posteroanterior cephalogram, determined that the midline was deviated 1 mm skeletally, and added a functional appliance to correct it. Patient B. Three participants ordered posteroanterior cephalograms that changed their diagnosis or ireat-
ment plans. One thought the posteroanterior cephalogram indicated a dental rather than a skeletal deviation of the midline and chose an asymmetric extraction pattern. Another orthodontist decided to extract third molars, use a splint, and perform a surgical visual treatment objective (VTO). A third was less concerned about the skeletal asymmetry after reviewing the posteroanterior cephalogram. Thus, for two participants, the posteroanterior cephalogram confirmed the impression that little skeletal asymmetry was present, and the other participant decided on third molar removal and the securing of more diagnostic records. The examination of lateral TMJ tomograms produced four changes. One orthodontist observed a skeletal problem and, hence, prescribed a surgical approach. A second observed a TMJ problem that he determined required treatment before commencement of orthodontic treatment. A third observed an anterior position of the condyle. And the fourth orthodontist suggested the patient may need splint therapy. The history for this patient stated that he had a click but no pain or limitation in movement. Patient C. One change was not accounted for by the radiographs recommended for patient C. One participant requested TMJ radiographs to check a possible condylar problem he observed on the lateral cephalogram and decided the problem did not exist after all. Patient D. Five changes resulted from radiographs not clinically indicated for patient D. Four changes occurred as a result of posteroanterior cephalograms. One orthodontist recommended the patient for an airway evaluation and counseled about surgery. Two participants observed an asymmetric mandible that they thought required surgery. The fourth noted an asymmetric mandible but did not choose to treat it. The addition of lateral TMJ radiographs led one participant to suggest a TMJ splint before surgery. Patient E. There were four changes reported by the
Volume 102
Number I
Air algorithm for ordering pretreatntent radiographs
"t
,
4
Fig. 4. Facial photographs, cephalometric tracing, and orthodontic models of patient D.
37
38
Atchison, Luke, and White
orthodontists on the basis of the posteroanterior cephalogram. The first noted an occluded airway, although the patient had no history of mouth breathing problems. Three clinicians were concerned about a possible skeletal asymmetry. Two of the three ruled asymmetry out on the basis of the radiograph, and one affirmed it, also on the basis of the posteroanterior cephalogram. Patient F. Two changes were not accounted for with the use of the algorithm, both resulting from the posteroanterior cephalogram. One participant recommended by an ear, nose and throat specialist. The other decided that the midline deviation was not skeletal.
DISCUSSION Although it is clear that the orthodontic radiographs provided substantial information to the participants, the study demonstrated considerable variation and inconsistency in the ordering of pretreatment orthodontic radiographs. These practice variations, in the face of the known biologic risk of ionizing radiation, suggests the need for guidelines for ordering orthodontic radiographs beyond those prepared by the AAO." The information needs listed in Table I, as well as the impact of the radiographs requested by the 39 orthodontists for six test cases, were used to develop criteria for ordering radiographs based on patient characteristics. These criteria were then organized into an algorithm that guides the orthodontist through the process of selecting appropriate radiographs for a patient. All criteria for requesting radiographs were developed by the orthodontists. We did not place a value on them, rather we used their self-reported data to determine which were productive criteria for developing the algorithm. In addition, current use of radiographs and the available literature were examined to help define relevant reasons to requestradiographs. It is recommended that previous radiographs should be obtained and used when suitable before ordering additional radiographs. Each of the radiographs considered will be discussed in depth.
Lateral cephalogram In addition to the high frequency of clinical indications noted in Table I, the lateral cephalogram is commonly used for comparison with standards, growth forecasting, and the development of VTO (especially if surgery is involved): Because of the broad nature of the information that a lateral cephalogram provides, its usefulness in treatment visualization and planning, a n d the impact that it demonstrated in changing the diagnosis and treatment plan for these six patient cases, the requesting of a pretreatment lateral cephalogram is sup-
Am. J. Orthod. Dentofac. Orthop. July 1992
ported by this study for all patients needing comprehensive orthodontic treatment.
Complete-mouth periapicals and bitewlngs versus panogram On the basis of the orthodontist's preference noted in Table I, the complete-mouth series is recommended if the patient is heavily restored or if periodontal problems, or root length, shape, and stage of formation questions need answering. If these situations do not exist, a panogram, posterior bitewings, and anterior periapicals are recommended to examine unerupted teeth and their development, eruption sequence, caries, and general oral pathosis. The advantage of this combination is that it incorporates the overall visualization provided by the panogram without forfeiting the detail necessary to examine caries, bone height, and root formation, particularly in the anterior region. This study supports the use of either a panogram or a complete-mouth series for every patient requiring comprehensive orthodontic treatment. What cannot be supported is the request of both a panogram and a complete-mouth series for the same patient. On the basis ofAtchison et al. ~~and the work of White and Forsythe, 8 Ignelzi et al., l~ and Kantor and Slome, 16 there is no rationale to order a panogram and a complete-mouth series together before interpretation of the radiographs. However, additional radiographs may occasionally be indicated on the basis of radiographic interpretation, to rule out disease observed in a radiograph in individual cases.
Posteroanterior cephalogram The posteroanterior cephalogram was a frequently requested radiograph. The most common indications reported were the transverse skeletal relationships and asymmetries, midline discrepancies, and airways (presumably nasal). Most of the changes produced by the posteroanterior cephalograms were related to diagnosis. Of the eight changes to treatment planning (by 5 of the 39 participants), three were referrals to otolaryngologist specialists, and three were related to a surgical decision. The posteroanterior cephalogram gives a two-dimensional view of the nasal airways and thus is questionable as a basis for referral to a medical specialist. Facial pattem, open bite tendency, and history of mouthbreathing or tonsillitis would seem to be more appropriate indications for such a referral. The posteroanterior cephalogram is very useful in visualizing the treatment changes in a patient who is undergoing orthognathic surgery. However, half of the orthodontists never ordered a posteroanterior cephalogram for any of the cases. In addition, an equal number
Volume 102 Number 1
An algorithm for ordering pretreatment radiographs
Fig. 5. Facial photographs, cephalometric tracing, and orthodontic models of patient E.
39
40
Atchison, Luke, atul White
Am. J. Ortho,I. Dentofac. Orthop.
July
~.
..J
Fig. 6. Facial photographs, cephalometric tracing, and orthodontic models of patient F.
