Individualized periapical radiography determined by clinical and panoramic examination M. Roblin * and A• .Akerblomt "Department of Oral Radiology and 'Department of Endodontics, Centre for Oral Health Sciences, Lund University, Sweden

Received 17 September 1991 and in final form 9 February 1992 The efficiency of panoramic radiography compared with full-mouth periapical examination is an unresolved problem. The diagnostic yield of periapical lesions when the clinical signs and symptoms and the findings from a panoramic radiograph served as the basis for an individualized periapical radiographic examination was studied. Two hundred patients were examined clinically and radiographically. The periapical status was assessed step by step with access to increasing numbers of radiographs. For the clinical examination, the sensitivity was 0.24, the positive predictive value 0.62, the specificity 0.98, the negative predictive value 0.90 and the likelihood ratio for the positive test result 12. For radiographs indicated by the clinical examination plus the panoramic radiograph and selected periapical radiographs, both the sensitivity and the positive predictive value were 0.91, the specificity and the negative predictive value 0.99 each and the likelihood ratio was 91. False findings were twice as frequent in the upper as in the lower arch and particularly found in the incisor and premolar regions. In 30% of the patients no periapical radiograph was needed to supplement the panoramic radiograph. In the other patients, two supplementary periapical radiographs were needed on average. We conclude that the information obtained from the clinical and panoramic examinations supplemented with no more than two periapical radiographs will result in a high diagnostic yield on the periapical status. Keywords: Periapical diseases; decision making; radiography, dental; radiography, panoramic

Dentomaxillofac. Radiol., 1992, Vol. 21,135-41, August

An increase in the use of panoramic radiographs in general dental practice has been re~orted from Great Britain 1.2, the USA3-S and Sweden . The reasons for this have yet to be established, but it is likely that its relative ease of use and significantly lower radiation dose makes it a potential substitute for the traditional full-mouth intraoral examination (FMX). However, this can only be justified if the diagnostic yield is comparable. The results of previous comparative studies of periapical lesions have been contradictory>!'. Furthermore, their conclusions could well be invalidated by improvements in imaging technology. A recent study'? based on ROC analysis failed to demonstrate any overall significant difference. However, for specific regions, such as the maxillary premolar and mandibular molar regions, periapical radiography was definitely superior. Similarly, both techniques have been found to give comparable results for the assessment of marginal bone levels'P-!", with the exception of certain specific regions and those sites with marked bone loss. We recommended as a consequence that panoramic radiography could be used as the primary examination for periodontal disease, and as the basis for any subsequent decision as to the need for supplementary intra-oral radiographs. The present study was undertaken to examine whether a similar approach

could be applied to the assessment of periapical pathology. The diagnostic yield of periapical lesions has been assessed in the situation where the clinical signs and symptoms and a panoramic radiograph served as the basis for an individualized radiographic examination.

Materials and methods Patients This study was based on a sample of 200 patients chosen at random from those attending for total dental care. The mean age for the women was 53.5 years (s.d. ± 5.8) and for the men 51.0 years (s.d. ± 6.3). The mean number of teeth present was 2Z.9 per patient (range 8-32; s.d. ± 5.5) and the distribution is shown in Table I.

Clinical examination This examination was carried out by an endodontist (A.A.) using a mirror and explorer. The following findings were recorded: pulpal exposure, sinus, swelling, pain, tenderness to percussion and palpation, extensive caries, and retained roots. Dentomaxillofac. Radiol., 1992, Vol. 21, August

135

Individualized radiographic examination: M. Rohlin and A . .4.kerblom Table I Distribution of teeth with and without periapical lesions and teeth with and without endodontic treatment (n = 200 patients). The percentage of standing teeth with a periapical lesion is shown in parentheses Mandible

Maxilla

Total Molars

Teeth with lesions With endodontic treatment Without endodontic treatment Total

45 41 86 (14.9)

Premolars

45 54 99 (17.5)

Teeth without lesions With endodontic treatment Without endodontic treatment Total

42 451 493

82 385 467

Total no. of teeth

579

566

Canine

Incisors

Molars

36 65 101 (18.8)

31 65 12 38 43 (11.4) 103 (14.3)

Premolars

36 38 74 (11.3)

Canine

9 6 15 (3.8)

8 25 33 (4.4)

275 279 554 (12.1)

380

16 702 718

431 3589 4020

395

751

4574

43 392 435

91 489 580

29 351

333

89 525 614

376

717

536

654

39 294

Incisors

Radiographic examination

Analysis

Technique. Following the clinical examination, a panoramic radiograph was obtained together with an intra-oral examination which, depending on the number of teeth present, consisted of 6-14 periapical and up to four bitewing radiographs (FMX). The dental Xray unit operated either at 60 kV and 10 rnA (Heliodent EC60, Siemens, Bensheim, Germany) or at 65 kV and 7 rnA (Oralix 65, Philips, Eindhoven, The Netherlands) and had a total filtration of 2 mm Al equivalent. The focus-film distance was 20 ern, E-speed film (Ektaspeed, Eastman-Kodak, Rochester, USA) was used. The panoramic radiographs were obtained with an Orthopantornograph'" Model OP 5 (Palomex Instrumentarium Corp. Helsinki, Finland) with an SR 90/15 FN tube with a stationary anode (Siemens, Erlangen, Germany). The collimator was 0.6-0.9 mm x 39.5 rnrn, and the focal spot-film distance 0.47 m. Screens of speed group 4-5 (Titan 2 HS screens, Siemens, Erlangen , Germany) and Wicor-XRP film, (CEA Strangnas, Sweden) were used.

