Review of Difficulty Indices for Removal of Impacted Third Molars and a New Classification of Difficulty Indices

Abstract

Background

A number of efforts have been made to establish a reliable assessment model for the surgical removal of impacted third molars. Although many such models have been proposed, none is considered universally applicable, and controversy remains. The earlier attempts were based exclusively on radiographic variables, whereas recent evidence is associated with nonradiographic and demographic variables.

Purpose

This article aimed to prepare review of the relevant literature to summarize the important indices given till date, tabulating only new indices as old indices are mentioned in almost all text books, their merits and demerits and their specificity and sensitivity (if assessed in the literature). In addition to it, authors also aimed to introduce a new descriptive classification of all indices based on pre- and postoperative assessment and further classify based on clinical, radiographic and demographic factors, surgical technique and time. This article will be of academic and practical help for residents and clinicians.

Method

A literature search was conducted in Medline, PubMed, Google Scholar, Science Direct and Cochrane electronic database with keywords—impacted mandibular third molar, difficulty indices, validity of, sensitivity, specificity. No publication year limit was used. The abstracts of the articles were retrieved, reviewed and sorted based on the following inclusion and exclusion criteria. Articles which include a new index, validate difficulty index, compare two or more difficulty indices, specificity and sensitivity of particular index are included. Duplicate articles, articles on complication, flap design, drug study, radiographic relation of inferior alveolar nerve, suturing, classification, only assessing factor affecting difficulty of third molar surgery are not included.

Results

Total 39 (36 articles + 3 books) are included. A total of 20 individual indices were found in the literature; each has been described in brief in this review. These indices are also listed in chronological order with their specificity and sensitivity if accessed in the literature.

Conclusion

There is a definite need to derive and validate an user-friendly index that could be used for preoperative prediction of difficulty. To the best of our knowledge, this review including all indices till date and classification proposed is the first on this subject.

Introduction

Surgical removal of mandibular third molar is one of the most common surgical events [1]. Both the patient and dentist must therefore have scientific evidence-based information concerning the estimated level of surgical difficulty of every case. There are a number of previous studies to evaluate surgical difficulty in the extraction of impacted mandibular third molars [1–6]. Prominent among the proposed models are the Winter’s [7], Pell and Gregory’s [8], Pederson’s [9] and the WHARFE [10] classification/scoring systems. These earlier attempts were based exclusively on radiographic variables, whereas recent evidence has associated a wide variety of non-radiographic variables [5, 11–13]. This is a review of relevant literature to summarize the important indices given till date, tabulating only new indices as old indices are mentioned in almost all text books, their merits and demerits and their specificity and sensitivity (if assessed in the literature). In addition to it, authors introduce their own descriptive classification (Fig. 1) of all indices based on pre- and postoperative assessment and further classify based on clinical, radiographic and demographic factors, surgical technique and time. To the best of our knowledge, this kind of review and the classification given by authors is the first on this subject.

Fig. 1.

Descriptive classification of difficulty indices of impacted third molar

A literature search was conducted in Medline, PubMed, Google Scholar, Science Direct and Cochrane electronic database with keywords-impacted mandibular third molar, difficulty indices and validity of sensitivity and specificity (Fig. 2). No publication year limit was used, so that the search could include the first available year of the particular database to April 2020. Additionally, a manual search of books was performed. The abstracts of the articles were retrieved, reviewed and sorted based on the following inclusion and exclusion criteria. Inclusion criteria are as follows: articles which include a new index, validate difficulty index, compare two or more difficulty indices and check for specificity and sensitivity of particular index. Exclusion criteria are as follows: duplicate articles, articles on complication, flap design, drug study, radiographic relation of inferior alveolar nerve, suturing, classification and only assessing factor affecting difficulty of third molar surgery. A total of 39 (36 articles + 3 books) were included in this review. A total of 20 independent indices were found which scale difficulty of third molar removal. These indices are listed in chronological order along with specificity and sensitivity of indices if accessed in the literature (Table 1).

Fig. 2.

Systematic review process. Flowchart of article selection through the systematic review process

Table 1.

