- Email: [email protected]
Asymptomatic mandibular third molars: Oral Surgeons' judgment of the need for extraction
J Oral Maxillofac
Surg
50:329-333,1992
Asymptomatic Mandibular Third Molars: Oral Surgeons’ Judgment of the Need for Extraction KERSTIN KNUTSSON, DDS,* BERNDT BREHMER, PtiD,t LEIF LYSELL, DDS, ODONTDR,$ AND MADELEINE ROHLIN, DDS, ODONT DR§ Ten oral surgeons were asked to judge the need for extraction of asymptomatic mandibular third molars. Thirty-six mandibular third molars with equal distribution of angular position, impaction status, and patient’s sex and age were selected. To estimate the consistency of judgment, the 36 cases were duplicated so that, in all, 72 cases were judged. The judgment of the oral surgeons was compared with that of 30 general dental practitioners (GDPs). The number of mandibular third molars the oral surgeons proposed to extract varied from 3 to 21 of 36 teeth. The mean number of molars proposed for extraction was 12 for the oral surgeons and 13 for the GDPs. There was no third molar that all the observers in the two groups agreed should be extracted. About three times as many obsewers in both groups proposed extraction of molars partially covered by soft tissue. The oral surgeons were unanimous in their judgment not to extract 11 molars, and the GDPs were also unanimous in judgment not to extract two of these. The mean intraobsewer agreement within the two groups was comparable, 94% for the oral surgeons and 92% for the GDPs. We conclude that there is a great variation among oral surgeons in their judgment on the need for removal of asymptomatic mandibular third molars. A similar variation in judgment also was observed among GDPs.
is to relate the policy-capturing process to cues characterizing the judgment.’ This approach has been applied in a study on 30 general dental practitioners (GDPs) who judged the need for extraction of asymptomatic mandibular third molars3 It was shown that the judgment whether to extract asymptomatic impacted mandibular third molars differed substantially among the GDPs. In many cases, the GDP is the person who both makes the decision and performs the third molar extraction. Not infrequently, however, after identifying an asymptomatic third molar, the GDP has to decide whether or not to refer the patient to an oral surgeon. The second level in the decision-making process on the relevance of the prophylactic indication for removal is therefore rather often the opinion of the oral surgeon. The aim of this study was to analyze how a group of oral surgeons judge the need for extraction of asymptomatic mandibular third molars. The oral surgeons’ judgment was also compared with that of a group of GDPs.
Clinical judgments have traditionally been considered an intuitive process whereby the clinician combines information about the patient and new observations and test results to make a diagnosis or treatment plan.’ Another way to analyze this judgment process
* Instructor, Department of Oral Radiology, School of Dentistry, Centre for Oral Health Sciences, Lund University, Malmii, Sweden. t Professor, Department of Psychology, University of Uppsala, Uppsala, Sweden. $ Chief Consultant, Department of Oral Surgery, Central Hospital, Kristianstad, Sweden. 0 Associate Professor, Department of Oral Radiology, School of Dentistry, Centre for Oral Health Sciences, Lund University, Malmii, Sweden. Supported by grants from the Swedish Dental Association, Gunnar Bjorlin’s Foundation, and Lund University. Address correspondence and reprint requests to Dr Rohlin: Department of Oral Radiology, School of Dentistry, University of Lund, S-2 14 2 1 Malmb, Sweden. 0 1992 AmericanAssociationof Oral and MaxillofaciaiSurgeons 02782391/92/5004-ooo3$3.00/0
329
EXTRACTIONOF MANDIBULARTHIRD MOLARS
330 Material and Methods CASES
Table 1. Number of Asymptomatic Mandibular Third Molars Proposed to be Extracted by 10 Oral Surgeons (N = 36) Observer
Thirty-six asymptomatic mandibular third molars were selected so as to represent an equal distribution of males and females, age groups, angular positions, and impaction status. Thus, half of the subjects were female patients. Three age groups were selected: 19 to 25 years, 26 to 40 years, and 4 1 to 60 years. The angular position was classified as vertical, mesioangular, distoangular, or horizontal according to the criteria presented by Winter.4 The impaction status was divided into three categories: 1) partially covered by soft tissue, 2) completely covered by soft tissue, and 3) completely covered by bone tissue. To assess the consistency of the judgment, the 36 caseswere duplicated. However, when presented to the observer, the duplicate case was turned to the opposite side compared with the original case. CASE PRESENTATION
Each of the 72 cases was presented as a radiograph, as shown in Figure 1, with a short text informing the observer about the sex and age of the patient, the impaction status of the molar, and that there were no symptoms. The observers were asked to judge whether or not to extract the illustrated third molar. OBSERVERS The 72 cases were sent to 10 oral surgeons at 10 different oral surgery clinics. They were all certified oral surgeons and active members of the Swedish As-
Rofessiond
wsw
EXpaienCe
NO.
