A comparison of the horizontal and vertical bitewing images in detecting approximal caries and interdental bone loss in posterior teeth: A diagnostic accuracy randomized cross over clinical trial

Abstract

Background/purpose

Horizontal bitewing radiographs are widely and frequently used in dentistry and are very reliable in diagnosing proximal caries and interproximal alveolar bone level. However, it is challengeable in detecting interproximal root caries, horizontal and/or vertical alveolar bone loss, and furcation involvements. The aim of this article was to assess the accuracy of vertical bitewing images in the diagnosis of caries and alveolar bone level compared to the horizontal bitewing technique.

Materials and methods

Each one of the 20 patients had eight bitewing radiographs to get four horizontal bitewing (control) and four vertical bitewing (experimental) images for the same posterior area; a steel wire (3 mm) was used on the sensor plate to help measure the magnification later on. The radiographs were processed digitally and were evaluated for caries by two expert restorative specialists and for bone loss by two experienced periodontists. They were also compared to the “gold standard,” which is using of both clinical and radiographic examination for diagnosis. They were blinded to each other during images evaluation.

Results

Of the 20-patient sample size, 70% were male and 30% were female, with a mean age of 29.9. The average number of radiographs taken to achieve four standard bitewing radiographs was 5.9 ± 1.7 for vertical bitewings and 5.3 ± 1.3 for horizontal bitewing radiograph. The measurements from the cementoenamel junction (CEJ) to the level of crestal bone didn't show a significant difference between the horizontal and vertical bitewing radiographs. The detection of furcation area in the molar teeth was much higher in the vertical bitewing (100%) compared to the horizontal bitewing (57.5%) (P < 0.0001). Conclusion: The vertical bitewing radiograph has the upper hand over the horizontal bitewing radiograph in the detection of furcation involvement, caries detection, and alveolar bone loss. Therefore, it is highly recommended to use vertical bitewing in caries and patients with periodontal disease rather than the conventional horizontal bitewing.

Introduction

Intraoral radiographs are essential tools for the diagnosis of oral lesions. One of the most accurate X-ray types is intraoral bitewing radiographs, as they are a very valuable method in detecting proximal decay and evaluating periodontal status. Because of an angle of projection through the interproximal spaces, allowing the contact areas to be explored, bitewing radiographs are successful in proximal caries and calculus detection.¹ They are also very useful in the assessment of the anatomical level of crest of the alveolar bone. This can be measured due to the parallelism between the long axis of the teeth and the sensor plate, which makes the beam of the X-ray fall perpendicularly on the tooth, giving an accurate reading on the film. ^2,3 The long axis of the bitewing plate can be positioned horizontally or vertically.

Horizontal bitewing radiographs are widely and frequently used in dentistry and are very reliable in diagnosing proximal caries and interproximal alveolar bone level in which there is slight bone loss .⁴ In comparison, vertical bitewing radiographs allow more of the root area to be seen, and thus are particularly useful diagnostic tools for detecting interproximal root and coronal caries, horizontal and/or vertical alveolar bone loss, and furcation involvements.⁵ Visualization of the crestal bone is important for the assessment of periodontitis, and vertical bitewing radiographs are increasingly valuable when patients have moderate to severe horizontal or vertical periodontal bone loss.^6,7 Thus, in response to the research question (“Are vertical bitewing radiographs more effective and more comprehensive for the diagnosis of periodontal bone loss and equally or more effective in visualizing and diagnosing dental caries?”), the authors’ hypotheses were vertical bitewing radiographs are more effective than the horizontal bitewing technique in diagnosing interproximal caries and in diagnosing alveolar bone loss from periodontal diseases. The aim of this article is to assess the accuracy of vertical bitewings in the diagnosis of caries and alveolar bone level compared to the horizontal bitewing technique for patients over the age of 18 at the Faculty of Dentistry, King Abdulaziz University (KAUFD).

Materials and methods

Study design

This study is a randomized crossover clinical trial with two arms and 1:1 ratio, conducted at the Faculty of Dentistry, King Abdulaziz University (KAUFD). It was conducted according to the guidelines of the Declaration of Helsinki, was approved by the ethical research committee at KAUFD (#053-06-17), and was registered at clinicaltrials.gov (NCT04341636) and followed CONSORT checklist (Fig. 1).

Figure 1.

CONSORT flowchart.

Study population

To be included in this study, patients were at least 18 years of age and were being treated in KAUFD dental clinics. The patients met the requirements of the KAUFD clinic for treatments, had a comprehensive treatment plan, were free of acute periodontal infection, were not smokers, and their opposing teeth were not missing, with the presence of posterior teeth and canine in each quadrant.