Chronological list of indices proposed and their specificity and sensitivity

S. no.	Index	Year of publication	Specificity	Sensitivity
1	Winter's lines (WAR)	1926	Not accessed
2	The Pell–Gregory classification	1933	Class c 88% Class 3 62%	Class c 15% Class 3 50%
3	Parant scale	1974	Not accessed
4	WHARFE index	1976	Not accessed
5	Pederson, modified Pederson	1988, 1994	45%, 89%	94.9%,20%
6	Koerner index	1994	Not accessed
7	Modified Parant scale by Garcia et al.	2000	Not accessed
8	Penarrocha et al. index	2000	Not accessed
9	Yuasa index	2002	92%	85%
10	Gbotolorun index	2007	97%	80%
11	Carvalho et al. scale	2011	Not accessed
12	Juodzbalys G (MRACBS) scale	2013	Not accessed
13	Kharma scale	2014	68.4%	18.2%
14	Mario Vicente-Barrero (SPESD)	2014	Not accessed
15	Roy et al. index	2015	Not accessed
16	Conti scale	2015	Not accessed
17	Pernambuco index	2017	87.9%	93.1%
18	Ribes Lainez N et al. index	2017	Not accessed
19	Sammartino G et al. index	2017	Not accessed
20	Zang et al. index	2019	Not accessed
21	Kim JY et al. index	2019	Not accessed

Winter's Lines (WAR)

Three imaginary lines to determine the depth of the mandibular third molars in bone have been described earlier in 1926 [7]. This method is taught to most undergraduate students, but is reported to be used little in practice [1]. IsidoroCortell-Ballester et al. [6] have used computer-assisted system in the classification of lower third molars which allowed objective classification of the third molars within the Winter subclasses, as follows [6]^:

• Third molars with a negative angle (< 0º) were considered to be inverted.

• Third molars with an angle between 0º and 30º were considered to be horizontal.

• Third molars with an angle between 31º and 60º were considered to be mesioangular.

• Third molars with an angle between 61 º and 90º were considered to be vertical.

• Third molars with an angle > 90º were considered distoangular.

WHARFE Index

Winter’s classification was expanded by Macgregor [10], to WHARFE which includes the Winter’s lines along with other factors and has recently been used in several studies [14, 15]. Using orthopantomograms, MacGregor 1976 was the first to develop a model to predict operative difficulty, using length of operation as a proxy for surgical difficulty. Although MacGregor was unable to develop a multivariate model of predictive value from the comprehensive set of variables, he was able to construct a semiquantitative model based on relatively few radiographic factors that were not interrelated. The resulting model showed that an increase in radiographic scoring of difficulty according to the WHARFE classification system was associated with increased operating time [11]. Edwards et al. [15] found a poor correlation between the WHARFE index and the surgeons anticipated difficulty. Chandler et al. [16] preferred to assess difficulty during the procedure and felt that experienced surgeons overestimated the surgical difficulty based on radiographic assessment. Unfortunately, WHARFE index is related to radiological features alone; the details of the surgical procedures are not considered, and in view of their complexity, they are rarely used in routine practice [17].

Pell–Gregory Classification

The Pell–Gregory classification [8] is widely cited in standard texts on oral and maxillofacial surgery [18] as a useful set of criteria for predicting difficulty in the extraction of an impacted lower third molar. The classification of such molars is based on their spatial relationships (as shown by radiography) to the ascending ramus of the mandible and to the occlusal plane. Garcia et al. studied using Pell–Gregory index class C to indicate “difficult,” specificity was 88%, but sensitivity was low at 15%. Using Pell–Gregory class 3 to indicate “difficult,” sensitivity was somewhat better (50%), but at the expense of specificity (62%) [4]. But this classification is not a reliable predictor of surgical difficulty in the extraction of vertical impacted lower third molars (which are difficult to classify on the Pell–Gregory scales), suggesting that this classification is of little value in clinical practice [4, 5, 17]. According to Kyeong-Lok Park et al., Pell–Gregory and Winter’s classification, there was no difference between the probability of moderately difficult cases and that of very difficult cases in each class [19].

Pederson Scale

Pederson proposed a modification of the Pell–Gregory scale [9]. The total scores by which difficulty is judged are based mainly on local anatomy and radiographs. Akadiri studied the sensitivity and specificity of Pederson index of 94.9% and 45%, respectively [20]. The Pederson index though reproducible is not a reliable instrument for predicting surgical difficulty of third molar surgery when compared with actual surgical difficulty as determined by the operation time [16, 20–23]. This scale was not proportional to difficulty in moderately difficult cases, but it was directly proportional to very difficult cases. When the Pederson index score is low, other causes, except position-related causes, might not impact extraction difficulty. It does not take various relevant factors into account, such as bone density, flexibility of the cheek, buccal opening and inverted tooth [1, 5]. Some authors seeking to develop an index of difficulty based on the Pederson index have introduced new variables, including root morphology, as reported by Kharma et al. [23], and subjective clinical variables, as reported by Roy et al. [24]. These authors claim to have attained more reliable results than those found using the Pederson index.