The total yield from the panoramic radiograph and the FMX was considered as the reference standard as described elsewhere 'V'". Sensitivity, specificity, predictive values and likelihood ratio were calculated according to Weinstein and Fineberg'? for each of the two steps against the reference standard. Thus, in STEP 1 the radiographs indicated by the clinical examination were assessed. In STEP 2, these findings together with those from both the panoramic radiograph and the selected additional periapical radiographs were correlated with the reference standard. The predictive values for periapical lesions were calculated both for the prevalence of 12% found in the present study and a lower prevalence of 7%, which represents the mean of those reported in three other studies !X-20.

Interpretation. All osteolytic or sclerotic periapical lesions were recorded as either definite, uncertain or absent. The assessment of the periapical status was made step by step by a single oral radiologist who had access to increasing numbers of radiographs. 1. The panoramic radiograph was assessed for any teeth which could be associated with the clinical signs and symptoms defined above (Figure 1a). If the radiographic evidence was considered uncertain, then the periapical radiograph of the tooth in question was assessed (Figure 1b). STEP 2. The panoramic radiograph (Figure 1a) was assessed as a whole. Where the status of a particular tooth was considered uncertain, then the periapical radiograph was also assessed (Figure l c, d). STEP

REFERENCE STANDARD. Both the FMX and the panoramic radiograph were assessed together. The total number of periapical radiographs in the FMX was noted for each patient. The additional periapical radiographs making up the FMX are presented for one case in Figure l e.

136

Dentomaxillofac. Radiol., 1992, Vol. 21, August

Results Figure 2 presents an overview of teeth. with clinical signs and symptoms, periapical lesions and endodontic treatment. One hundred and sixty-seven patients had a total of 554 teeth with periapical lesions, of which 47 (9%) were wholly sclerotic. The mean number of teeth with periapical lesions was 2.8 per patient (range 0-12; s.d. ± 2.2). Fifteen per cent of the remaining teeth had been endodontically treated" Table I shows the distribution of teeth with and without periapical lesions as well as teeth with and without endodontic treatment. A periapical lesion was found on 12.1% of the teeth present, with a frequency of 39% on the endodontically treated teeth and 7.2% on those that were not. Half of the teeth (49.6%) with a periapical lesion had received endodontic treatment (Figure 2). Periapical lesions were most frequent on mandibular molars (18.8%) and least so on mandibular canines and incisors (3.8% and 4.4%).

Clinical examination Ninety patients presented with clinical signs and symptoms on 213 teeth. Most patients had one (38 patients) or two affected teeth (32 patients) but five had more than seven teeth involved. Table II presents the

Individualized radiographic examination: M. Rohlin and A. Akerblom

Figure I The radiographs of one patient to illustrate the step-by-step interpretation process. Step 1. Clinical examination revealed the left maxillary lateral incisor and the right mandibular second premolar to have extensive caries. The panoramic radiograph (a) revealed a periapical lesion on the lateral incisor whereas the periapical status of the premolar was assessed as uncertain. The supplementary periapical radiograph (b) confirmed the latter as certain. Step 2. The panoramic radiograph did not reveal any other periapical lesions, but the periapical status around the maxillary and mandibular central incisors was assessed as uncertain. The two supplementary radiographs (c, d) demonstrated a lesion on the right maxillary central incisor, but none on the other teeth. Reference standard. Simultaneous interpretation of the panoramic and full-mouth examination. No additional lesions were observed. Eleven periapical radiographs (e) were needed to complete the full-mouth examination

distribution of teeth together with their different clinical signs and symptoms. The latter were most often found on maxillary incisors and molars as well as premolars of both arches. In relation to the actual number of standing teeth, they were most frequent on mandibular molars.

Table III presents the correlation of the clinical examination with the radiographic reference standard. One hundred and thirty-three out of the total of 554 periapical lesions (24%) were observed on teeth with clinical signs and symptoms, a sensitivity of 0.24. The positive predictive value was 0.62. As shown in Figure Dentomaxillofac. Radiol., 1992, Vol. 21, August

