Orthodontists’ criteria for prescribing cone-beam computed tomography—a multi-country survey

Abstract

Objectives

This study aimed to describe and compare CBCT imaging prescription in clinical practice among orthodontists from five countries in Europe and America. Additionally, it investigated factors associated with the prescribing and the use of guidelines for CBCT imaging.

Materials and methods

A cross-sectional survey was carried out using an online questionnaire sent to all registered orthodontists in Belgium, Brazil, Canada, Romania, and the United States of America (USA). The data were analyzed by descriptive statistics, bivariate tests, and Poisson regression.

Results

The final sample consisted of 1284 participants. CBCT was prescribed by 84.4% of the participants for selected cases (84.9%), mainly for impacted teeth (92.4%), presurgical planning (54.1%), and root resorption (51.9%). High cost was most frequently the limiting factor for CBCT prescription (55.4%). Only 45.2% of those who were using CBCT imaging reported adhering to guidelines. CBCT imaging prescription was associated with the orthodontists’ countries (p < .009, except for Belgium, p = .068), while the use of guidelines was associated with the respondents’ country and additional training on CBCT imaging (p < .001).

Conclusions

Orthodontists refer patients for CBCT for selected indications (impacted teeth, root resorption, presurgical planning, dentofacial deformities, as suggested by the international guidelines, and also for upper airway and temporomandibular joint evaluation). Many do not adhere to specific guidelines. There are substantial variations between the countries about the orthodontists’ referral for CBCT and guideline usage, irrespective of gender. CBCT prescription may be limited by financial barriers, adhering to specific guidelines and prior CBCT training.

Clinical relevance

CBCT prescription among orthodontists must be based on prescription criteria and current guidelines. It is advised to improve CBCT education and training to enhance CBCT selection, referral, analysis, and interpretation in orthodontic practice.

Introduction

Cone-beam computed tomography (CBCT) has been introduced more than two decades ago and has been extensively documented for its use in various dental specialties. If 3D information is required, CBCT may be preferred to multi-slice computed tomography due to its accessibility, lower ionizing radiation levels, and high spatial resolution. On the other hand, its use is also limited by the higher biological effect than conventional 2D radiographs, the higher financial cost, and the potentially limited availability in rural areas [1–4].

In orthodontics, the standard of care for diagnosis and treatment planning of malocclusions consists of 2D radiographs (lateral cephalometric, panoramic, and periapical), extra- and intraoral photographs, and model (plaster or digital) analysis [5]. Due to the inherent limitations of 2D imaging modalities, such as anatomical superimposition and distortion, marketing reasons, and integration to a digital treatment approach, an exponential increase in the use of CBCT imaging has been noted over the last decade [6]. Three-dimensional (3D) images are used in orthodontic treatment planning for adult and pediatric patients [7–10].

Knowing the risk associated with the stochastic effects of ionizing radiation [3, 7], a significant downside of CBCT imaging may be the increasing often unnecessary prescription [7]. Children and adolescents are frequent patients in orthodontic practices and can be two to ten times more sensitive to ionizing radiation than adults. The risk of radiation effects is five times greater in children than in adults for the same exposure setting [7]. Moreover, children present a considerably higher risk of radiation carcinogenesis [8–11]. For these reasons, the prescription of CBCT imaging for pediatric patients is only justified in those cases where conventional radiography offers limited information for diagnosis and “the potential clinical benefits of the advanced imaging modalities outweigh risks of ionizing radiation exposure” [12, 13]. Also, exposure parameters, such as field of view (FOV) and voxel size, should be carefully selected, while maintaining the image with a sufficient diagnostic value [14–16]. As a general basic principle, international guidelines do not recommend CBCT as a routine examination in orthodontics [3, 7, 17]. However, not all professionals follow this recommendation, resulting in the indiscriminate use of CBCT [18].

Considering the unknown and possible divergence on the prescription and justification of CBCT imaging in orthodontics in different countries [8–10, 16, 19], further research is needed to evaluate the reasoning behind a professional’s judgment about when to prescribe CBCT imaging in orthodontics. Therefore, this study aimed to describe and compare CBCT imaging in clinical practice among orthodontists in Brazil related to the USA, Canada, Belgium, and Romania. Also, a secondary aim was to investigate factors associated with the prescription and use of CBCT guidelines. It was hypothesized that there are variations in CBCT imaging prescription in orthodontics when considering samples from the five countries.