1992
Volume 102
Number I
of participants considering surgery as a treatment option chose not to order a posteroanterior cephalogram. Therefore this study does not support the routine requesting of a posteroanterior cephalogram presurgically for all patients. A posteroanterior cephalogram may be indicated in severe asymmetry cases (such as hemifacial microsomia), but a recommendation for ordering a posteroanterior cephalogram could not be supported for the relatively minor symmetry problems encountered in this study. Lateral TMJ radiographs At the time these patients were seen in the dental school clinic, either lateral tomograms or transcranial radiographs were exposed for viewing the TMJ. The lateral corrected tomograms would now be the radiograph of choice for this situation. The submental vertex (SMV) was separately requested by two orthodontists during this study. We have chosen to include it as part of a TMJ series. Radiographs of the TMJ were frequently requested by the participants, but produced very few changes to diagnosis or treatment for the six test cases, only one of whom demonstrated any obvious signs or symptoms of TMJ disorder. Such signs and symptoms are quite common in the general population. Some reports have found the incidence to be more than 50% of the population. 17"ts Even children and adolescents have been found to have a high incidence of TMJ symptoms, only in a milder form. ~9 Discussion of recent legal proceedings 2~ has caused considerable concern among orthodontists as to the indications for pretreatment tomography. Other indications that have been suggested for tomograms might include mandibular displacements (both anterior as in a pseudo Class III and lateral as in a unilateral crossbite), prefunctional appliance therapy, and premandibular orthognathic surgery. This study does not support the taking of TMJ radiographs in the absence of significant TMJ signs or symptoms. There is no question that the appearance of osteoarthritis or aberrant position of a condyle could significantly affect orthodontic treatment. However, since the incidence of these findings in the absence of significant signs or symptoms of dysfunction is quite rare, screening for occult TMJ disease is not indicated. We accept the position of the American Academy of Pediatric Dentistry that "imaging is warranted when the clinical examination or history (or both) indicates that a recent or progressive pathological joint condition exists, such as: trauma; significant dysfunction/alteration in ROM; sensory/motor alterations; significant changes in occlusion (anterior open bite, posterior open bite, mandibular shift, and related factors); joint imaging is
An algorithm for orderbzg pretreatment radiographs
41
not indicated for joint sounds in the absence of other temporomandibular disorder signs and symptoms. ''2t On the basis of the tomographic findings, it may be necessary to consider further imaging possibilities. The clinician treating temporomandibular dysfunction can usually suspect the presence of an internal derangement on the basis of a clinical examination. In these instances the role of the radiographic examination of the TMJ is to rule out the presence of osseous disease and to evaluate the condylar position. This is accomplished well with tomography. When it is desired to examine the joints for the presence of osseous disease, both lateral and frontal tomograms are recommended. Computed tomography would be most appropriate if there is suggestive evidence of neoplasm or trauma. Arthrography is an invasive examination usually used to document soft tissue abnormality before TMJ surgery or after osseous abnormality has been excluded. We believe that arthrography should be limited to selected temporomandibular disorder cases of disk displacement when the dynamic imaging results will alter the course of treatment because of the invasiveness of the procedure (arthrography), radiation, and discomfort that can occur. 2-' Magnetic resonance imaging (MRI) examinations are invariably expensive and are highly technique sensitive. The spatial resolution of bone on a MRI examination is inferior to tomography. In our experience their indications are comparable to those for arthrography. Whether TMJ radiographs provide sufficient information in the presence of significant TMJ signs or symptoms to warrant routine requesting was not addressed in this study. This study does not support the taking of TMJ radiographs in patient cases with midline discrepancies or unilateral posterior crossbite. None of the patient cases exhibited severe facial asymmetry so that no conclusions could be made regarding the appropriateness of TMJ radiographs in such patients. However, since such a patient would already have received both a panoramic and a posteroanterior and lateral cephalometric examination, the clinician will have ample opportunity to screen for TMJ problems and order a follow-up TMJ examination, if warranted. The use of pretreatment TMJ radiographs was suggested by some participants considering functional appliance treatment for the test cases. Although the majority of orthodontists considering functional appliance therapy did not order TMJ radiographs, there was a trend showing that the percent of orthodontists who did was greater than the orthodontists who did not consider such treatment. This suggests that the potential for changes to the condyles from functional appliances is of concern to many orthodontists who use them. However, the number of such ortho-
42
Am. J. Orthod. Dentofac. Orthop. July 1992
Atchison, Luke, and White
~ Obtain Panoramic, Anterior Periapicals and Bitewing radiographs
Obtain Lateral
~ No
Cephalo~'am
No
Obtain Complete-mouth survey
Does the patient have a severe facial asymmetry?
Obtain Posterior-Anterior
~ram
No
"Does the patient have significant TMJ signs 9 or symptoms? .
Obtain TMJ Tomograms
end growth, is significant growth required for treatment or is $urget~ contemplated? J
Obtain Carpal Index
treat patient
Fig. 7. Algorithm for ordering pretreatment orthodontic radiographs.
Voh,me 102
Number 1
dontists was small, and this should be replicated in a larger sample. There is an absence in the dental literature of reports of frequent detrimental effects to condyles from functional appliances. Thus the rare occurrence would suggest that routine ordering of pretreatment TMJ radiographs is not indicated. Similarly, the proportion of participants ordering presurgical TMJ radiographs was even smaller, and their routine requesting in presurgical cases with no significant signs or symptoms of TMJ problems is not indicated by this study.
Carpal index The carpal index was requested 13 times for five of the patient cases and produced one change in treatment. Because of its very low impact on diagnosis and treatment planning, this study does not support the routine requesting of this examination. Suitable selection criteria may include patients at the end stages of growth where significant additional growth is required for treatment to succeed and patients who require orthognathic surgery where the orthodontist wishes to be relatively certain that growth has been completed.
Occlusals Of the 31 occlusal radiographs requested in this study, only one film caused a change in diagnosis or treatment plan. The change was to place a lower lingual arch and to extract maxillary primary canines. The reason typically given for requesting occlusals was to screen for midline pathosis. Since such pathosis is rare, and screening for occult pathosis is not indicated, 23 the routine use of occlusal radiographs in the absence of specific indications (such as impacted canines) is contraindicated unless they are used to replace anterior periapicals in visualizing the roots of anterior teeth. The recommendation that a lateral cephalogram is to be ordered for all patients undergoing comprehensive orthodontic treatment is neither surprising nor controversial. However, this recommendation may eventually become replaced as new techniques that provide information about the patient's skeletal characteristics are perfected that do not involve the use of ionizing radiation. The selection of the remaining radiographs, on the basis of patient characteristics and treatment approach, is consistent with work in the field of diagnostic radiography. Comparing this algorithm with the AAO guidelines, t3 the algorithm accounts for the radiographs that would not be routinely taken such as the posteroanterior cephalogram and TMJ radiographs. Also, the algorithth rejects the use of the complete-mouth series and the panogram for the same patient, which the guidelines do not. It is accepted that patient radiographs should not be ordered before completion of a clinical examination. Orthodontics lends itself to the use of selection