Patients excluded from the study included¹ those that carried a disease that could complicate the treatment planned for any procedures in the KAUFD clinics,² lactating or pregnant females,³ those with uncontrolled malignant tumors, and⁴ individuals refusing to contribute in the study. Patients were also excluded if they had a prosthesis that blocked the caries boarder or cementoenamel junction (CEJ), had a gagging reflex that complicated taking X-rays, or lacked the ability to open his or her mouth.

X-ray exposure

A convenient sample of twenty patients were randomly placed into two groups (10 patients per group) to get four diagnostic horizontal bitewing (control) images, and four diagnostic vertical bitewing (experimental) images for the same posterior area using a computer randomization technique random.org (Randomness and Integrity Services Ltd, Premier Business Centres, Dublin, Ireland). All steps of randomization, patient enrollment and assigned intervention were conducted by ZN. This study was an open label because it is difficult to blind the X ray to the participants and investigators. A crossover was then conducted for both groups. X-rays were taken using a specific type of film holder (BiteWing Intraoral Dental Sensor Holder Tab, Dentsply Sirona, Charlotte, NC, USA) and a stainless steel orthodontic wire was cut into equal 3-mm pieces in order to be fixed on the X-ray sensor plate (Kodak RVG 6100 size 2 sensor, Carestream Dental, Atlanta, GA, USA). The film holder aimed to parallelize the orthodontic wire to the long axis of the tooth being captured. This steel wire helped to adjust for measurements of magnification later on (Fig. 2). The X-ray tube was in the same distance in all radiographs (4 mm) taken from the sensor plate. Fig. 2 shows the horizontal and vertical bitewings for the same area with a 3-mm steel wire.

Figure 2.

a. Horizontal bitewing with a 3 mm steel wire, and b. Vertical bitewing with a 3 mm steel wire.

Caries detection

All of the 160 bitewing radiographs were evaluated for caries by two expert restorative specialists. The evaluators were familiar with the criteria of the rating scale prior to the evaluation sessions. The following five-point confidence scale was used: 1—decay definitely absent; 2—decay probably absent; 3—equal chance of decay being absent or present; 4—decay probably present; 5—decay definitely present.^4,8

If caries was detected, a second five-point scale was used as follows: 1—decay mostly missing (>25%); 2—decay slightly present (25–50%); 3—about half of the borders are present; 4—decay probably clear (most borders are present); 5—decay definitely clear (all borders are present).^4,8

The observers were unaware of each other while evaluating the bitewing images. They also weren't told which bitewing belonged to which patient. All radiographs were assessed in a random sequence, and the evaluation process was repeated again in two weeks. Intra- and inter-observer reliability were assessed with the examination of all X-ray images obtained from the patients. Time consumed in each evaluation visit was also calculated and documented (time consumed in the first visit was 1 h and 5 min; time consumed in the second visit was 56 min). Evaluations were entered electronically using an online survey after each radiograph.

Bone loss detection

All of the 160 bitewing radiographs were evaluated by two investigators whom had done a calibration twice before conducting the measurements on the real sample to assess the consistency of the measurements and to enhance the reliability and validity of this research. All of the bitewing radiographs were then evaluated for bone level measurements. The stainless steel wire was introduced to create a factor of magnification as discussed by Li et al.⁹ CEJ was considered as reference points. Each tooth was evaluated twice at each site (mesially and distally) and then adjusted using the steel measurement.¹⁰ The measurements were performed using IC Measure INK software. Intra-observer reliability was assessed with the examination of all X-ray images obtained from the patients. Observers used the following five-point confidence scale: 1—furcation definitely missing; 2—furcation probably missing; 3—equal chance of furcation being existent or missing; 4—furcation probably exists; 5—furcation definitely exists.¹⁰

If furcation was detected, a second five-point scale was used as follows: 1—furcation mostly missing (>25%); 2—furcation slightly exists (25–50%); 3—about half of the furcation are existing; 4—furcation probably clear (most borders are existent); 5—furcation definitely clear (all borders are existent).¹⁰

Patient recruitment

Patients were recruited from the patients enrolled in the postgraduate and undergraduate clinics at KAU. Investigators also informed clinic patients and students about the study.

Sample size/power calculations

G∗power system (version 3.1.9, University, Düsseldorf, Germany) was used to determine the sample size. Supposing a mean difference of 2.1 and standard deviation of 5.2 in caries detection among both,^{6, 7} a sample size of 20 patients was suitable to get a Type I mistake rate of 5% and power of 80%.