Materials and methods

This international multicentric study was carried out among orthodontists registered in five countries, selected by convenience, in Europe and South and North America: Belgium, Brazil, Canada, Romania, and the United States of America (USA). Each country’s Institutional Ethics Committee approved the research protocol (Brazil, #15,742,619.7.0000.5083; Belgium, #S63051; Canada, #Pro00092421; USA, #HUM00162784). In Romania, this type of research is exempt from ethical clearance. The survey was carried out following the principles outlined in the Declaration of Helsinki.

Study design, sample size calculation and data collection

A cross-sectional survey was conducted. The study population was composed of orthodontists in each country, as registered in their respective professional associations. The sample size was calculated based on simple random sampling methods and stratified sampling with proportional allocation using R software, version 3.6.2. A sampling error value of 3% with a 95% confidence was determined to be acceptable. On that basis, a minimum of 670 participants was required: 11 in Belgium (of 450 registered), 570 in Brazil (of 23,872 registered), 16 in Canada (of 636 registered), 20 in Romania (of 800 registered), and 53 in the USA (of 2,204 registered).

Data were collected between February and September 2020, using a self-administered electronic questionnaire sent to the participants through the SurveyMonkey platform. A reminder was sent after 1 week and then another after 30 days. A questionnaire was drafted explicitly for the present study due to the lack of validated data collection instruments from previous studies. Figure 1 shows a flowchart with the methodological steps of this research phase. The questionnaire was initially designed in American English, and then translated and cross-culturally adapted for each language: Portuguese for Brazil, Dutch and French for Belgium, and Romanian for Romania. The pretest and the pilot study were carried out among orthodontists from Brazil and Canada. The participants in these two tests were different professionals. The retest assessment to validate questionnaire reliability involved the same Brazilian orthodontists who had participated in the pilot study 1.

Fig. 1.

Flowchart of the development of the questionnaire

The final version of the questionnaire consisted of three sections: (a) demographic characteristics of the orthodontists, (b) criteria for prescribing CBCT imaging in orthodontic practice, and (c) participant’s technical knowledge and personal opinion of CBCT. The questionnaire was composed of one open-ended and 23 multiple-choice questions. The only mandatory question was the research participation confirmation. Through the SurveyMonkey platform, participation was voluntary and confidential.

Data analysis

Answers to the questionnaires were automatically registered in a database and analyzed using the SPSS 22 software (IBM Corp. Armonk, NY, USA). Descriptive statistics of the variables obtained from the closed questions were initially performed and were all analyzed as categorical. Chi-square tests were used in the bivariate comparisons between the five countries.

To identify the variables associated with each outcome (prescription of CBCT and use of guidelines for prescription), Poisson regression models with robust variance were undertaken, with prevalence ratios (PR) and their respective 95% confidence intervals (CI). An unadjusted regression analysis (bivariate) was initially performed to identify each dependent variable’s associated independent variables (demographic and professional characteristics of the respondents and training/course for using CBCT imaging). Those with p < 0.25 in the unadjusted analysis were included in the adjusted models, which considered p < 0.05 for statistically significant results. Non-respondents were excluded and only those who answered “yes” or “no” to the questions regarding the outcomes were initially selected. Non-respondents of the independent variables were also excluded, which resulted in smaller sample sizes in the final models.

The single open-ended question placed at the end of the questionnaire (“Please share any additional thoughts that you may have regarding the prescription of CBCT in orthodontic practice”) was descriptively analyzed and results were summarized to identify the main aspects that emerged.

Results

Of the 27,962 questionnaires sent out to orthodontists, a total of 1359 were returned. Response rates in each of the five countries studied ranged from 4.2% (Brazil) to 11.8% (Romania). Participants who reported not being orthodontists (n = 35), who declined to participate (n = 18), and those who gave incomplete answers (n = 22) were further excluded from the analysis. The final sample consisted of 1284 participants (Fig. 2). Besides, we identified 180 participants (14%) who did not prescribe CBCT imaging or did not respond to this question (n = 21), yielding a total of 1083 respondents for the final analyses.

Fig. 2.