An algorithm for ordering pretreatment radiographs
43
criteria on the basis of patient characteristics because of the comprehensiveness of the history and clinical evaluation conducted. Indeed, much of the treatment is based on those same patient characteristics. What effect would the use of selection criteria have on the profession? Are orthodontists placing themselves at risk if they order less than a complete set of radiographs for their patients? We believe that use of the proposed algorithm would reduce unnecessary patient exposure without compromising patient care. There were 21 radiographs ordered that effected diagnosis or treatment and that were not indicated by the algorithm. All but one were due to the posteroanterior cephalogram and the TMJ radiographs. Several factors are intriguing about these two radiographs. First, the usefulness of these radiographs was not agreed on by the orthodontists. Of the 36 orthodontists, 17 (44%) never ordered a posteroanterior cephalogram, and 19 (49%) never ordered TMJ radiographs. Second, some of these "conditions," such as facial asymmetry, were diagnosed by other members of the study on the basis of the study models and the facial slides. This is consistent with findings by Han24 who found that 55% of the information needed by orthodontists for treatment planning was obtained from study models alone. Atchison et al. t4 found that the majority of confidence in diagnosis came from nonradiographie sources. Finally, the majority of the "problems" missed or the action required by the "problems" were not agreed on by the orthodontists. Several orthodontists demonstrated a close referral relationship with ENT specialists predicated by radiographic information provided by the posteroanterior cephalometric radiograph. However, the majority of the orthodontists did not recommend such action. It is difficult to assess what the actual clinical significance of these remaining few missed "problems" would be. Obviously, to the majority of the orthodontists it was not a significant "problem" requiring action, suggesting that the radiographs may have been ordered to provide medicolegal protection. Guided by this algorithm, a practitioner would be free to use his/her professional judgement to determine the actual number and type of radiographs required by each specific patient, depending on the nature and severity of the patient's complaint and not medicolegal protection. At the conclusion of each interview, the orthodontist was debriefed. In addition to answering any questions he had about the study, we asked "what possible risk could the patient face if treated with the initial treatment plan before you reviewed any radiographs?" Although there were some replies regarding increased treatment duration or the possibility of a missed pathologic condition, the most frequent response was "no risk." We asked whether the participant would be willing to ini-
44
Atchison, Luke, and White
tiate treatment for a patient without first reviewing radiographs. Overwhelmingly, the orthodontists stated that because of medicolegal pressures, they could not treat without radiographs. While we would not suggest that orthodontists treat without radiographs, the risks involved from exposure to any ionizing radiation mandates that unnecessary radiographs should not be ordered. The proposed algorithm is an attempt to suggest what are "necessary" radiographs. Further studies to test the algorithm with a broader sampling of patient problems would need to be completed to establish the appropriateness of this algorithm. We express our appreciation to the 39 orthodontists who participated in the study, to Dr. Patrick Turley for his advice, and to Dr. Jeff Gombein for his statistical assistance. REFERENCES 1. Bell RS, Loop .IW. The utility and futility of radiographic skull examinations for trauma. N Engl J bled 1971;284:236-9. 2. Phillips LA. A study of the effects of high yield-criteria for emergency room skull radiography. HEW Publications (FDA) 1978:78-8069. 3. Phillips L. Comparative evaluation of the effects of a high-yield criteria list upon skull radiography. JACEP 1979;8:106-9. 4. Jergens ME, Morgan MT, McElroy CE. Selective use of radiology of the skull and cervical spine. West J Med 1977;127: I-4. 5. Phillips L. Emergency room skull radiography. In National Conference on referral criteria for x-ray examinations. HEW Publication (FDA) 1979:79-8083. 6. Merton K. Upper GI examinations. In: National conference on referral criteria for x-ray examinations. HEW Publication (FDA) 1979:79-8083. 7. Brill P, Euwing M, Dunn A. The value (?) of routine chest radiography in children and adolescents. Pediatrics 1973;52: 125-7. 8A. White S, Forsythe A. High-yield criteria for panoramic radiography. HHS Publication (FDA) 1982:82-8186. 8B. White S, Forsythe A, Joseph L. Patient selected criteria for panoramic radiography. Oral Surg Oral Med Oral Pathol 1984;57:681-90. 9. Kog~ SL' Stephens PG" Selective radiography instead ~ screen"
Am. J. Orthod. Dentofac. Orthop. July 1992
10. l 1. 12.
13. 14.
15.
16.
17.
18. 19.
20. 21.
22.
23.
24.
ing pantomography-a risk/benefit evaluation. Can Dent Assoc J 1982;48:271-5. Brooks SL. A study of selection criteria for intraoral dental radiography. Oral Surg Oral Med Oral Pathol 1986;62:234-9. The selection of patients for x-ray examinations: dental radiographic examinations. IlttS Publication (FDA) 1987:88-8273. Recommendations in radiographic practices: an update, 1988. Council on Dental blaterials, Instruments, and Equipment, J Am Dent Assoc 1989;I 18:115-7. American Association of Orthodontists. Guidelines for Quality Assessment of Orthodontic Care. St Louis, blo., 1988. Atchison KA, Luke LS, White SC. Contribution of pretreatment radiographs to the orthodontists' decision making. Oral Surg Oral bled Oral Pathol 1991;71:238-45. Ignelzi blA, Fields HW, Vann WE Screening panoramic radiographs in children: prevalence data and implications. Pediatr Dent 1989;I 1:279-85. Kantor ML, Slome BA. Efficacy of panoramic radiography in dental diagnosis and treatment planning. J Dent Res 1989;68:810-2. Grosfeld O, Jackowska M, Czarnecka B. Results of epidemiological examinations of the temporomandibular joint in adolescents and young adults. J Oral Rehabil 1985;12:95-105. Solberg WK, Wood MW, Houston JB. Prevalence of mandibular dysfunctions in young adults. J Am Dent Assoc 1979;98:25-34. Egermark-Erickson I, Carlsson GE, Ingerwall B. Prevalence of mandibular dysfunction and orofacial parafunction in 7-, ll-, and 15-year old Swedish children. Eur J Orthod 1981;3:163-72. Laskin DM. Establishing standards of care. Guest editorial. AM J ORTItOD DEN'rOFACORTItOP 1988;93:171. American Academy of Pediatric Dentistry. Treatment of temporomandibular disorders in children: summary statements and recommendations. J Am Dent Assoc 1990;120:265-9. McNeill C, Mohl N, Rugh J, Tanaka T. Temporomandibular disorders: diagnosis, management, education, and research. J Am Dent Assoc 1990;120:253-63. Zeichner S J, Ruittimann UR, Weber RL. Dental radiography: efficacy in the assessment of intraosseous lesions of the face and the jaws in asymptomatic patients. Radiology 1987;162:691-5. Han UK. A study on the use of diagnostic records in orthodontic treatment decision. Masters thesis. University of blichigan 1989.
Reprint requests to: Dr. Kathryn Atchison UCLA School of Dentistry Center for the tlealth Sciences 63-025 Los Angeles, CA 90024-1668
Los Angeles, Ca~if. A study was conducted to identify selection criteria for ordering pretreatment orthodontic radiographs. Thirty-nine orthodontists evaluated six test cases. They provided information on the rationale for ordering each specific radiograph and the impact of the radiograph on the diagnosis and treatment plan. Skeletal relationshi p of the jaws was the most common indication for a radiograph request, followed by root formation/length and molar position or development. Of the radiographs, 16% produced a change in diagnosis, and 20% produced a change in treatment plans. The criteria specified for the radiographs, their impact, and relevant information in the literature were used to develop a n algorithm or set of decision rules that, when tested on the six test cases, resulted in a 36% reduction in the total number of radiographs. (AM J ORTHOD DENTOFACORTHOP 1992;102: 29-44.)