137

Individualized radiographic examination: M. Rohlin and A. Akerblom STEP 2: assessment of radiographs indicated by the clinical examination plus panoramic radiograph plus selected periapical radiographs. Table IV presents the findings of the radiographs indicated by the clinical examination plus the panoramic radiograph and selected periapical radiographs correlated with the radiographic reference standard. Both the sensitivity and the positive predictive value were high, 0.91, indicating that 91% of the findings were true positive and 9% false positive. The specificity as well as the negative predictive values were also very high. As shown in Figure 3a, the sensitivity was somewhat lower for premolars and maxillary incisors than for the other teeth. The specificity varied between 0.97 and 1.0. The positive predictive value was lower for incisors and premolars (Figure 3b). The overall likelihood ratio of a positive test was 91, but was much lower, however, at 29 and 42 respectively, for the maxillary incisors and premolars. In most patients (70%) the panoramic radiograph was considered to be inadequate in some region for a definite diagnosis and in these patients, 284 supplementary periapical radiographs (mean 2 radiographs/ patient; s.d. ± 1.1) were judged necessary. The distribution of these radiographs is shown in Figure 4. Most of them were needed in the incisor region of both arches or the maxillary premolar region. The additional radiographs resulted in the detection of a lesion more frequently in the upper arch (47 radiographs), in the premolar region in particular, whereas the diagnostic yield in the lower arch was low (17 radiographs).

Periapical lesion (n 554)

=

Clinical signs and symptoms (n= 213)

Endodontic treatment (n= 706)

Figure 2 Total number of teeth with clinical signs and symptoms, periapical lesions and endodontic treatment

3a, the sensitivity was somewhat higher for mandibular than for maxillary teeth. The positive predictive value was high for the mandibular molars and low for the incisors of both arches (Figure 3b). The likelihood ratio of a positive test was 12.

Radiological examination STEP I: assessment of radiographs indicated by the clinical examination. A periapical radiograph had to be taken of 26 teeth with clinical signs and symptoms to supplement the panoramic radiograph, and in eight of these there was a positive radiographic finding.

Table II Distribution of teeth with different types of clinical signs and symptoms. The numbers of teeth with both clinical and radiological findings are shown in parentheses Lower arch

Upper arch Molars

Sinus Pain Tenderness to percussion and/or palpation Extensive caries Retained root

Premolars

Canine

Incisors

Molars

Premolars

Canine

3 (1)

4 (3)

1 (1) 1 (1)

1 (1)

Incisors

Total

1 (1) 18 (9)

2 (1)

7 (2)

9 (4) 17 (9) 7 (7)

12 (7) 6 (4) 9 (9)

4 (1) 3 (2) 30)

12 (7) 20 (6) 7 (5)

9 (9) 12 (6) 15 (14)

3 (1) 15 (10) 10 (8)

3 (1) 3 (2)

4 (3) 9 (3) 2 (2)

53 (32) 85 (41) 56 (50)

35 (21)

34 (22)

10 (6)

42 (19)

40 (32)

30 (21)

7 (4)

15 (8)

213 (133)

1.8

2.0

Total Percentage of teeth with clinical signs and symptoms

6.0

6.0

2.7

5.7

4.6

7.5

4.7

Table III Sensitivity, specificity and likelihood ratio of the clinical examination in detecting the periapical lesions revealed by the reference standard. The values in A are based on the prevalence of periapical lesions found in the present study (12%) and in B on a lo\yer prevalence of7% A. Reference standard

Clinical examination

Total Sensitivity: 0.24 Specificity: 0.98 Likelihood ratio for a positive test: 12.0

B. Reference standard

Lesion

No lesion

Total

Lesion

No lesion

Total

+133 -421

80 3940

213 4361

+77 -243

85 4169

162 4412

554

4020

4574

320

4254

4574

Positive predictive value: 0.62 Negative predictive value: 0.90

+. Number of teeth with positive findings; -. number of teeth with negative findings.

138 Dentomaxillofac. Radiol., 1992, Vol. 21, August

Positive predictive value: 0.48 Negative predictive value: 0.94

Individualized radiographic examination: M. Rohlin and A. Akerblom

1.0 -

1.0 -

0

0 x

x

-

.

0

0

a

0

0.8 -

'" >

>> :;; 0.6 'iii c +'

0.2

0

6

0

0.8 -

"

6

6

~

'iii 0

c,

I

I

Canine

Incisors

b

0.2

f0-

I

I

0.0

Premolars

Molars

Canine

Incisors

Molars Figure 3 Tooth group-specific sensitivities (a) and positive predictive values (b) for the clinical examination and the panoramic radiograph plus selected periapical radiographs for diagnosing periapical lesions. Numbers of teeth with and without periapical lesion are shown in Table I and numbers of false findings for each tooth group in Table V. Radiological examination step 2: 6, lower arch; 0, upper arch. Clinical examination: D, lower arch; x, upper arch

0c

66

III

s: c.

'en"

I-

.~

positive and false negative, were twice as frequent in the upper as in the lower arch (Table V). These were most frequent in the maxillary incisor and premolar regions. The majority of the false-positive findings were in the maxillary incisor region, whereas falsenegative findings were more frequent in the maxillary premolar region. The mean number of periapical radiographs per patient needed to make up the FMX was 12 (s.d. ± 1.8): in all, 2401 were required for the total 200 patients.

80 63 60

"tl

46

'"

l-

t;; .~

40

34

c.

26

.!l!

I-

"c. t;;

20

.. c 0

'ii "0

-c

0 ~7>~

~o

..,

Discussion

~..,

~,0e . ..,o~'" 0\7> c,' ~ 07>