Flowchart of number of participants in each study phase

Demographic and professional characteristics of respondents

Table 1 presents the orthodontists’ demographic and professional characteristics and variables regarding their CBCT imaging prescription. Male orthodontists made up 51.53% of the total sample, while in the USA and Canada, the vast majority was female (80.9% and 70.2%, respectively). Sex differences were not significant in Belgium and the total sample. The comparisons between the countries showed no significant difference between the USA and Canada, but both countries were significantly different from Brazil, Belgium, and Romania.

Table 1.

Orthodontist’s characteristics, prescription of CBCT imaging, and use of guidelines in each country (n = 1284)

Variable	Countries												P-value*
	Total		Belgium (n = 51)		Brazil (n = 939)		Canada (n = 47)		Romania (n = 85)		USA (n = 162)
	N	%	n	%	n	%	n	%	n	%	n	%
Gender
Female	618	48.1	19	38.0a,b	420	45.0a	33	70.2c	15	17.6b	131	80.9c	< .001
Male	659	51.3	31	62.0a,b	514	55.0a	14	29.8c	69	81.2b	31	19.1c	< .001
Other**	1	0.1	0		0		0		1	1.2	0
Not answered**	6	0.5	1		5		0		0		0
Age group (years)													< .001
20–29	94	7.3	0a,b		91	9.7a	0a,b		3	3.6a,b	0b		< .001
30–39	333	25.9	4	7.8c	267	28.6a	7	14.9a,c	41	48.8b	14	8.6c	< .001
40–49	404	31.5	16	31.4	304	32.5	12	25.5	26	31.0	46	28.4	NS
50–59	303	23.6	26	51.0b	193	20.6a	12	25.5a,b	14	16.7a	58	35.8b	< .001
60–69	126	9.8	5	9.8a,c	69	7.4a,b	12	25.5c	0	0.0b	40	24.7c	< .001
> 70	19	1.5	0a,b		11	1.2a	4	8.5b	0	0.0a,b	4	2.5a,b	< .001
Not answered**	5	0.4	0		4		0		1		0
Number of years as a licensed orthodontist
≤ 5	225	17.5	3	5.9a,b	196	20.9a	5	10.6a	20	23.5a	1	0.6b	< .001
6–10	213	16.6	0b		164	17.5a	7	14.9a,b	27	31.8b	15	9.3a,c	< .001
11–20	383	29.8	11	21.6	289	30.8	9	19.1	26	30.6	48	29.6	NS
≥ 21	463	36.1	37	72.5c	290	30.9a	26	55.3c	12	14.1b	98	60.5c	< .001
Prescribe CBCT
No	180	14.0	4	8.0a,d	169	18.1a	3	6.4a,b,d	4	4.8b,c,d	0c		< .001
Yes, instead of 2D images	114	8.9	2	4.0a	58	6.2a	7	14.9a,b	7	8.3a	40	27.2b	< .001
Yes, as a complement to 2D images	969	75.5	44	88.0	708	75.7	37	78.7	73	86.9	107	72.8	NS
Not answered**	21	1.6	1		4		0		1		15
Follow guidelines	N	%	n = 46		n = 766		n = 44		n = 80		n = 147
	1083		n	%	n	%	n	%	n	%	n	%
No	576	53.2	24	53.3a	349	46.2a	22	50.0a	66	83.5b	115	80.4b	< .001
Yes	490	45.2	21	46.7a	406	53.8a	22	50.0a	13	16.5b	28	19.6b	< .001
Not answered**	17	1.6	1		11		0		1		4

Different letters represent statistically significant differences between countries (p < .05) and the same letters represent no statistically significant differences between countries(p > .05) NS = not significant (p > .05)

^*Chi-square test.

^**Categories not included in the chi-square test

Most of respondents were aged between 30 and 59 years with more than 10 years as a licensed orthodontist. Bivariate comparisons showed significant differences among the countries.