I d e a l l y , each time an orthodontist examines a new patient, he or she decides which radiographs are needed to complement the clinical information. The decision to order a particular radiographic examination may be influenced by several factors. A specific diagnostic examination may be ordered because a clinician wishes to perform a very thorough examination or believes there is a possibility of disease. This may be done to meet the standards of care for the community or to protect the orthodontist medicolegally. These factors are tempered, however, by the desire to expose the patient to the least amount of radiation necessary. Competing factors such as these interject a degree of difficulty to the decision of which or how many radiographs are appropriate as part of a diagnostic examination. In an effort to assist clinicians in making these difficult decisions regarding the use of diagnostic imaging, selection criteria have been developed for a variety of medical and dental radiographic examinations. 1"1~Selection criteria, or high-yield criteria, are guidelines or decision rules that guide the clinician in the choice of radiographs. Specifically, selection criteria are those signs, symptoms, or historic findings, which, if present, indicate there is a reasonable probability that a radiograph will be of value to the clinician in managing the care of a patient. The goal of developing selection criteria is to identify the optimal clinical circumstances for ordering radiographs necessary for diagnosis and
From the School of Dentistry, University of California, Los Angeles. *Assistant Professor, Section of Public Health Dentistry. ~'Clinieal Professor, Sections of Pediatric Dentistry and Orthodontics. *Professor, Section of Oral Radiol~y. 8/1/26113
treatment planning and to balance the clinicians diagnostic needs and desire to expose the patient to as little unnecessary radiation as possible. The value of selection criteria have been well documented. AS an example, in a classic study of skull radiographs obtained to evaluate a traumatic head injury, Bell and Loop t found that the use of clinical criteria for ordering radiographs significantly improved the percent of positive examinations. Of the adults admitted to the emergency room for head trauma, 58% had skull radiographs. However, evidence of fracture was seen in only 1.7%. When patients were assigned to high- or low-risk groups on the basis of clinical criteria, the detection of positive findings became 1 in 11.5 patients for the high-risk group and 1 in 435 patients for the low-risk group. A subsequent study demonstrated a 40% reduction in the number of skull radiographs ordered after adoption of ordering by clinical criteria. 2.3 Two studies have been conducted concerning the use of clinical criteria for ordering periapical radiographs for dental diagnosis. 9"t~Both demonstrated significant reductions in the number of radiographs ordered when clinical criteria were followed. A set of guidelines adopted by the American Dental Association tt.tz were recently developed that specify selection criteria for ordering general dental radiographs for new and recall patients. The American Association of Orthodontics (AAO) prepared recommendations that the appropriate radiograph examination for all patients initiating orihodontic treatment consist of a minimum of a lateral cephalogram, a panogram, and a complete-mouth set of periapicals and bitewings, t3 Atchison et al.~4 in their study of the ordering pat29
30
Atchison, Luke, and White .
terns of orthodontists on simulated cases, reported that about three-fourths of the radiographs ordered did not contribute original information to the orthodontist's certainty or content of his diagnosis and treatment plan. Further, it was found that the number of radiographs ordered by the orthodontists showed a stronger relationship to the orthodontist ordering the films than to the patient under consideration. These findings suggest the need for a detailed examination of the individual radiographic examinations ordered during the course of a comprehensive orthodontic patient examination. The purpose of this article is to identify the selection criteria for ordering orthodontic radiographs and then to organize these criteria into a set of easily followed decision rules or an algorithm. Because there are a number of radiographic examinations used for an orthodontic examination, one must consider not only the information influencing the decision to order a radiograph but also which radiograph best provides this information.
MATERIALS AND METHODS The design for this study has been previously described. ~4Briefly, 39 orthodontists, members of the Pacific Coast Society of Orthodontists and/or faculty at dental schools, were each interviewed for approximately 2 hours during which the participant evaluated six patient cases selected from the postdoctoral orthodontic program at the UCLA School of Dentistry. The records were duplicated maintaining patient confidentiality. The records included extraoral and intraoral slides, trimmed study models, and a patient history and examination summary. Composites of the extraoral photographs, study models, and a cephalometric tracing for each of the six cases are included in Figs. 1 through 6. The orthodontist was shown the case records and asked to formulate a diagnosis (enumerate the problems presented), which was recorded. The process was repeated for the treatment plan. The orthodontist was then told that he could request a radiograph if needed to complete his diagnosis or treatment plan. Each radiograph was requested individually and evaluated before another radiograph was requested. The complete-mouth series was given whenever a portion of it was requested. The Ricketts, Downs, Steiner, Wits, Harvold, and Sassouni analyses were presented along with a traced lateral cephalogram. The information sought from the radiograph was recorded. Once the respondent had examined the radiograph, he was asked whether the radiograph caused a change or a confirmation of the diagnosis. Any reSulting changes were noted. Again, the same questions were asked about the treatment plan. Radiographs were
Am. J. Orthod. Dentofac. Orthop. July 1992
selected until the orthodontist was confident he had all the diagnostic information he needed. Case A (Fig. 1), because of its relative simplicity, was used as the first case for each participant. The order of the remaining cases was randomized to prevent differences among the cases resulting from fatigue. Information was obtained regarding the number and type of radiographs ordered for the patients. Data was also collected regarding the information sought by the orthodontist from each radiograph for diagnosis and treatment planning, and the impact of the radiograph on the diagnosis and treatment plan. The impact of the radiograph was the self-reported assessment of the information provided by the radiograph on the diagnosis and treatment plan. Thus, a radiograph could provide information that changed the diagnosis or treatment plan, confirmed information provided by the clinical records or another radiograph, or had no effect on the diagnosis or treatment plan. Descriptive statistics were computed for each variable. Responses were grouped by orthodontist, test case, and radiograph. On the basis of the information requested by the participants and the impacts of each radiograph, a set of decision rules (an algorithm) was developed and tested for each case.
RESULTS The information sought by participants Thirty-nine participants completed the interview process, for a total of 234 case evaluations. Table I shows the information that participants reported needing in an open-ended format to complete a diagnosis and treatment plan for the six test cases, as well as the radiographs they requested to provide it. Skeletal relationship of the jaws was the most common indication for a radiograph, cited 216 times or 14.4% of all reasons for ordering a radiograph. Eight other indications each accounted for more than 5% of the requests for information: root formation or length, molar position and development, incisor angulation, general pathosis, presence of teeth, tooth shape, size and position, mandibular plane angle, and periodontal problems. Three of the indications, incisor angulation, skeletal relationship, and mandibular plane angle, involve information provided primarily by a lateral cephalogram. The lateral cephalogram alone accounts for 34% of the information sought by the orthodontists. The other commonly requested information was provided by the panogram or complete-mouth series examinations. The complete-mouth series and the panogram together provided about 51% of the information, but critical examination of the indications for them demonstrate con-
Volume 102 Number 1
An algorithm for ordering pretreatment radiographs
Fig. 1. Facial photographs, cephalometric tracing, and orthodontic models of patient A.