Koerner Index (Table 2)

Table 2.

Koerner index

Classification difficulty index	Value
Angulation
Mesioangular	1
Horizontal	2
Transverse	2
Vertical	3
Distoangular	4
Depth
Level A	1
Level B	2
Level C	3
Ramus relationship
Class I	1
Class II	2
Class III	3
Difficulty index
Slightly difficult	3–4
Moderately difficult	5–7
Very difficult	8–10

Koerner KR. The removal of impacted third molars—Principles and procedures. Dent Clin North Am. 1994;38:255–78

Koerner et al. proposed a difficulty index scale for removal of these teeth on the basis of local anatomy and radiographs, which were helpful in predicting the difficulty that would be encountered intra-operatively, as well as the postoperative complications. It covers indications for surgery and how indications and surgical difficulty correspond with the age of the patient. The difficulty of the procedure depends not only on age, but on how the tooth is categorized radiographically (classification, position, angulation) and on the patient's facial characteristics (tapering versus compact facial form). The operator's familiarity with the anatomy of the surgical site is indispensable for uneventful third molar surgery [25]. Khanal et al. [26] stated that this index can be used as a tool for predicting the frequency of the postoperative complications, especially swelling and trismus.

Yuasa Scale

Yuasa et al. proposed a new scale that takes into account not only relative depth and relation with the ramus of the mandible, but also width of the root [5]. Depth which is deep occlusal level (level c), ramus relationship/space available which is no space (class 3) and width (bulbous) are associated with great difficulty. It was better than Pederson’s index in terms of odds ratio, sensitivity and specificity. It had recorded 8.3% false + ve and 15% false −ve sensitivity of 85% and specificity of 92% compared with 50% and 92% on Pederson’s index. The main difference between Yuasa index and the Pederson index is the inclusion of width of the root. This suggests that the width is a more important factor than weight of score. Yuasa new index is simple and valid, particularly in checking difficult cases. Advantages of this index are that even busy practitioners can use this new index and this does not require any calculation and training. One limitation is the existence of an unpredictable factor such as abnormal root curvature that cannot be detected on panoramic radiographs [5].

Modified Parant Scale

Parant scale was implemented to predict postoperative difficulties [27]. Garcia et al. modified Parant scale and assessed the reliability of the Pell–Gregory classification for predicting surgical difficulty against their own postoperative difficulty scale [4, 28]. According to author, scale is a reliable, consistent measure and can be considered an appropriate “gold standard.” Modified Parant scale is considered to be relatively more reliable than Pederson scale [22]. In this scale, Grade I is extraction with forceps only; Grade II is extraction by ostectomy; Grade III is extraction by ostectomy and coronal section; and Grade IV is complex extractions. However, this index cannot be used for preoperative evaluation [5]. Pederson and modified Parant scales fail to predict difficulty level of tooth removal accurately, especially in cases of distoangular impactions [22].

Gbotolorun Index (Table 3)

Table 3.

Gbotolorun et al. index

Variables	Value	Range
Age (years)	1	< 24
	2	25–34
	3	> 34
BMI	1	< 24
	2	25–30
	3	> 30
Depth from point of elevation	1	0–3 mm
	2	4–6 mm
	3	> 6 mm
Curvature of roots	1	Incomplete
	2	Straight/favorably curved
	3	Unfavorably curved
Total	12

New index score: easy: 4–6, moderately difficult: 7–9, very difficult: 10–12

Gbotolorun et al.: Assessment of factors associated with surgical difficulty in impacted mandibular third molar extraction. J Oral Maxillofac Surg 2007;65:1977–83