Criteria for prescribing CBCT imaging

The clinical criteria were divided into three parts: dental conditions, craniofacial morphology, and orthodontic appliances (Table 2). Dental conditions and craniofacial morphology, in particular: impacted teeth (92.4%), surgical cases (54.1%), root resorption (51.9%), dental anomalies (44.3%), dentofacial deformities (40.8%), and temporomandibular joint (TMJ) evaluation (39.8%), showed the highest prevalence for the prescription of CBCT imaging, in all countries. There were no statistically significant differences in terms of the criteria “root resorption” between countries. Dental anomalies which showed a higher prevalence for prescription in Canada (68.2%) and the USA (65.3%) were statistically different from Belgium (46.6%), Brazil (41.9%), and Romania (40.5%). As for CBCT prescription for the dental position, there were statistically significant differences between Romania, Belgium, Brazil, and the USA with prevalence rates of 59.5%, 44.4%, 34.4%, and 22.4%, respectively.

Table 2.

Clinical criteria for prescribing CBCT imaging in each country

Variable (more than one answer)	Total N = 1083		Countries
			Belgium n = 46		Brazil n = 766		Canada n = 44		Romania n = 80		USA n = 147
	n	%	n	%	n	%	n	%	n	%	n	%
None of the answer options	4	0.4	0		0		2	4.5	2	2.5	0
At least one answer option	1029	95.0	45	100a,b	718	100.0a	42	95.5b	77	97.5b	147	100.0a,b
Not answered	50	4.6	1		48		0		1		0
Dental conditions
Impacted teeth	1001	92.4	45	100.0	699	97.4	42	95.5	75	94.9	140	95.2
Root resorption	562	51.9	31	68.9	389	54.2	25	56.8	38	48.1	79	53.7
Dental anomalies	480	44.3	21	46.67a,b	301	41.9a	30	68.2b	32	40.5 a	96	65.3b
Dental position	358	33.1	20	44.4a,b	247	34.4a	11	25.0a,c	47	59.5b	33	22.4c
Craniofacial morphology
Facial asymmetry	403	37.2	8	17.8b	298	41.5a	16	36.4a,b	25	31.6a,b	56	38.1a,b
Upper airways	206	19.0	2	4.4b,d	148	20.6a,b,c,d	10	22.7a,b,c,d	7	8.9c,d	39	26.5a
Palatal cleft	320	29.5	10	22.2	208	29.0	11	25.0	32	40.5	59	40.1
Surgical cases	586	54.1	6	13.3b	434	60.4a	19	43.2%a	44	55.7a	83	56.5a
Dentofacial deformities	442	40.8	10	19.3b	324	45.1a	17	38.6a,b	17	21.5b	74	50.3a
TMJ analysis (n/%)	431	39.8	5	11.1b	303	42.2a	22	50.0a	44	55.7a	57	38.8a
Orthodontic appliances
Anchoring with mini-implants	191	17.6	5	11.1a	109	15.2a	7	15.9a	34	43.0b	36	24.5a
Digital smile image	101	9.3	3	6.7	81	11.3	2	4.5	6	7.6	9	6.1
Other	60	5.5	5	11.1	40	5.6	1	2.3	3	3.8	11	7.5

Chi-square test. Different letters represent statistically significant differences between countries (p < .05) and the same letters represent no statistically significant differences between countries (p > .05)

Factors associated with the prescription of CBCT imaging

According to the Poisson regression, Table 3 shows the associations between the outcome variable “prescription of CBCT imaging” and the respondents’ demographic and professional features. In the unadjusted analysis, “age,” “years as a licensed orthodontist,” and “country” were associated with the outcome. In the final (adjusted) model, only the variable “country” remained associated with the “prescription of CBCT imaging.” When the groups were compared with the reference category “Brazil,” which had the lowest frequency, the proportions of respondents from Canada, Romania, and the USA who prescribed CBCT were 1.1, 1.2, and 1.2 times respectively higher than those from Brazil (p < 0.05). Belgium showed a result similar to Brazil (1.09).

Table 3.

Poisson regression analysis of the association between the prescription of cone-beam computed tomography (CBCT) and demographic and professional variables (N = 1252^*)