31
32 Atchison, Luke, and White
Am. J. Orthod. Dentofac. Orthop. July 1992
Table I. Information needs reported by orthodontists (11 = 39) when ordering radiographs to complete
diagnostic process for six test cases, by radiograph Itformation sought Incisor angulation Soft tissue profile Mandibular plane angle Facial type (pattern) Skeletal relationship of jaws Airway Skeletal asymmetry Midline deviation Condylar shape and position Osteoarthritis Molar position and development Tooth size, shape, position Presence of teeth Eruption sequence General pathosis Periodontal problems Root formation/ lengdl Caries Sutures Growth/skeletal age Other
tero, Cephalogram
Cephalogranl
AFI FMX
Panogrant
T~U
SMV
Occlusal
I Corpo,I index
127 34 91
127 34 91 17 202 17 2
17 216
14
4
7 36 14 1
3
5
I 4 19
25 44 14 67
1
43
14 144
14
3
I
2
2
17
119
38
59
4
104
26 7 42 63
79 14 60 24
I
106 22 121 87
93
54
59
9
13
I
147
13 11
4
5
Total
8
2
siderable overlap. The complete-mouth series was generally ordered for tasks requiring high image resolution, such as evaluation of periodontal problems, root formation or length, and dental caries. Participants requested a panogram for tasks where a broad view of the jaws was desirable, such as evaluation of molar position and development, tooth size, shape and position, and the presence of teeth or eruption problems. Impact of the radiograph on the diagnosis and treatment plan The participant determined the contribution of each radiograph toward the development of the diagnosis and treatment plan. Each radiograph could influence a diagnosis in one of three ways. A positive yield was defined as a radiograph that the participant stated caused a change to the diagnosis. A neutral yield was one that the participant stated confirmed diagnostic information already available, either as a result of the clinical information or from a previous radiograph. A radiograph was rated as having no effect if the participant reported it contributed no information to the diagnosis. The effect of the radiographs on diagnosis is shown in Table II. There were 120 radiographs (16%) that produced positive yields. These radiographs provided
2
I
2
68 13 15 20
359 information changes to the diagnosis (one radiograph could provide more than one piece of information). In terms of formulating a diagnosis, the radiographs were found to be most useful in the assessment of skeletal problems (138 information changes), and in the assessment of molar problems, including crowding, missing teeth, or impactions (85 information changes). Radiographs were also demonstrated to have beneficial effects on the development of the treatment plan. Table II shows that 148 radiographs, 20% of all radiographs ordered, produced changes to the treatment plan. The largest single category of changes dealt with molar, premolar, or other extraction decisions (123 treatment changes or 41%). This is not surprising since molar crowding was evident in three test cases. Ordering patterns observed in the study To provide the information needed, participants requested 741 radiographs, of which 192 (26%) produced a positive yield on diagnosis or treatment plan. To comprehensively examine the influence of practitioner versus patient characteristics on the number and type of radiographs ordered for the examination of a patient, the ordering patterns demonstrated by the 39 participants for the six test cases were examined. Table III
An algorithm for ordering pretreatnlent radiographs 33
Volume 102 Number 1
Fig. 2. Facial photographs, cephalometric tracing, and orthodontic models of patient B.
compares the distribution of the radiographs by test case, as well as the number of radiographs that produced a positive yield. It shows that the lateral cephalogram, panogram, and the complete-mouth series were ordered
by a majority of the practitioners (the denominator) for each test case. Less than half of the participants for any specific case ordered any of the other radiographs. Variability across case is noted, particularly in the number
34
Atchison, Luke, and White
Am. J. Orthod. Dcnt~zc'. Orthop. July 1992
t \. !
9
.
Fig. 3. Facial photographs, cephalometric tracing, and orthodontic models of patient C.
Volume 102
Number l
of times the posteroanterior cephalometric and TMJ radiographs were ordered for one case compared with another. Variability is also noted in the impact of the radiographs from one case to another. Although the number of participants who ordered a lateral cephalogram did not vary substantially across case, the number of lateral cephalograms that produced a positive yield did, ranging from six productive lateral cephalograms for case A or E to a high of 22 productive lateral cephalograms for case F. A similar range of productivity was found for the complete-mouth series examination.
Development of an algorithm for ordering orthodontic radiographs On the basis of the information needs reported by the participants, as well as the impact ofeach radiograph on the diagnosis and treatment plan, an algorithm was developed for ordering orthodontic radiographs (Fig. 7). Certain basic assumptions were made in developing the algorithm. 1. A radiograph is indicated only if the patient exhibits a clinical characteristic suggestive of a problem requiring radiographic examination or if there is a treatment decision that requires a radiograph. ~' The probability of occult disease being detected in the absence of any clinical sign or symptom is so low as to contraindicate radiographic examination solely on this basis. ~ 2. The algorithm described has been considered exclusively for prctreatment orthodontic radiographic examinations. 3. The algorithm is based on evaluation of patient cases that exhibit orthodontic problems and that would eventually require comprehensive (usually fully banded and/or bonded appliance) treatment. Simple tooth movement cases were not considered and may not require a lateral cephalogram radiograph. We propose in this algorithm that a patient should first be examined and a decision made whether the patient is a candidate for comprehensive orthodontic treatment. If so, the patient will require a lateral cephalogram. If the patient is heavily restored or exhibits periodontal problems, then a complete-mouth series is indicated to allow for examination of changes in alveolar bone and teeth. If not, a panogram, anterior periapicals, and posterior bitewings are indicated. If the patient displays a severe facial asymmetry, a posteroanterior cephalogram should be requested to assess whether the asymmetry is dental or skeletal, as well as the extent of asymmetry. Significant signs or symptoms of TMJ problems indicate the need for laterally cor-
An algorithm for ordering pretreatment radiographs
35
Table IIA. Impact of the 741 radiographs on orthodontist diagnoses and treatment plans Impact on diagnosis Radi~raphs with: Positive yield Neutral yield No effect
hnpaet on treatment plan 120 544 77
Radiographs with: Positive yield Neutral yield No effect
148 471 122
Table liB. Information changes
No. of changes
Diagnosis Skeletal problem Molar problem Caries or periodontal problem Incisor problem TMJ, condyle problem Root formation Profile considerations Eruption Midline problem Nonanterior crowding Other TOTAL
138 85 40 28 25 9 11 5 7 28 50 359
Treatment plan Molar extraction Premolar extraction Other extraction Functional appliance or headgear Treatment timing Ortbognathie surgery Treat nonextraction Expand arch Anterior bite correction Refer Other TOTAL
64 39 20 31 16 32 9 3 3 13 71 301
rected TMJ tomograms to evaluate for osseous or positional abnormalities in the joints. Finally, if the patient is thought to be at the end stages of growth potential and significant growth is needed for treatment success, or if orthognathic surgery is contemplated and the patient may still have growth potential, a carpal index should be requested. Table IV shows that if this algorithm had been followed there would have been a 36% reduction in the number of radiographs ordered for the six test cases. The extent of missed information had the algorithm been used was then determined for each test case. If no radiograph was indicated by the patient's characteristics, but one was ordered, the impact of the radiograph
36
Atchison, Luke, and White
Am. J. Orthod. Dentofac. Orthop. July 1992
Table Ill. Comparison of the number of radiographs with a positive yield* and the number and type of
radiographs ordered for each test case Number of changeslmtmber of radiographs ordered by patient test case Radiograph
a
Lateral cephal~ram Panogram Full-mouth series Posterioanterior cephalogram TMJ r a d i ~ m p h s Maxillary and mandibular occlusal Carpal index Submental vertex
6/39 8/32 2/30 1/4 014 0/l 1/4 0/1
I
I 13138 17/32 6/26 3/15 4115 0/3 0/3 0/0
C I 1/39 9/33 2/28 0/6 1/8 1/4 0/3 0/0
I
O 13139 9/29 10/29 4/11 l / 15 0/6 0/1 !/2
I
I 6/39 7/21 12134 4/12 4/10 0/7 0/0 0/0
TOTAL
r 22136 9/31 3/23 2/11 0/6 0/7 0/2 0/0
I
Total 71/230 59/178 35/170 14159 10158 1/28 1 / 13 1/3 1921741
*A positive yield was defined as a change produced in the diagnosis or treatment plan resulting from a radi~raph.