Gbotolorun et al. proposed a preoperative index which uses four variables (both clinical and radiologic) [12]. The major difference of this index and the Pederson index is the incorporation of the clinical variables (namely age and BMI). It was also associated with the curvature of roots of the impacted tooth and the depth from point of elevation. Depth from the point of elevation has been described as the single most important indicator for prediction of the difficulty [2, 10]. It is comparable to those of Akinwande who divided his score into similar groups [2]. However, the present study contrasts those of Renton et al. [1] and Yuasa et al. [5] although these studies used different modes of classification of intraoperative difficulty. The sensitivity and specificity of the new index to determine actual intraoperative difficulty as follows: In determining easy extraction, the new index had 74% and 79% of sensitivity and specificity, respectively (accuracy 76%), while those of the Pederson index were 43% and 74%, respectively (accuracy 49%). In determining moderately difficult extractions, the sensitivity and specificity of the new index were 70% and 75%, respectively (accuracy 73%), while those of the Pederson index were 52% and 48%, respectively (accuracy 49%). For difficult cases, the new index was 80% of sensitivity and 97% of specificity (accuracy 98%), while the Pederson was 20% of sensitivity and 89% of specificity (accuracy 86%) [12]. However, this index has not been validated [29].

Carvalho et al. Scale

Carvalho et al. have conducted a study to adjust a multivariate model to explain each of the response variables for the occurrence of surgical difficulty [30]. The odds ratios revealed that the likelihood of difficulty is greater (1) if classified by Winter as horizontal in comparison with those classified as mesioangular by Pell and Gregory or classified as C3 in comparison with those classified as A1; (2) if it has two roots or a germ in comparison with those with a fused root; (3) if the root is bent; (4) if the periodontal space is completely radiopaque in comparison with those with a mixed or completely radiolucent image; and (5) if there is a close relation with the crown and root of the second molar in comparison with those only in contact with the root of the second molar. Scale proposed is based on surgical technique employed for extraction: 1: use of elevator alone (low difficulty), 2: ostectomy (moderate difficulty), 3: ostectomy and tooth sectioning (high difficulty) and time elapsed between incision and suturing of tissues: 1: < 15 min low difficult, 2: 15–30 min moderate difficult, 3: > 30 min high difficult [30].

Juodzbalys and Daugela ((M, R, A, C, B and S) scale (Table 4)

Table 4.

Juodzbalys and Daugela (MRACB) index

Position of the mandibular third molar	Risk degree of presumptive intervention (score)
	Conventional (0)	Simple (1)	Moderate (2)	Complicated (3)
Mesiodistal position in relation to the second molar—M and the mandibular ramus—R
Relation to the second molar—M Relation to the mandibular ramus—R	Crown directed at or above the equator of the second molar Sufficient space in dental arch	Crown directed below the equator to the coronal third of the second molar root Partially impacted in the ramus	Crown/roots directed to the middle third of the second molar root Completely impactedin the ramus	Crown/roots directed to the apical third of the second molar root Completely impacted in the ramus in distoangular or horizontal position
Apicocoronal position in relation to the alveolar crest—A and the mandibular canal—C (IAN injury risk)
Relation to the adjacent alveolar crest (from the uppermost point of the tooth)—A Relation to the mandibular canal (from the lowest point of the tooth)	Tooth is completely erupted  >  = 3 mm to the mandibular canal	Partially impacted, but widest part of the crown (equator) is above the bone Contacting or penetrating the mandibular canal, wall of the mandibular canal may be identified	Partially impacted, but widest part of the crown (equator) is below the bone Contacting or penetrating the mandibular canal, wall of the mandibular canal is unidentified	Completely encased in the bone Roots surrounding the mandibular canal
Buccolingual position in relation to mandibular lingual and buccal walls—B (LN injury risk)
Relation to mandibular lingual and buccal walls—B	Closer to buccal wall	In the middle between lingual and buccal walls	Closer to lingual wall	Closer to lingual wall, when the tooth is partially impacted or completely encased in the bone (A2 or A3)
Spatial position—S
Spatial position—S	Vertical (90°)	Mesioangular ≤ 60°	Distoangular ≥ 120°	Horizontal (0°) or inverted (270°)

Juodzbalys G, Daugela P. Mandibular Third Molar Impaction: Review of Literature and a Proposal of a Classification. J Oral Maxillofac Res 2013;4(2):e1-12

This classification describes wisdom tooth relation to the adjacent anatomical structures: mandibular ramus, second molar, alveolar crest, mandibular canal and the spatial position of the tooth. Wisdom tooth position assessment should be performed clinically and using cone beam computed tomography and panoramic radiographic images [31]. Risk degree of presumptive intervention is scored as follows: • conventional extraction is determined, when all parameters are equal to score 0; • simple, when at least one parameter is equal to score 1 and surgical extraction with coronectomy and/or sectioning of roots is determined; • moderate, when at least one parameter is equal to score 2 and surgical extraction with coronectomy and/or sectioning of roots is determined; • complicated, when at least one parameter is equal to score 3 and surgical extraction with coronectomy and/or sectioning of roots is determined. Extraoral approach can be indicated. To make the classification more informative, each component of the indices (M, R, A, C, B and S) is described independently. This index is promising to be a helpful for assessment as well as for planning for surgical operation [31]. But they did not verify its reliability and validate the index. It is of limited clinical application due to the need to the cone beam computed tomography in classification [29].