Statistical analysis

Intra- and inter-observer reliability grades were analyzed with counting of Cronbach's alpha amounts and interclass correlation (ICC) for bone loss measurements. To determine the intra- and inter-observer contract for each image set regarding caries and furcation detection, weighted Kappa coefficients were calculated. The amount of bone loss was evaluated using the Wilcoxon sign-rank test. The existence or missing of decay/furcation was evaluated by the use of the McNemar test. The accuracy of the diagnosis of each imaging modality for caries and furcation detection was analyzed with asymptotic 95% confidence intervals and a receiver operating characteristic (ROC) test. The authors compared vertical and horizontal bitewing with the “gold standard,” which consists of both clinical and radiographic examination. Using the Statistical Package for the Social Sciences (SPSS) statistical program (Version 23.0; SPSS Inc., Chicago, IL, USA), the mean of area under ROC curves (AUC) was counted for each observer for each imaging pattern.

Results

Sample characteristics

The study was conducted on 20 patients (70% male, 30% female) with a mean age of 29.9 ± 11.6. The average number of radiographs taken to achieve four diagnostic standard bitewing radiographs was 5.9 ± 1.7 for vertical bitewing and 5.3 ± 1.3 for horizontal bitewing (Table 1).

Table 1.

Demographic characteristics of study sample.

Variable	N(%)
Age (mean ± SD)	29.9 ± 11.6
Gender
Male	14 (70.0)
Female	6(30.0)
Average of vertical x-ray (mean ± SD)	5.9 ± 1.7
Average of horizontal x-ray (mean ± SD)	5.3 ± 1.3
Type of Jaw
Mandible	50(62.5)
Maxilla	30(37.5)
Type of tooth
1st molar	23(28.75)
2nd molar	17(21.25)
1st premolar	12(15.0)
2nd premolar	28(35.0)

SD, Standard deviation.

To enhance the reliability of the research, the authors distributed their sample between both jaws; the number of mandibular teeth were 50, while the maxillary teeth were 30 (62.5% mandibular teeth, 37.5% maxillary teeth). The involved types of teeth were 23 first molar teeth (28.75%), 17 s molar teeth (21.25%), 12 first premolar teeth (15%), and 28 s premolar teeth (35%) (Table 1).

Radiographic characteristics and bone loss measurements

The mesial and distal radiographic measurements from CEJ to the level of crestal bone didn't show a significant difference between horizontal and vertical bitewing radiographs (P > 0.05), although the horizontal bitewing measurements were less in both mesial and distal surfaces (Table 2). However, in molar teeth, the detection of furcation area was much higher in vertical bitewing (100%) compared to horizontal bitewing (57.5%) (Table 2).

Table 2.

Bone loss measurements of vertical and horizontal bitewings.

Variable	Vertical N = 80	Horizontal N = 80	P value
Mesial	1.06 ± 0.78	0.99 ± 0.61	0.768
Distal	1.30 ± 0.91	1.26 ± 0.78	0.847
Total	1.15 ± 0.69	1.12 ± 0.63	0.930

∗P value < 0.05.

In the characteristics of obtained bitewings, both methods did not show any difference, with the exception of furcation, which was 100% detected in the vertical bitewing compared to the horizontal bitewing (P < 0.001) (Table 3).

Table 3.

Characteristics of vertical and horizontal bitewings.

Variable	Vertical N = 80	Horizontal N = 80	P value
Overlap all	70(87.5)	70(87.5)	1.000
Overlap focused tooth	30(38.5)	23(29.5)	0.310
Both Jaws furcation	40(100.0)	13(35.9)	<0.001a

^aP < 0.05.

Caries and furcation detections

For more accurate confirmation regarding the reliability of both methods in caries and furcation detection, statistical tests were included (e.g., Kappa coefficients) for intra-observer readings among the first and second reading, and for inter-observer reliability between the two observers. It was shown that the inter- and intra-reliability of the vertical bitewings were higher than those for the horizontal one in caries detection, while it was the opposite in furcation detection (Table 4).

Table 4.

Intra-observer agreement calculated for each observer by image type and Inter-observer kappa coefficients among observers for the first and second readings.

	Weighted Kappa-SE
	Observer 1 First reading –second reading	Observer 2 First reading –second reading	Observer 1- Observer 2 First reading	Observer 1- Observer 2 Second reading
Caries
Horizontal bitewing	0.379–0.072	0.475–0.073	0.246–0.063	0.322–0.062
Vertical bitewing	0.469–0.070	0.510–0.068	0.249–0.061	0.314–0.063
Furcation
Horizontal bitewing	0.382–0.095	0.414–0.095	0.285–0.092	0.429–0.093
Vertical bitewing	0.234–0.085	0.389–0.101	0.280–0.089	0.298–0.093

SE, standard error.