Table IV. Number and type of radiographs eliminated when ordering pretreatment orthodontic radiographs on the basis of the algorithm
Radiograph
Reduction in number
Reduction (%)
Lateral cephal~ram Pan.ram Complete-mouth series* P-A cephalogram TMJ radiographs Occlusals Carpal index TOTAL
0 - 22 -92 - 59 -58 - 28 - 13 -- 265
0 12 54 100 1130 100 1013 35.8
*When reducing the FMX, only posterior periapicals were removed because the algorithm calls for anterior periapicals and posterior bitewings to accompany all p a n . r a m s .
was checked to determine whether the orthodontist thought it provided original information that changed either the diagnosis or the treatment plan. The elimination of films not indicated by the algorithm would have failed to account for 21 changes or 8% of the 268 changes which were made to either the diagnosis.or the treatment plan. The changes not accounted for by using the algorithm are as follows: Patient A. One participant ordered a carpal index for a 12-year-old boy and added a functional appliance to speed up treatment. Although the carpal index showed considerable growth remaining, this could have been predicted by the child's age. Another participant ordered a posteroanterior cephalogram, determined that the midline was deviated 1 mm skeletally, and added a functional appliance to correct it. Patient B. Three participants ordered posteroanterior cephalograms that changed their diagnosis or ireat-
ment plans. One thought the posteroanterior cephalogram indicated a dental rather than a skeletal deviation of the midline and chose an asymmetric extraction pattern. Another orthodontist decided to extract third molars, use a splint, and perform a surgical visual treatment objective (VTO). A third was less concerned about the skeletal asymmetry after reviewing the posteroanterior cephalogram. Thus, for two participants, the posteroanterior cephalogram confirmed the impression that little skeletal asymmetry was present, and the other participant decided on third molar removal and the securing of more diagnostic records. The examination of lateral TMJ tomograms produced four changes. One orthodontist observed a skeletal problem and, hence, prescribed a surgical approach. A second observed a TMJ problem that he determined required treatment before commencement of orthodontic treatment. A third observed an anterior position of the condyle. And the fourth orthodontist suggested the patient may need splint therapy. The history for this patient stated that he had a click but no pain or limitation in movement. Patient C. One change was not accounted for by the radiographs recommended for patient C. One participant requested TMJ radiographs to check a possible condylar problem he observed on the lateral cephalogram and decided the problem did not exist after all. Patient D. Five changes resulted from radiographs not clinically indicated for patient D. Four changes occurred as a result of posteroanterior cephalograms. One orthodontist recommended the patient for an airway evaluation and counseled about surgery. Two participants observed an asymmetric mandible that they thought required surgery. The fourth noted an asymmetric mandible but did not choose to treat it. The addition of lateral TMJ radiographs led one participant to suggest a TMJ splint before surgery. Patient E. There were four changes reported by the
Volume 102
Number I
Air algorithm for ordering pretreatntent radiographs
"t
,
4
Fig. 4. Facial photographs, cephalometric tracing, and orthodontic models of patient D.
37
38
Atchison, Luke, and White
orthodontists on the basis of the posteroanterior cephalogram. The first noted an occluded airway, although the patient had no history of mouth breathing problems. Three clinicians were concerned about a possible skeletal asymmetry. Two of the three ruled asymmetry out on the basis of the radiograph, and one affirmed it, also on the basis of the posteroanterior cephalogram. Patient F. Two changes were not accounted for with the use of the algorithm, both resulting from the posteroanterior cephalogram. One participant recommended by an ear, nose and throat specialist. The other decided that the midline deviation was not skeletal.
DISCUSSION Although it is clear that the orthodontic radiographs provided substantial information to the participants, the study demonstrated considerable variation and inconsistency in the ordering of pretreatment orthodontic radiographs. These practice variations, in the face of the known biologic risk of ionizing radiation, suggests the need for guidelines for ordering orthodontic radiographs beyond those prepared by the AAO." The information needs listed in Table I, as well as the impact of the radiographs requested by the 39 orthodontists for six test cases, were used to develop criteria for ordering radiographs based on patient characteristics. These criteria were then organized into an algorithm that guides the orthodontist through the process of selecting appropriate radiographs for a patient. All criteria for requesting radiographs were developed by the orthodontists. We did not place a value on them, rather we used their self-reported data to determine which were productive criteria for developing the algorithm. In addition, current use of radiographs and the available literature were examined to help define relevant reasons to requestradiographs. It is recommended that previous radiographs should be obtained and used when suitable before ordering additional radiographs. Each of the radiographs considered will be discussed in depth.
Lateral cephalogram In addition to the high frequency of clinical indications noted in Table I, the lateral cephalogram is commonly used for comparison with standards, growth forecasting, and the development of VTO (especially if surgery is involved): Because of the broad nature of the information that a lateral cephalogram provides, its usefulness in treatment visualization and planning, a n d the impact that it demonstrated in changing the diagnosis and treatment plan for these six patient cases, the requesting of a pretreatment lateral cephalogram is sup-
Am. J. Orthod. Dentofac. Orthop. July 1992
ported by this study for all patients needing comprehensive orthodontic treatment.
Complete-mouth periapicals and bitewlngs versus panogram On the basis of the orthodontist's preference noted in Table I, the complete-mouth series is recommended if the patient is heavily restored or if periodontal problems, or root length, shape, and stage of formation questions need answering. If these situations do not exist, a panogram, posterior bitewings, and anterior periapicals are recommended to examine unerupted teeth and their development, eruption sequence, caries, and general oral pathosis. The advantage of this combination is that it incorporates the overall visualization provided by the panogram without forfeiting the detail necessary to examine caries, bone height, and root formation, particularly in the anterior region. This study supports the use of either a panogram or a complete-mouth series for every patient requiring comprehensive orthodontic treatment. What cannot be supported is the request of both a panogram and a complete-mouth series for the same patient. On the basis ofAtchison et al. ~~and the work of White and Forsythe, 8 Ignelzi et al., l~ and Kantor and Slome, 16 there is no rationale to order a panogram and a complete-mouth series together before interpretation of the radiographs. However, additional radiographs may occasionally be indicated on the basis of radiographic interpretation, to rule out disease observed in a radiograph in individual cases.
Posteroanterior cephalogram The posteroanterior cephalogram was a frequently requested radiograph. The most common indications reported were the transverse skeletal relationships and asymmetries, midline discrepancies, and airways (presumably nasal). Most of the changes produced by the posteroanterior cephalograms were related to diagnosis. Of the eight changes to treatment planning (by 5 of the 39 participants), three were referrals to otolaryngologist specialists, and three were related to a surgical decision. The posteroanterior cephalogram gives a two-dimensional view of the nasal airways and thus is questionable as a basis for referral to a medical specialist. Facial pattem, open bite tendency, and history of mouthbreathing or tonsillitis would seem to be more appropriate indications for such a referral. The posteroanterior cephalogram is very useful in visualizing the treatment changes in a patient who is undergoing orthognathic surgery. However, half of the orthodontists never ordered a posteroanterior cephalogram for any of the cases. In addition, an equal number
Volume 102 Number 1
An algorithm for ordering pretreatment radiographs
Fig. 5. Facial photographs, cephalometric tracing, and orthodontic models of patient E.