Independent variables	CBCT prescription
	Unadjusted PR (95% CI)	Adjusted**
		p ***	PR (95% CI)	p ***
Gender
Male	1		1
Female	1,03 (0.99–1.08)	.146	1.01 (0.96–1.06)	.794
Age group
20–39	1		1
40–59	1.06 (1.00–1.11)	.040	1.02 (0.95–1.10)	.530
≥ 60	1.08 (1.00–1.16)	.038	0.98 (0.88–1.08)	.659
Years as a licensed orthodontist
< 5	1		1
6–20	0.97 (0.91–1.04)	.355	0.94 (0.86–1.01)	.099
> 20	1.07 (1.01–1.14)	.035	1.02 (0.93–1.12)	.674
Country
Brazil	1		1
Belgium	1.12 (1.03–1.22)	.009	1.09 (0.99–1.19)	.068
Canada	1.14 (1.05–1.24)	.001	1.13 (1.04–1.23)	.003
Romania	1.16 (1.10–1.23)	< .001	1.18 (1.11–1.25)	< .001
USA	1.22 (1.18–1.26)	< .001	1.21 (1.17–1.26)	< .001

^*Of the 1263 respondents, 11 were excluded due to non-responses in the independent variables

^**adjusted by the variables included in the table

^***Wald test

PR, prevalence ratio; 95% CI, confidence interval of 95%

Factors associated with the use of guidelines for CBCT prescription

The association between the use of guidelines for prescribing CBCT imaging and respondents’ demographic and professional features is presented in Table 4. In the unadjusted analysis, the variables “training/course for using CBCT imaging” and “country” were associated with the outcome. After adjustment, these variables remained significant, and “age” was also associated with the outcome. The proportion of respondents in the oldest age group (≥ 60 years) who reported following guidelines was 1.3 times higher than that in the youngest group. Compared with respondents from Romania, the prevalence of those from Brazil, Belgium, and Canada who used guidelines was 3.2, 3.0, and 2.9 times higher, respectively. Those who reported having undergone CBCT training/courses had a 1.5 times higher prevalence of using guidelines than those who did not have any related training.

Table 4.

Poisson regression analysis of association between the use of guidelines for the prescription of CBCT imaging and professional variables (N = 1064)

Independent variable	Guideline followers
	Yes	No	Unadjusted		Adjusted*
	N (%)	N (%)	PR (95%CI)	p **	PR (95%CI)	p **
Gender
Male	243 (44.9)	298 (55.1)	1
Female	239 (46.2)	278 (53.8)	1.03 (0.90–1.17)	.668
Not answered (N = 6)***
Age
20–39	158 (45.9)	186 (54.1)	1		1
40–59	336 (56.3)	261 (43.7)	0.95 (0.82–1.10)	.509	1.05 (0.91–1.22)	.481
≥ 60	56 (47.5)	62 (52.5)	1.14 (0.93–1.41)	.201	1.28 (1.04–1.58)	.019
Not answered (N = 5)***
Years as a licensed orthodontist
< 5	87 (46.5)	100 (53.5)	1
6–20	206 (42.7)	276 (57.3)	0.92 (0.76–1.11)	.369
> 20	192 (48.6)	203 (51.4)	1.04 (0.87–1.26)	.641
Training/course for using CBCT imaging
No	250 (37.2)	422 (62.8)	1	< .001	1	< .001
Yes	206 (58.5)	146 (41.5)	1.57 (1.38–1.80)		1.48 (1.30–1.69)
Country
Belgium	21 (46.7)	24 (53.3)	3.03 (1.65–5.57)	< .001	3.02 (1.62–5.64)	.001
Brazil	402 (53.3)	352 (46.7)	3.47 (2.05–5.86)	< .001	3.19 (1.85–5.51)	< .000
Canada	22 (50.0)	22 (50.0)	3.25 (1.79–5.91)	.421	2.94 (1.59–5.43)	.001
Romania	12 (15.4)	66 (84.6)	1		1
USA	28 (19.6)	115 (80.4)	1.27 (0.69–2.36)	.444	1.20 (0.63–2.28)	.585

PR, prevalence ratio; 95%CI, confidence interval of 95%

^*Adjusted for age, training, and country variables

^**Wald test

^***Not included in the analysis

Knowledge of technical parameters

Training/course in the use of CBCT imaging in orthodontic practice was reported by 354 participants (32.7%), namely Belgium, 10 (22.2%); Brazil, 270 (37.5%); Canada, 18 (42.9%); Romania, 16 (20.8%); and USA, 40 (27.2%). CBCT training had been performed as part of a specialization course in both the overall sample (43.0%) and in Brazil (49.4%), and during dedicated CBCT courses in other countries (78.5% on average).