w
1 2 3 4 5 6 1 8 9 10
13 20 30 23 12 25 8 21 3 19
Intraobserver Agreement
No. Judged to Extract on 1st Observation
No. Judged to Extract on 2nd Observation
TO
Total
Extract
EVdUatCd
3 4
3 3 10 9 11 16 17 17 17 21
3 3 1 8 11 11 14 15 17 21
36 35 33 35 35 28 32 34 36 36
1
8 12 14 15 15 17 21
sociation of Oral and Maxillofacial Surgeons. One of the authors (LL) selected the observers from the association’s list of members as being oral surgeons with a good professional reputation and with a declared interest in the question of extraction of third molar. They were also selected so as to represent clinics in different geographic regions. There were 3 female and 7 male oral surgeons. Their length of professional experience in oral surgery is presented in Table 1. They had been certified as oral surgeons for 3 to 30 years. Before entering an oral surgery residency, they had had between 2 and 14 years’ experience with general dentistry. The same cases also were judged by another 30 observers who practised as general dental practitioners (GDP). This group of observers has been described in detail in a previous study.3 The coefficient of correlation between years of professional experience and number of molars proposed for extraction was calculated for the oral surgeons and the GDPs, respectively. Results ORAL SURGEONS’ JUDGMENT
FIGURE 1. Woman, 19 years. The molar is partially covered by sot? tissue and in a mesioangular position. Nine of 10 oral surgeons proposed extraction.
The number of mandibular third molars the observers proposed to extract varied from 3 to 21 of 36 teeth (Table 1). There was no third molar that all oral surgeons agreed should be extracted (Table 2). In general, however, the oral surgeons as a group seemed more inclined to propose extraction of third molars partially covered by soft tissue than for other impaction statuses. Thus, 9 of 10 observers were unanimous on the decision to extract one molar, which was partially covered by soft tissue and mesioangularly positioned (Fig 1).
331
KNUTSSON ET AL
Table 2. Number of Oral Surgeons (N = 10) Unanimous on Repeated Observations and With Each Other on Judgment to Extract Asymptomatic Mandibular Third Molars of Different Impaction Status and Position for Patients in Three Different Age Groups Age Group (yr) Impaction Status
Position
19-25
Partially covered by soft tissue
Vertical Mesioangular Distoangular Horizontal
Completely covered by soft tissue
Completely covered by bone tissue
26-w
41-60
8 9 7 7
6 7 5 8
1 3 3 3
Vertical Mesioangular Distoangular Horizontal
5 4 6 2
I 1 0 0
0 0 4 0
Vertical Mesioangular Distoangular Horizontal
8 3 2 2
0 0 4 0
0 0 1 0
FIGURE 2. A, Man, 46 years. The molar is completely covered by soft tissue and in a mesioangular position. All oral surgeons agreed that it should not be extracted. B, Man, 35 years. The molar is completely covered by bone tissue and in a vertical position. All oral surgeons and GDPs agreed that it should not be extracted. C, Man, 60 years. The molar is completely covered by bone tissue and in a vertical position. All oral surgeons and GDPs agreed that it should not be extracted.