The Az values for different evaluators were calculated for caries detection in Table 5. The Az values for vertical bitewings were higher compared to horizontal bitewing. Figure 3, Figure 4 show the ROC curves for both observers for the second readings for each image type. It showed vertical bitewings have better sensitivity to detect caries compared to horizontal bitewings.

Table 5.

Az values, their standard errors, 95% confidence intervals and significance levels (P-value) for each observer.

	Observer 1		Observer 2
	First reading	Second reading	First reading	Second reading
Horizontal bitewing
AZ-SE	0.648–0.063	0.569–0.064	0.538–0.070	0.610–0.066
95% CI	0.524–0.772	0.434–0.704	0.400–0.675	0.480–0.740
P value	0.029a	0.310	0.579	0.105
Vertical bitewing
AZ-SE	0.722–0.062	0.849–0.043	0.888–0.041	0.861–0.042
95% CI	0.600–0.844	0.765–0.934	0.807–0.969	0.778–0.944
P value	0.001a	<0.001a	<0.001a	<0.001a

Az, area under the receiver operating characteristic curve; SE, standard error; CIs, confidence intervals.

^aP < 0.05.

Figure 3.

Receiver operating characteristic (ROC) curves for Observer 1 for the second reading of caries detection method.

Figure 4.

Receiver operating characteristic (ROC) curves for Observer 2 for the second reading of caries detection method.

Discussion

The results of this article, which compared sets of vertical and horizontal bitewing images taken from the same patients, showed that vertical bitewings were superior in assessing the caries, furcation involvement, and interproximal alveolar bone loss. Hence, the vertical contained significantly more information regarding the extent of caries lesion and furcation and provided greater visualization of apico-coronal dimension than the horizontal bitewings. These results were confirmed in Hong et al., 2020 results, which used retrospective analysis without any standardize measurements for the X-rays.

Accurate diagnosis is critical in any medical field and the treatment is highly based on the diagnosis. In standard oral radiographs, the posterior interproximal bone levels are important for periodontal and peri-implant diagnoses based on the 2017 Classification.^{6, 7} Specifically, it will help with staging and grading of periodontitis, as well as those with slight-to-moderate periodontal bone loss, which can be difficult to catch in conventional horizontal bitewings.

One of the challenges with the vertical bitewing is the difficulty in catching all posterior teeth in two bitewings (as can be done with conventional horizontal bitewings). Therefore, it is recommended to conduct three vertical exposures for diagnostic purposes if a patient has second or third molars, due to the narrower horizontal dimension of the sensor than when it is placed vertically.⁷ However, this narrow diameter may help to minimize the superimposition.

Another challenge is the total number of exposures taken to get an acceptable diagnostic bitewing—it was slightly higher for vertical compared to horizontal. However, both methods needed more than five exposures to reach to the requirement for four sets of X-rays. This could be due to the inability of the horizontal bitewing to precisely expose the interproximal alveolar bone levels or apico-coronal dimension of the crowns, possibly due to the difficulty of placing the sensor in the vertical position. However, it can be argued that vertical had higher exposure and difficulty of placement, because the investigators did not have enough training and experience in its application compared to the conventional horizontal bitewing. This result contradicts a previous study where they found the opposite to be true.^{6, 7} However, in that study, they used a retrospective analysis for X-ray records. In another study, the horizontal bitewings were better in separating the interproximal contacts than the vertical.¹¹ It should be noted, however, that variation in technique regarding film or sensor position could play a role.

Horizontal bitewing is the standard acceptable radiographic protocol for new or maintenance patients worldwide. However, some universities, such as the University of Pennsylvania, require vertical bitewings for all patients with alveolar bone loss.^{7,12, 13, 14, 15, 16, 17, 18, 19} This study showed vertical bitewings were more effective in assessing caries and periodontal disease.

A limitation of this study is the potential reporting bias in caries, furcation, and alveolar bone that may have been revealed by the investigators due to the difficulty of blinding the X-rays being evaluated. To overcome this bias, several calibrations were conducted to detect caries and alveolar bone loss. Additionally, the investigators did not detect any bone loss and did not use staging and grading tools for periodontal disease diagnosis. Another limitation of the study is that the exposure was done by dental interns, whom may have had little or no prior experience with vertical bitewings.

The vertical bitewing radiograph has the upper hand over horizontal bitewing radiographs in detection of furcation involvement, caries detection, and alveolar bone loss. The total number of exposures taken were comparable between the two methods. Therefore, it is highly recommended to use vertical bitewings in caries and periodontal disease patients rather than conventional horizontal bitewings. However, it is recommended to educate and train the students appropriately regarding how to use the vertical technique.

Declaration of competing interest

The authors have no conflicts of interest relevant to this article.