When asked about voxel size determination, most respondents (58.4%) declared, “I do not specify.” Small voxel size (< 0.3 mm) was the option chosen by 7.3% of respondents and “depends on clinical application” by 27%. As regards FOV, the most frequently found answers were “specific area” (60%) and “depends on clinical application” (44.8%) in all countries studied. The option “full skull,” which means a large FOV, was chosen by 3.2% of respondents from all countries, but included none from Belgium nor Romania. Most chose the option “I do not specify” (45.5%) from Romania.

Responses on referring or not to an oral and maxillofacial radiologist (OMR) for an image interpretation report are shown in Fig. 3. Canadian (36%) and Belgian (40.5%) respondents presented the highest frequencies of responses as “always refer to an OMR to interpret the CBCT images,” while no orthodontist from Romania indicated this category.

Fig. 3.

Proportion of orthodontists, in each country, who refer the CBCT to an oral and maxillofacial radiologist

High costs were indicated out as the reason for not prescribing CBCT imaging by most of the overall sample (55.4%), mainly for Brazilian (63.3%) and Romanian (52.6%) respondents. Concerns about ionizing radiation dose were raised by 26.3% of total respondents.

When asked, “When you refer patients for CBCT imaging, do you inform them about the relative ionizing radiation dose?,” the answer “never” was the most frequent (47.0%), except for Romanians, whose response was mainly “sometimes” (44.6%). Other responses in the overall sample were “sometimes” (29.2%) and “always” (16.9%).

The open-ended question regarding additional thoughts on the prescription of CBCT imaging in orthodontic practice was answered by 211 respondents (10 in Belgium, 111 in Brazil, 13 in Canada, 10 in Romania, and 67 in the USA). Some responses mentioned routinely prescribing CBCT imaging, even for children. However, many respondents were very conscientious, associating CBCT with high ionizing radiation dose and pointing out the method’s potential overuse. Others reported the use of CBCT imaging as a marketing tool by colleagues. The need for more knowledge about CBCT imaging was also mentioned.

Discussion

To our knowledge, this is the first multi-country study that assesses aspects of the prescription of CBCT imaging by orthodontists. The results are essential for orthodontists and oral and maxillofacial radiologists as they provide relevant information on the use of CBCT guidelines and their application in routine orthodontics settings. Taking CBCT imaging training does not seems to influence professionals’ adherence to CBCT imaging guidelines. The influence of such factors also highlights the need for orthodontics and OMR to establish joint efforts to unify and standardize worldwide consensus guidelines on the prescription of CBCT imaging to optimize the understanding of a potential balance between the risks and benefits of this imaging tool.

Concerns on the indiscriminate use of CBCT imaging have led to a mobilization of European countries, and the generation of a collaborative project—The SEDENTEXCT consortium in 2012, which is still being followed worldwide [3]. It was created to gather information for the rational use of CBCT imaging in dentistry and present guidelines based on scientific evidence [17]. In orthodontics, the consortium established that CBCT imaging was justified only for evaluating cleft palate, compromised tooth locations, evaluation of impacted teeth, and degree of tooth resorption related to nearby impacted teeth. The responders to the present survey largely followed these criteria. The results showed that CBCT imaging is broadly prescribed in orthodontics, with only 14% of total responders not using this imaging modality in their clinical practice. Beyond the guidelines [3, 7], these results showed a growing prescription in cases of TMJ and upper airway evaluation. As a general basic principle, the international organizations’ guidelines do not recommend CBCT imaging as a routine examination in orthodontics [3, 7, 17]. However, there are growing suggestions of the possible added value of CBCT imaging in other specific cases [1, 2, 8, 10, 18–23]. To fulfill these guidelines’ role, the updating of these recommendations should be examined, considering the recent related research progress [6].