Eight of the 10 observers also proposed extraction of one molar in a vertical position completely covered by bone tissue and one partially covered with soft tissue. The observers were unanimous on the decision not to extract 11 molars. All these teeth were completely covered by soft or bony tissue and belonged to patients in the two oldest age groups. Figures 2A, 2B, and 2C show three of these molars. Table 1 presents the intraobserver agreement on whether or not to extract for the 10 oral surgeons. The mean overall intraobserver agreement was 94% There were three oral surgeons ( 1,9, 10) having an individual agreement of 100%. COMPARISON AND
BETWEEN ORAL SURGEONS’
GDPs’ JUDGMENT
The mean number of molars proposed for extraction was 12 for the oral surgeons and 13 for the GDPs. There was no third molar which all oral surgeons or
332
EXTRACTION
OF MANDIBULAR
THIRD MOLARS
Table 3. Percentage of Oral Surgeons (N = 10) and GDPs (N = 30) Proposing Extraction of Asymptomatic Mandibular Third Molars of Different Impaction Status and Position for Patients in the Three Different Age Groups Age Groups (yr) 19-25
26-40
oral SUrgeOLl
41-60 oral
oral GDP
SUrgeOIl
GDP
surgeon
GDP
Position
(a)
(“p)
w)
@)
(W)
(96)
Partially covered by soft tissue
Vertical Mesioangular Distoangular Horizontal
80 90 70 70
37 60 57 83
60 70 50 80
43 77 43 73
10 30 30 30
3 43 30 33
Completely covered by soft tissue
Vertical Mesioangular Distoangular Horizontal
50 40 60 20
10 27 17 30
10 10 0 0
27 I 33 23
0 0 40 0
23 20 53 7
Completely covered by bone tissue
Vertical Mesioangular Distoangular Horizontal
80 30 20 20
50 27 10 30
0 0 40 0
0 10 70 7
0 0 10 0
0 10 20 3
Impaction Status
all GDPs agreed should be extracted (Table 3). However, about three times as many observers, both among the oral surgeons and the GDPs, proposed extraction of molars partially covered by soft tissue compared with molars with other impaction status. The GDPs were unanimous in their judgment not to extract two molars, whereas the oral surgeons were unanimous about 11 molars, including the two molars proposed by the GDPs. The two molars that all observers agreed should not be extracted were completely covered by bone tissue and in a vertical position (Figs 2B, C). Although the frequency of proposal for extraction in the two groups was approximately equal (oral sur-
geons, 10; GDPs, 12), the frequency difference between the groups exceeded 30% for some third molars. The oral surgeons were more inclined than the GDPs to propose extraction for molars completely covered by soft tissue and in a distoangular position, as well as molars in a vertical position in the youngest age group. Figure 3A shows one of these molars. On the other hand, the GDPs more often proposed extraction of molars in a distoangular position and completely covered by soft tissue (Fig 3B) or by bone tissue in the 26to 40-year age group (Table 3). The mean intraobserver agreement of the two groups was comparable, 94% for the oral surgeons, with a range of 78% to 100% (K index, 0.53 and l.O), and 92% for
FIGURE 3. A, Woman, 20 years. The molar is partially covered by SOBtissue and in a vertical position. Eight of 10 oral surgeons and 15 of 30 GDPs proposed extraction. B, Woman, 40 years. The molar is completely covered by soft tissue and in a distoangular position. Ten of the 30 GDPs proposed extraction, and all oral surgeons agreed that it should not be extracted.