39
40
Atchison, Luke, atul White
Am. J. Ortho,I. Dentofac. Orthop.
July
~.
..J
Fig. 6. Facial photographs, cephalometric tracing, and orthodontic models of patient F.
1992
Volume 102
Number I
of participants considering surgery as a treatment option chose not to order a posteroanterior cephalogram. Therefore this study does not support the routine requesting of a posteroanterior cephalogram presurgically for all patients. A posteroanterior cephalogram may be indicated in severe asymmetry cases (such as hemifacial microsomia), but a recommendation for ordering a posteroanterior cephalogram could not be supported for the relatively minor symmetry problems encountered in this study. Lateral TMJ radiographs At the time these patients were seen in the dental school clinic, either lateral tomograms or transcranial radiographs were exposed for viewing the TMJ. The lateral corrected tomograms would now be the radiograph of choice for this situation. The submental vertex (SMV) was separately requested by two orthodontists during this study. We have chosen to include it as part of a TMJ series. Radiographs of the TMJ were frequently requested by the participants, but produced very few changes to diagnosis or treatment for the six test cases, only one of whom demonstrated any obvious signs or symptoms of TMJ disorder. Such signs and symptoms are quite common in the general population. Some reports have found the incidence to be more than 50% of the population. 17"ts Even children and adolescents have been found to have a high incidence of TMJ symptoms, only in a milder form. ~9 Discussion of recent legal proceedings 2~ has caused considerable concern among orthodontists as to the indications for pretreatment tomography. Other indications that have been suggested for tomograms might include mandibular displacements (both anterior as in a pseudo Class III and lateral as in a unilateral crossbite), prefunctional appliance therapy, and premandibular orthognathic surgery. This study does not support the taking of TMJ radiographs in the absence of significant TMJ signs or symptoms. There is no question that the appearance of osteoarthritis or aberrant position of a condyle could significantly affect orthodontic treatment. However, since the incidence of these findings in the absence of significant signs or symptoms of dysfunction is quite rare, screening for occult TMJ disease is not indicated. We accept the position of the American Academy of Pediatric Dentistry that "imaging is warranted when the clinical examination or history (or both) indicates that a recent or progressive pathological joint condition exists, such as: trauma; significant dysfunction/alteration in ROM; sensory/motor alterations; significant changes in occlusion (anterior open bite, posterior open bite, mandibular shift, and related factors); joint imaging is
An algorithm for orderbzg pretreatment radiographs
41
not indicated for joint sounds in the absence of other temporomandibular disorder signs and symptoms. ''2t On the basis of the tomographic findings, it may be necessary to consider further imaging possibilities. The clinician treating temporomandibular dysfunction can usually suspect the presence of an internal derangement on the basis of a clinical examination. In these instances the role of the radiographic examination of the TMJ is to rule out the presence of osseous disease and to evaluate the condylar position. This is accomplished well with tomography. When it is desired to examine the joints for the presence of osseous disease, both lateral and frontal tomograms are recommended. Computed tomography would be most appropriate if there is suggestive evidence of neoplasm or trauma. Arthrography is an invasive examination usually used to document soft tissue abnormality before TMJ surgery or after osseous abnormality has been excluded. We believe that arthrography should be limited to selected temporomandibular disorder cases of disk displacement when the dynamic imaging results will alter the course of treatment because of the invasiveness of the procedure (arthrography), radiation, and discomfort that can occur. 2-' Magnetic resonance imaging (MRI) examinations are invariably expensive and are highly technique sensitive. The spatial resolution of bone on a MRI examination is inferior to tomography. In our experience their indications are comparable to those for arthrography. Whether TMJ radiographs provide sufficient information in the presence of significant TMJ signs or symptoms to warrant routine requesting was not addressed in this study. This study does not support the taking of TMJ radiographs in patient cases with midline discrepancies or unilateral posterior crossbite. None of the patient cases exhibited severe facial asymmetry so that no conclusions could be made regarding the appropriateness of TMJ radiographs in such patients. However, since such a patient would already have received both a panoramic and a posteroanterior and lateral cephalometric examination, the clinician will have ample opportunity to screen for TMJ problems and order a follow-up TMJ examination, if warranted. The use of pretreatment TMJ radiographs was suggested by some participants considering functional appliance treatment for the test cases. Although the majority of orthodontists considering functional appliance therapy did not order TMJ radiographs, there was a trend showing that the percent of orthodontists who did was greater than the orthodontists who did not consider such treatment. This suggests that the potential for changes to the condyles from functional appliances is of concern to many orthodontists who use them. However, the number of such ortho-
42
Am. J. Orthod. Dentofac. Orthop. July 1992
Atchison, Luke, and White
~ Obtain Panoramic, Anterior Periapicals and Bitewing radiographs
Obtain Lateral
~ No
Cephalo~'am
No
Obtain Complete-mouth survey
Does the patient have a severe facial asymmetry?
Obtain Posterior-Anterior
~ram
No
"Does the patient have significant TMJ signs 9 or symptoms? .
Obtain TMJ Tomograms
end growth, is significant growth required for treatment or is $urget~ contemplated? J
Obtain Carpal Index
treat patient
Fig. 7. Algorithm for ordering pretreatment orthodontic radiographs.
Voh,me 102
Number 1
dontists was small, and this should be replicated in a larger sample. There is an absence in the dental literature of reports of frequent detrimental effects to condyles from functional appliances. Thus the rare occurrence would suggest that routine ordering of pretreatment TMJ radiographs is not indicated. Similarly, the proportion of participants ordering presurgical TMJ radiographs was even smaller, and their routine requesting in presurgical cases with no significant signs or symptoms of TMJ problems is not indicated by this study.
Carpal index The carpal index was requested 13 times for five of the patient cases and produced one change in treatment. Because of its very low impact on diagnosis and treatment planning, this study does not support the routine requesting of this examination. Suitable selection criteria may include patients at the end stages of growth where significant additional growth is required for treatment to succeed and patients who require orthognathic surgery where the orthodontist wishes to be relatively certain that growth has been completed.