One interesting finding of the present study was that many respondents from all countries had a conservative approach when prescribing CBCT imaging. In responses to the open question, some orthodontists stressed that it is a diagnostic modality that should be indicated only when necessary. This should be interpreted with caution since orthodontists concerned about the risk of CBCT may be more prone to respond. In addition, possible differences between responders and non-responders were not analyzed since data on the characteristics of the non-responders was not available. We should also consider that many non-specialized clinicians that practice orthodontics also prescribe CBCT. In general, most of the respondents were male. Female orthodontists in the USA were more eager to participate in the survey, even though two third of the AAO members are male. Response rate among US orthodontists was 80.9% of the females versus 19.1% of the males. On the other hand, in Romania, men orthodontists were more prone to answer the survey, even if orthodontists in Romania are majority women (84%). A critical guideline carried out by the American Academy of Oral Radiology (AAOMR) [7] established four general strategies for CBCT imaging prescription. One strategy is the prescription of CBCT imaging only when other 2D imaging diagnostic options are insufficient. However, in the respondents’ free opinions, noteworthy and worrying answers included remarks such as “I recommend CBCT for all cases over 10 years” and “I recommend for all cases over 12 years.” These comments indicate that some professionals may prescribe CBCT imaging as a screening exam, even for children, which breaches international guidelines and runs counter to the current literature [3, 7, 10, 11, 24]. The AAOMR guidelines [7] also suggest restricting the field of view (FOV) to the area of interest instead of the large FOV often prescribed in orthodontics. Only 2.3% of total respondents in this survey reported routine use of the full skull FOV. A limited FOV selection represents a significant reduction in the patient’ ionizing radiation [16, 25].

Another principle recommended by the AAOMR [7] agrees with the recently released National Council on Radiation Protection and Measurements Report (NCRP 177) [26] which suggests using the lowest resolution possible of the region of interest without prejudicing the diagnostically useful definition. A lower resolution scan represents less ionizing radiation, mostly related to an increase in voxel size. Such an imaging protocol parameter is very technical and even unknown according to some of the orthodontists’ responses. CBCT imaging prescription was also investigated among general dentists, and showed excessive prescription of CBCT imaging, thereby resulting in unnecessary radiographic exposure [27]. In all countries, most responses indicated that voxel size was not specified. This result raises a vital concern about the responsibility of the professional in terms of patient radiologic imaging. The technical parameters for acquiring CBCT images should be part of orthodontic residency training and education. AAOMR recommendations [7] also address the unnecessary taking of conventional radiographs before the CBCT exam in cases where tomography is an absolute indication. Interestingly, the survey presented responses from orthodontists who seemed to be unaware of this recommendation, as noted in open responses such as “when I request initial exams, including 2D exams such as periapical, lateral cephalometric and panoramic, I start with the simplest, but depending on what I find, CBCT may be requested.” Although such a response could appear to be good practice, when the orthodontist’s initial clinical exam determines the possible need for 3D imaging, an initial prescription of CBCT should follow instead of a series of conventional imaging modalities which might eventually still need a CBCT image.

The frequency with which orthodontic professionals prescribe CBCT imaging is crucial information retrieved by this study. This dental specialty mainly treats patients throughout the stages of their craniofacial development. The rational prescription of CBCT in orthodontics was reported earlier [28], pointing out the As Low As Reasonably Achievable (ALARA) principles. Nowadays, the evolution of the ALARA concept into ALADA (As Low As Diagnostically Acceptable) [29] and ALADAIP (As Low as Diagnostically Acceptable being Indication-oriented and Patient-specific) [14] demands continuous knowledge updating on the part of orthodontists. The practical application of these principles reinforces the need to balance the risks and benefits according to each clinical situation and patient age.

The optimization of CBCT imaging exposure in children and adolescents has been reported by a multicentre and multidisciplinary group of European researchers [13]. DIMITRA (dentomaxillofacial pediatric imaging: an investigation towards low-dose radiation-induced risks) consortium addressed the importance of limiting FOV and individualizing the orthodontic prescription of advanced imaging. It was noted that when asked which FOV they use, many orthodontists do not specify.

The differences between countries following the CBCT prescription guidelines in this study raise questions about which factors are associated with these differences. Brazil was the country with the lowest prevalence of CBCT imaging prescription. The higher cost of 3D exams was pointed out as a reason for not prescribing CBCT imaging. That could be explained by the fact that orthodontists in Brazil do generally not have CBCT scanners in their practices and request CBCT scans from an external radiology clinic. The prevalence of less prescription in Brazil could also be explained by an economic factor, as Brazil is economically less developed than Belgium, Canada, or the USA. Brazil and Romania present the lowest Human Development Index among all the countries studied. Other reasons could be related to regional diversity. In Brazil, the economic reality of its different states is strikingly diverse. In this study, it was not feasible to analyze each country’s regions individually. Further studies are needed to determine whether differences in CBCT imaging prescription could be related to contrasts between states or regions in each country or socioeconomic status of individual patients.