333
KNUTSSON ET AL
the GDPs, with a range of 69% to 100% (K index, 0.39 and 1.00). There was no significant correlation between the number of years of professional experience and the judgment regarding whether or not to extract either among the oral surgeons or among the GDPs as a group. There was a tendency for the oral surgeons to propose less extractions with increased experience (Y= - .3 1), whereas this tendency was not found among the GDPs (Y = .2 1). Discussion The validity of the prophylactic indication for third molar extraction can be evaluated from different viewpoints. The GDP or “family dentist” follows a limited group of patients over time, giving them comprehensive dental care, whereas the oral surgeon sees a selected group of patients referred because of their need for a specialist’s opinion or surgical treatment. These different clinical working conditions might at least theoretically influence the assessment of indications for different interventions. One may expect the oral surgeons to be those members of the dental profession with the most thorough knowledge and understanding of indications for removal of third molars. Having noted the wide individual differences among 30 GDPs in the judgment whether or not to extract an asymptomatic third molar,3 we found it warranted to examine the judgments of a group of oral surgeons. Our method of selection of observers resembles the so-called Delphi technique,5 where the observers are selected for their knowledge on the subject and standing within the profession. The degree of variation as to whether to extract or not among the oral surgeons is in agreement with studies concerning judgment analyses among other groups of medical experts. ‘s6 Investigators have related this variation to differences in how medical experts weight clinical information in making diagnostic or management decisions. ’ The oral surgeons’ disagreements could be related to differences in the calculated importance of each cue, so that the cues used and the cue weights differed considerably from oral surgeon to oral surgeon. Most important, the variation of their use of cues involves differences in therapeutic (whether to extract or not) decision in each case, not just differences in their use of cues. Some of the oral surgeons and the GDPs showed substantial intraobserver disagreement. This is in agreement with other investigations concerning medical judgments’ and endodontic treatment decisions.’
The mean number of molars proposed for extraction was similar for the two groups of observers, even though the oral surgeons were expected to be more clinically experienced concerning third molar problems. Wigton, Hollerich, and Partid’ had predicted that variation in diagnostic strategies decreased as clinical experience increased. They found, however, that experienced faculty members were as different from one another in their strategies as medical students. The difference in some cases was so great that the dominant cue for one expert was not used at all by another. Reit et al’ also found that the decision strategy of GDPs was similar to that of endodontists concerning endodontic treatment decisions. One reason for the differences in judgment between the observers is the scant knowledge of the natural history of asymptomatic third molars. There are few longitudinal studies of impacted third molars. However, recent reportsg,io showed that less than 10% to 15% of impacted third molars will develop a pathological change. It is, therefore, difficult to predict that the patient benefit outweighs the surgical risk when a third molar is extracted. There is clearly also a need to study how patients evaluate prophylactic extraction of a mandibular third molar. The most important decision maker will, after all, be the patient. References 1. Wigton RS: Application of judgment analysis and cognitive feedbacks to medicine, in Brehmer B, Joyce CRB (eds): Human Judgment. The SJT view. Amsterdam, Elsevier, 1988, pp 221-245 2. Brehmer A, Brehmer B: What have we learned about human judgment from thirty years of policy capturing?, in Brehmer B, Joyce CRB (eds): Human Judgment. The SJT view. Amsterdam, Elsevier, 1988, pp 75-l 14 3. Knutsson K, Brehmer B, Lysell L, et al: General dental practitioners’ evaluation of the need for extraction of asymptomatic mandibular third molars. Community Dent Oral Epidemiol 1992 (accepted) 4. Winter GB: Principles of Exodontia as Applied to the Impacted Third Molar. St Louis, MO, American Medical Books, 1926 5. Linstrone HA, Turoff HM: Introduction, in Linstrone HA, Turoff HM (eds): Delphi method: Techniques and Applications. Reading, MA, Addison-Wesley, 1975, pp 3-12 6. Reit C, Grijndahl H-G: Management of periapical lesions in endodontically treated teeth. Swed Dent J 8: 1, 1984 7. Reit C, Grondahl H-G, Engstrijm B: Endodontic treatment decisions. A study of the clinical decision-making process. Endod Dent Traumatol 1:102, 1985 8. Wigton RS, Hollerich VL, Partil KD: How physicians use clinical information in diagnosing pulmonary embolism: An application of conjoint analysis. Medical Decision Making 6:2, 1986 9. Stanley HR, Alattar M, Colett WK, et al: Pathological sequelae of “neglected” impacted third molars. J Oral Path01 17:113, 1988 10. Ahlqwist M, Grondahl H-G: Prevalence of impacted teeth and associated pathology in middle-aged and older Swedish women. Community Dent Oral Epidemiol 19:116, 1991