Occlusals Of the 31 occlusal radiographs requested in this study, only one film caused a change in diagnosis or treatment plan. The change was to place a lower lingual arch and to extract maxillary primary canines. The reason typically given for requesting occlusals was to screen for midline pathosis. Since such pathosis is rare, and screening for occult pathosis is not indicated, 23 the routine use of occlusal radiographs in the absence of specific indications (such as impacted canines) is contraindicated unless they are used to replace anterior periapicals in visualizing the roots of anterior teeth. The recommendation that a lateral cephalogram is to be ordered for all patients undergoing comprehensive orthodontic treatment is neither surprising nor controversial. However, this recommendation may eventually become replaced as new techniques that provide information about the patient's skeletal characteristics are perfected that do not involve the use of ionizing radiation. The selection of the remaining radiographs, on the basis of patient characteristics and treatment approach, is consistent with work in the field of diagnostic radiography. Comparing this algorithm with the AAO guidelines, t3 the algorithm accounts for the radiographs that would not be routinely taken such as the posteroanterior cephalogram and TMJ radiographs. Also, the algorithth rejects the use of the complete-mouth series and the panogram for the same patient, which the guidelines do not. It is accepted that patient radiographs should not be ordered before completion of a clinical examination. Orthodontics lends itself to the use of selection
An algorithm for ordering pretreatment radiographs
43
criteria on the basis of patient characteristics because of the comprehensiveness of the history and clinical evaluation conducted. Indeed, much of the treatment is based on those same patient characteristics. What effect would the use of selection criteria have on the profession? Are orthodontists placing themselves at risk if they order less than a complete set of radiographs for their patients? We believe that use of the proposed algorithm would reduce unnecessary patient exposure without compromising patient care. There were 21 radiographs ordered that effected diagnosis or treatment and that were not indicated by the algorithm. All but one were due to the posteroanterior cephalogram and the TMJ radiographs. Several factors are intriguing about these two radiographs. First, the usefulness of these radiographs was not agreed on by the orthodontists. Of the 36 orthodontists, 17 (44%) never ordered a posteroanterior cephalogram, and 19 (49%) never ordered TMJ radiographs. Second, some of these "conditions," such as facial asymmetry, were diagnosed by other members of the study on the basis of the study models and the facial slides. This is consistent with findings by Han24 who found that 55% of the information needed by orthodontists for treatment planning was obtained from study models alone. Atchison et al. t4 found that the majority of confidence in diagnosis came from nonradiographie sources. Finally, the majority of the "problems" missed or the action required by the "problems" were not agreed on by the orthodontists. Several orthodontists demonstrated a close referral relationship with ENT specialists predicated by radiographic information provided by the posteroanterior cephalometric radiograph. However, the majority of the orthodontists did not recommend such action. It is difficult to assess what the actual clinical significance of these remaining few missed "problems" would be. Obviously, to the majority of the orthodontists it was not a significant "problem" requiring action, suggesting that the radiographs may have been ordered to provide medicolegal protection. Guided by this algorithm, a practitioner would be free to use his/her professional judgement to determine the actual number and type of radiographs required by each specific patient, depending on the nature and severity of the patient's complaint and not medicolegal protection. At the conclusion of each interview, the orthodontist was debriefed. In addition to answering any questions he had about the study, we asked "what possible risk could the patient face if treated with the initial treatment plan before you reviewed any radiographs?" Although there were some replies regarding increased treatment duration or the possibility of a missed pathologic condition, the most frequent response was "no risk." We asked whether the participant would be willing to ini-
44
Atchison, Luke, and White
tiate treatment for a patient without first reviewing radiographs. Overwhelmingly, the orthodontists stated that because of medicolegal pressures, they could not treat without radiographs. While we would not suggest that orthodontists treat without radiographs, the risks involved from exposure to any ionizing radiation mandates that unnecessary radiographs should not be ordered. The proposed algorithm is an attempt to suggest what are "necessary" radiographs. Further studies to test the algorithm with a broader sampling of patient problems would need to be completed to establish the appropriateness of this algorithm. We express our appreciation to the 39 orthodontists who participated in the study, to Dr. Patrick Turley for his advice, and to Dr. Jeff Gombein for his statistical assistance. REFERENCES 1. Bell RS, Loop .IW. The utility and futility of radiographic skull examinations for trauma. N Engl J bled 1971;284:236-9. 2. Phillips LA. A study of the effects of high yield-criteria for emergency room skull radiography. HEW Publications (FDA) 1978:78-8069. 3. Phillips L. Comparative evaluation of the effects of a high-yield criteria list upon skull radiography. JACEP 1979;8:106-9. 4. Jergens ME, Morgan MT, McElroy CE. Selective use of radiology of the skull and cervical spine. West J Med 1977;127: I-4. 5. Phillips L. Emergency room skull radiography. In National Conference on referral criteria for x-ray examinations. HEW Publication (FDA) 1979:79-8083. 6. Merton K. Upper GI examinations. In: National conference on referral criteria for x-ray examinations. HEW Publication (FDA) 1979:79-8083. 7. Brill P, Euwing M, Dunn A. The value (?) of routine chest radiography in children and adolescents. Pediatrics 1973;52: 125-7. 8A. White S, Forsythe A. High-yield criteria for panoramic radiography. HHS Publication (FDA) 1982:82-8186. 8B. White S, Forsythe A, Joseph L. Patient selected criteria for panoramic radiography. Oral Surg Oral Med Oral Pathol 1984;57:681-90. 9. Kog~ SL' Stephens PG" Selective radiography instead ~ screen"
Am. J. Orthod. Dentofac. Orthop. July 1992
10. l 1. 12.
13. 14.
15.
16.
17.
18. 19.
20. 21.
22.
23.
24.
ing pantomography-a risk/benefit evaluation. Can Dent Assoc J 1982;48:271-5. Brooks SL. A study of selection criteria for intraoral dental radiography. Oral Surg Oral Med Oral Pathol 1986;62:234-9. The selection of patients for x-ray examinations: dental radiographic examinations. IlttS Publication (FDA) 1987:88-8273. Recommendations in radiographic practices: an update, 1988. Council on Dental blaterials, Instruments, and Equipment, J Am Dent Assoc 1989;I 18:115-7. American Association of Orthodontists. Guidelines for Quality Assessment of Orthodontic Care. St Louis, blo., 1988. Atchison KA, Luke LS, White SC. Contribution of pretreatment radiographs to the orthodontists' decision making. Oral Surg Oral bled Oral Pathol 1991;71:238-45. Ignelzi blA, Fields HW, Vann WE Screening panoramic radiographs in children: prevalence data and implications. Pediatr Dent 1989;I 1:279-85. Kantor ML, Slome BA. Efficacy of panoramic radiography in dental diagnosis and treatment planning. J Dent Res 1989;68:810-2. Grosfeld O, Jackowska M, Czarnecka B. Results of epidemiological examinations of the temporomandibular joint in adolescents and young adults. J Oral Rehabil 1985;12:95-105. Solberg WK, Wood MW, Houston JB. Prevalence of mandibular dysfunctions in young adults. J Am Dent Assoc 1979;98:25-34. Egermark-Erickson I, Carlsson GE, Ingerwall B. Prevalence of mandibular dysfunction and orofacial parafunction in 7-, ll-, and 15-year old Swedish children. Eur J Orthod 1981;3:163-72. Laskin DM. Establishing standards of care. Guest editorial. AM J ORTItOD DEN'rOFACORTItOP 1988;93:171. American Academy of Pediatric Dentistry. Treatment of temporomandibular disorders in children: summary statements and recommendations. J Am Dent Assoc 1990;120:265-9. McNeill C, Mohl N, Rugh J, Tanaka T. Temporomandibular disorders: diagnosis, management, education, and research. J Am Dent Assoc 1990;120:253-63. Zeichner S J, Ruittimann UR, Weber RL. Dental radiography: efficacy in the assessment of intraosseous lesions of the face and the jaws in asymptomatic patients. Radiology 1987;162:691-5. Han UK. A study on the use of diagnostic records in orthodontic treatment decision. Masters thesis. University of blichigan 1989.
Reprint requests to: Dr. Kathryn Atchison UCLA School of Dentistry Center for the tlealth Sciences 63-025 Los Angeles, CA 90024-1668