Orthodontists’ CBCT imaging training was associated with a higher prevalence of prescribing CBCT imaging following the guidelines, 48% more than orthodontists without CBCT training (Table 3). One possible explanation could be the knowledge of many advantages of CBCT images which make diagnosis more straightforward, such as enhanced anatomical definition, proportion, and size. Another aspect is the advantage of 3D images in orthodontic communication. It is undeniable that 3D evaluation leads to a more thorough explanation of a professional diagnosis and helps patients understand their treatment, using several tools, such as 3D models, 3D superimpositions between different times 1:1-dimension scale. However, that should not be the reason for prescribing CBCT, and such benefits must be balanced with the stochastic risks of ionizing radiation. In an ideal scenario, additional CBCT imaging training would improve decision-making regarding the prescription of advanced imaging technology, and not an automatic increase in 3D imaging requests.

The proportion of orthodontists in each country who refer their scans for a licensed radiologist’s interpretation showed the differences between clinical approaches in diagnosis using CBCT imaging. Belgium (41%), Canada (36%), and the USA (30%) were the countries that most refer to a radiologist for image interpretation of CBCT. Romania showed the lowest frequency of referred image interpretation to oral radiologists (29.7%). The availability of oral radiologists could account for such differences. While oral radiology is a specialty per se in countries such as Brazil, Canada, and the USA, Romania’s radiologist is trained in general radiology. In maxillofacial radiology, the formal training involves CT and MRI, but not CBCT. In Romania, training in radioprotection is a prerequisite for operating radiological equipment in private practice. However, specific training for prescribing CBCT is neither included in the dental curriculum nor in the postgraduate orthodontics programs. Clinicians acquire their knowledge mainly from the private sector through courses and lectures.

In Belgium, it is the responsibility of the person taking the CBCT scan to write a report. A radiologist interprets all the images acquired in a hospital as part of standard procedure. In Canada, CBCT imaging certification is mandatory for dentists interested in acquiring a scanner and its operators. Patients should be informed of the limitations and benefits of CBCT imaging, and informed consent should be signed accordingly. In the USA, no CBCT imaging certification requirements exist, and the CBCT imaging recommendations were endorsed only by the AAOMR but not by the American Association of Orthodontists.

There were some differences between countries concerning the average years of professional experience of orthodontists. A tendency towards the more significant CBCT imaging prescription was identified in a younger orthodontic population in Romania. This contrasts with a more conservative approach which more experienced professionals might have (i.e., the higher percentage of orthodontists with more than 10-year professional experience in Belgium, Canada, Brazil—countries where CBCT imaging is less prescribed).

One of the limitations of this study refers to the external validity of the results, since a convenience sample was used and our results cannot be generalized to other contexts. Further studies are necessary to check for the global worldwide use of CBCT in orthodontic practice, both in relation to CBCT prescription and its direct use. Also, it was not possible to identify the differences between responders and not responders.

These results highlighted the need for CBCT training, especially for professionals who do not support specialized image interpretation. Furthermore, the problems surrounding CBCT imaging in an orthodontic practice involve more than the prescription itself. They also involve radiation dose, costs, image storage, user-friendly viewer sources, and proper training. This study also raised questions about financial costs and standard care regulations of each country as influencing factors associated with CBCT imaging prescription. Despite the continuous evolution of the literature on CBCT use, this study showed that the current international guidelines are not being fully utilized in their role of guiding orthodontists in clinical practice. These findings have implications for clinical practice and dental education. Courses and training in orthodontics should include the recommended CBCT prescription and guidelines.

Conclusions

This multi-country study suggests differences between the selected countries in terms of CBCT prescription. It also suggests the need to improve CBCT imaging education and training on radiation safety, precautions, and protection for improved CBCT prescription in orthodontics. Orthodontists in different countries often prescribe CBCT imaging for dental abnormalities, surgical cases, and dentofacial deformities, as suggested by the international guidelines, and also for upper airway and TMJ evaluation. Orthodontists with prior training/certification seem to follow imaging guidelines more closely. There were no significant associations between gender when prescribing CBCT imaging.