Surg
50:329-333,1992
Asymptomatic Mandibular Third Molars: Oral Surgeons’ Judgment of the Need for Extraction KERSTIN KNUTSSON, DDS,* BERNDT BREHMER, PtiD,t LEIF LYSELL, DDS, ODONTDR,$ AND MADELEINE ROHLIN, DDS, ODONT DR§ Ten oral surgeons were asked to judge the need for extraction of asymptomatic mandibular third molars. Thirty-six mandibular third molars with equal distribution of angular position, impaction status, and patient’s sex and age were selected. To estimate the consistency of judgment, the 36 cases were duplicated so that, in all, 72 cases were judged. The judgment of the oral surgeons was compared with that of 30 general dental practitioners (GDPs). The number of mandibular third molars the oral surgeons proposed to extract varied from 3 to 21 of 36 teeth. The mean number of molars proposed for extraction was 12 for the oral surgeons and 13 for the GDPs. There was no third molar that all the observers in the two groups agreed should be extracted. About three times as many obsewers in both groups proposed extraction of molars partially covered by soft tissue. The oral surgeons were unanimous in their judgment not to extract 11 molars, and the GDPs were also unanimous in judgment not to extract two of these. The mean intraobsewer agreement within the two groups was comparable, 94% for the oral surgeons and 92% for the GDPs. We conclude that there is a great variation among oral surgeons in their judgment on the need for removal of asymptomatic mandibular third molars. A similar variation in judgment also was observed among GDPs.
is to relate the policy-capturing process to cues characterizing the judgment.’ This approach has been applied in a study on 30 general dental practitioners (GDPs) who judged the need for extraction of asymptomatic mandibular third molars3 It was shown that the judgment whether to extract asymptomatic impacted mandibular third molars differed substantially among the GDPs. In many cases, the GDP is the person who both makes the decision and performs the third molar extraction. Not infrequently, however, after identifying an asymptomatic third molar, the GDP has to decide whether or not to refer the patient to an oral surgeon. The second level in the decision-making process on the relevance of the prophylactic indication for removal is therefore rather often the opinion of the oral surgeon. The aim of this study was to analyze how a group of oral surgeons judge the need for extraction of asymptomatic mandibular third molars. The oral surgeons’ judgment was also compared with that of a group of GDPs.
Clinical judgments have traditionally been considered an intuitive process whereby the clinician combines information about the patient and new observations and test results to make a diagnosis or treatment plan.’ Another way to analyze this judgment process
* Instructor, Department of Oral Radiology, School of Dentistry, Centre for Oral Health Sciences, Lund University, Malmii, Sweden. t Professor, Department of Psychology, University of Uppsala, Uppsala, Sweden. $ Chief Consultant, Department of Oral Surgery, Central Hospital, Kristianstad, Sweden. 0 Associate Professor, Department of Oral Radiology, School of Dentistry, Centre for Oral Health Sciences, Lund University, Malmii, Sweden. Supported by grants from the Swedish Dental Association, Gunnar Bjorlin’s Foundation, and Lund University. Address correspondence and reprint requests to Dr Rohlin: Department of Oral Radiology, School of Dentistry, University of Lund, S-2 14 2 1 Malmb, Sweden. 0 1992 AmericanAssociationof Oral and MaxillofaciaiSurgeons 02782391/92/5004-ooo3$3.00/0
329
EXTRACTIONOF MANDIBULARTHIRD MOLARS
330 Material and Methods CASES
Table 1. Number of Asymptomatic Mandibular Third Molars Proposed to be Extracted by 10 Oral Surgeons (N = 36) Observer
Thirty-six asymptomatic mandibular third molars were selected so as to represent an equal distribution of males and females, age groups, angular positions, and impaction status. Thus, half of the subjects were female patients. Three age groups were selected: 19 to 25 years, 26 to 40 years, and 4 1 to 60 years. The angular position was classified as vertical, mesioangular, distoangular, or horizontal according to the criteria presented by Winter.4 The impaction status was divided into three categories: 1) partially covered by soft tissue, 2) completely covered by soft tissue, and 3) completely covered by bone tissue. To assess the consistency of the judgment, the 36 caseswere duplicated. However, when presented to the observer, the duplicate case was turned to the opposite side compared with the original case. CASE PRESENTATION
Each of the 72 cases was presented as a radiograph, as shown in Figure 1, with a short text informing the observer about the sex and age of the patient, the impaction status of the molar, and that there were no symptoms. The observers were asked to judge whether or not to extract the illustrated third molar. OBSERVERS The 72 cases were sent to 10 oral surgeons at 10 different oral surgery clinics. They were all certified oral surgeons and active members of the Swedish As-
Rofessiond
wsw
EXpaienCe
NO.
w
1 2 3 4 5 6 1 8 9 10
13 20 30 23 12 25 8 21 3 19
Intraobserver Agreement
No. Judged to Extract on 1st Observation
No. Judged to Extract on 2nd Observation
TO
Total
Extract
EVdUatCd
3 4
3 3 10 9 11 16 17 17 17 21
3 3 1 8 11 11 14 15 17 21
36 35 33 35 35 28 32 34 36 36
1
8 12 14 15 15 17 21
sociation of Oral and Maxillofacial Surgeons. One of the authors (LL) selected the observers from the association’s list of members as being oral surgeons with a good professional reputation and with a declared interest in the question of extraction of third molar. They were also selected so as to represent clinics in different geographic regions. There were 3 female and 7 male oral surgeons. Their length of professional experience in oral surgery is presented in Table 1. They had been certified as oral surgeons for 3 to 30 years. Before entering an oral surgery residency, they had had between 2 and 14 years’ experience with general dentistry. The same cases also were judged by another 30 observers who practised as general dental practitioners (GDP). This group of observers has been described in detail in a previous study.3 The coefficient of correlation between years of professional experience and number of molars proposed for extraction was calculated for the oral surgeons and the GDPs, respectively. Results ORAL SURGEONS’ JUDGMENT
FIGURE 1. Woman, 19 years. The molar is partially covered by sot? tissue and in a mesioangular position. Nine of 10 oral surgeons proposed extraction.
The number of mandibular third molars the observers proposed to extract varied from 3 to 21 of 36 teeth (Table 1). There was no third molar that all oral surgeons agreed should be extracted (Table 2). In general, however, the oral surgeons as a group seemed more inclined to propose extraction of third molars partially covered by soft tissue than for other impaction statuses. Thus, 9 of 10 observers were unanimous on the decision to extract one molar, which was partially covered by soft tissue and mesioangularly positioned (Fig 1).
331
KNUTSSON ET AL
Table 2. Number of Oral Surgeons (N = 10) Unanimous on Repeated Observations and With Each Other on Judgment to Extract Asymptomatic Mandibular Third Molars of Different Impaction Status and Position for Patients in Three Different Age Groups Age Group (yr) Impaction Status
Position
19-25
Partially covered by soft tissue
Vertical Mesioangular Distoangular Horizontal
Completely covered by soft tissue
Completely covered by bone tissue
26-w
41-60
8 9 7 7
6 7 5 8
1 3 3 3
Vertical Mesioangular Distoangular Horizontal
5 4 6 2
I 1 0 0
0 0 4 0
Vertical Mesioangular Distoangular Horizontal
8 3 2 2
0 0 4 0
0 0 1 0
FIGURE 2. A, Man, 46 years. The molar is completely covered by soft tissue and in a mesioangular position. All oral surgeons agreed that it should not be extracted. B, Man, 35 years. The molar is completely covered by bone tissue and in a vertical position. All oral surgeons and GDPs agreed that it should not be extracted. C, Man, 60 years. The molar is completely covered by bone tissue and in a vertical position. All oral surgeons and GDPs agreed that it should not be extracted.
Eight of the 10 observers also proposed extraction of one molar in a vertical position completely covered by bone tissue and one partially covered with soft tissue. The observers were unanimous on the decision not to extract 11 molars. All these teeth were completely covered by soft or bony tissue and belonged to patients in the two oldest age groups. Figures 2A, 2B, and 2C show three of these molars. Table 1 presents the intraobserver agreement on whether or not to extract for the 10 oral surgeons. The mean overall intraobserver agreement was 94% There were three oral surgeons ( 1,9, 10) having an individual agreement of 100%. COMPARISON AND
BETWEEN ORAL SURGEONS’
GDPs’ JUDGMENT
The mean number of molars proposed for extraction was 12 for the oral surgeons and 13 for the GDPs. There was no third molar which all oral surgeons or
332
EXTRACTION
OF MANDIBULAR
THIRD MOLARS
Table 3. Percentage of Oral Surgeons (N = 10) and GDPs (N = 30) Proposing Extraction of Asymptomatic Mandibular Third Molars of Different Impaction Status and Position for Patients in the Three Different Age Groups Age Groups (yr) 19-25
26-40
oral SUrgeOLl
41-60 oral
oral GDP
SUrgeOIl
GDP
surgeon
GDP
Position
(a)
(“p)
w)
@)
(W)
(96)
Partially covered by soft tissue
Vertical Mesioangular Distoangular Horizontal
80 90 70 70
37 60 57 83
60 70 50 80
43 77 43 73
10 30 30 30
3 43 30 33
Completely covered by soft tissue
Vertical Mesioangular Distoangular Horizontal
50 40 60 20
10 27 17 30
10 10 0 0
27 I 33 23
0 0 40 0
23 20 53 7
Completely covered by bone tissue
Vertical Mesioangular Distoangular Horizontal
80 30 20 20
50 27 10 30
0 0 40 0
0 10 70 7
0 0 10 0
0 10 20 3
Impaction Status
all GDPs agreed should be extracted (Table 3). However, about three times as many observers, both among the oral surgeons and the GDPs, proposed extraction of molars partially covered by soft tissue compared with molars with other impaction status. The GDPs were unanimous in their judgment not to extract two molars, whereas the oral surgeons were unanimous about 11 molars, including the two molars proposed by the GDPs. The two molars that all observers agreed should not be extracted were completely covered by bone tissue and in a vertical position (Figs 2B, C). Although the frequency of proposal for extraction in the two groups was approximately equal (oral sur-
geons, 10; GDPs, 12), the frequency difference between the groups exceeded 30% for some third molars. The oral surgeons were more inclined than the GDPs to propose extraction for molars completely covered by soft tissue and in a distoangular position, as well as molars in a vertical position in the youngest age group. Figure 3A shows one of these molars. On the other hand, the GDPs more often proposed extraction of molars in a distoangular position and completely covered by soft tissue (Fig 3B) or by bone tissue in the 26to 40-year age group (Table 3). The mean intraobserver agreement of the two groups was comparable, 94% for the oral surgeons, with a range of 78% to 100% (K index, 0.53 and l.O), and 92% for
FIGURE 3. A, Woman, 20 years. The molar is partially covered by SOBtissue and in a vertical position. Eight of 10 oral surgeons and 15 of 30 GDPs proposed extraction. B, Woman, 40 years. The molar is completely covered by soft tissue and in a distoangular position. Ten of the 30 GDPs proposed extraction, and all oral surgeons agreed that it should not be extracted.
333
KNUTSSON ET AL
the GDPs, with a range of 69% to 100% (K index, 0.39 and 1.00). There was no significant correlation between the number of years of professional experience and the judgment regarding whether or not to extract either among the oral surgeons or among the GDPs as a group. There was a tendency for the oral surgeons to propose less extractions with increased experience (Y= - .3 1), whereas this tendency was not found among the GDPs (Y = .2 1). Discussion The validity of the prophylactic indication for third molar extraction can be evaluated from different viewpoints. The GDP or “family dentist” follows a limited group of patients over time, giving them comprehensive dental care, whereas the oral surgeon sees a selected group of patients referred because of their need for a specialist’s opinion or surgical treatment. These different clinical working conditions might at least theoretically influence the assessment of indications for different interventions. One may expect the oral surgeons to be those members of the dental profession with the most thorough knowledge and understanding of indications for removal of third molars. Having noted the wide individual differences among 30 GDPs in the judgment whether or not to extract an asymptomatic third molar,3 we found it warranted to examine the judgments of a group of oral surgeons. Our method of selection of observers resembles the so-called Delphi technique,5 where the observers are selected for their knowledge on the subject and standing within the profession. The degree of variation as to whether to extract or not among the oral surgeons is in agreement with studies concerning judgment analyses among other groups of medical experts. ‘s6 Investigators have related this variation to differences in how medical experts weight clinical information in making diagnostic or management decisions. ’ The oral surgeons’ disagreements could be related to differences in the calculated importance of each cue, so that the cues used and the cue weights differed considerably from oral surgeon to oral surgeon. Most important, the variation of their use of cues involves differences in therapeutic (whether to extract or not) decision in each case, not just differences in their use of cues. Some of the oral surgeons and the GDPs showed substantial intraobserver disagreement. This is in agreement with other investigations concerning medical judgments’ and endodontic treatment decisions.’
The mean number of molars proposed for extraction was similar for the two groups of observers, even though the oral surgeons were expected to be more clinically experienced concerning third molar problems. Wigton, Hollerich, and Partid’ had predicted that variation in diagnostic strategies decreased as clinical experience increased. They found, however, that experienced faculty members were as different from one another in their strategies as medical students. The difference in some cases was so great that the dominant cue for one expert was not used at all by another. Reit et al’ also found that the decision strategy of GDPs was similar to that of endodontists concerning endodontic treatment decisions. One reason for the differences in judgment between the observers is the scant knowledge of the natural history of asymptomatic third molars. There are few longitudinal studies of impacted third molars. However, recent reportsg,io showed that less than 10% to 15% of impacted third molars will develop a pathological change. It is, therefore, difficult to predict that the patient benefit outweighs the surgical risk when a third molar is extracted. There is clearly also a need to study how patients evaluate prophylactic extraction of a mandibular third molar. The most important decision maker will, after all, be the patient. References 1. Wigton RS: Application of judgment analysis and cognitive feedbacks to medicine, in Brehmer B, Joyce CRB (eds): Human Judgment. The SJT view. Amsterdam, Elsevier, 1988, pp 221-245 2. Brehmer A, Brehmer B: What have we learned about human judgment from thirty years of policy capturing?, in Brehmer B, Joyce CRB (eds): Human Judgment. The SJT view. Amsterdam, Elsevier, 1988, pp 75-l 14 3. Knutsson K, Brehmer B, Lysell L, et al: General dental practitioners’ evaluation of the need for extraction of asymptomatic mandibular third molars. Community Dent Oral Epidemiol 1992 (accepted) 4. Winter GB: Principles of Exodontia as Applied to the Impacted Third Molar. St Louis, MO, American Medical Books, 1926 5. Linstrone HA, Turoff HM: Introduction, in Linstrone HA, Turoff HM (eds): Delphi method: Techniques and Applications. Reading, MA, Addison-Wesley, 1975, pp 3-12 6. Reit C, Grijndahl H-G: Management of periapical lesions in endodontically treated teeth. Swed Dent J 8: 1, 1984 7. Reit C, Grondahl H-G, Engstrijm B: Endodontic treatment decisions. A study of the clinical decision-making process. Endod Dent Traumatol 1:102, 1985 8. Wigton RS, Hollerich VL, Partil KD: How physicians use clinical information in diagnosing pulmonary embolism: An application of conjoint analysis. Medical Decision Making 6:2, 1986 9. Stanley HR, Alattar M, Colett WK, et al: Pathological sequelae of “neglected” impacted third molars. J Oral Path01 17:113, 1988 10. Ahlqwist M, Grondahl H-G: Prevalence of impacted teeth and associated pathology in middle-aged and older Swedish women. Community Dent Oral Epidemiol 19:116, 1991