Kharma Scale (Table 5)

Table 5.

Kharma scale

Relationship (position of mandibular third molars)
Mesioangular Horizontal/transverse Vertical Distoangular	0 1 2 3
Depth
Level A: high occlusal level Level B: medium occlusal level Level C: deep occlusal level	1 2 3
Ramus relationship/space available
Class 1: sufficient space Class 2: reduced space Class 3: no space	0 1 2
Root form
Convergent Divergent Bulbous	0 1 2
Difficulty index
Very difficult Moderately difficult Slightly difficult Easy	7–10 5–7 3–4 1–2

Kharma MY et al. Reliability of Pederson scale in surgical extraction of impacted lower third molars: Proposal of new scale. J Oral Diseases. 2014;2014:1–4

Kharma scale proposed a new index based on four factors: tooth angulation, the depth of the third molar in the mandible, the relationship with the ramus/space available and root form and postoperative difficulty indicated by Parant scale which has shown a significant correlation. It is similar to Pederson index; in that, it measures the same parameters in addition to root forms, and it is close to Yuasa scale as the former assess the same parameters in addition to tooth angulation. Kharma scale showed more accuracy and reliability in preoperative estimation of the difficulty than Pederson scale [23]. According to Al-Samman et al., Kharma scale was unreliable as there was no significant association between the Kharma score and duration of operation and both radiological and clinical information must be taken into account. In addition, curvature of the root is sometimes an unpredictable factor, as it is often not visible in panoramic radiographs. Kharma scale reported 85.7% false + ve and 25.7% false −ve and showed a very low sensitivity of 18.2% and a specificity of 68.4% [32].

Mario Vicente-Barrero et al. (SPESD) Scale (Table 6)

Table 6.

Mario Vicente-Barrero et al. index

Direction of eruption
Vertical	0
Mesioangular	1
Horizontal	2
Depth
Mild impaction	0
Moderate impaction	1
Severe impaction	2
Distance between the second molar and the anterior border of the ramus
Class I	0
Class II	1
Class III	2
Winter’s distance
< 5 mm	0
6–10 mm	1
> 10 mm	2
Proximity and situation of the third molar relative to the dental nerve canal
Separated	0
In contact	1
Overlapping	2
Root morphology
Germ, single root or several roots fused together	0
2 or more, parallel or convergent roots	1
2 or more, divergent or abnormal roots	2

Mario Vicente-Barrero et al. Is it Possible to Predict the Difficulty of Third-Molar Surgical Extraction? IJSR International journal of scientific research September 2014,3( 9):314–316.

In this study, six preoperative parameters were considered, every variable was assigned a score of 0, 1 or 2, and a scale for preoperative evaluation of surgical difficulty (SPESD) was created, which ranked from 1 to 12. This scale was combination of previously given scales. They classified it in a low-difficulty one corresponding to ABR (anterior border ramus) and WD (winter’s depth) equal to 0; a medium-difficulty one corresponding to ABR and WD equal to 1; and a high-difficulty cluster corresponding to ABR and WD equal to 2 [33].

Roy et al. Index (Table 7)

Table 7.

Roy et al. index

(A) Pederson’s index
1. Angulation of the tooth
Mesioangular	1
Horizontal	2
Vertical	3
Distoangular	4
2. Depth
Level A	1
Level B	2
Level C	3
3. Ramus relationship
Class I	1
Class II	2
Class III	3
(B) Depth from point of elevation
0–3 mm (slightly difficult)	1
4–6 mm (moderately difficult)	2
> 6 mm (very difficult)	3
(C) Preoperative clinical assessment chart
Mouth opening (adequate/reduced)	(1/2)
Tongue size (normal/large)	(1/2)
Angulation of external oblique ridge (obtuse/acute))	(1/2)
Cheek flexibility (flexible/non-flexible)	(1/2)
(D) Width of root
Thin (A ≥ B)—easy	1
Bulbous (B > A)—moderate	2
Thick (multiple roots B > A, B > thickness of all roots combined)	3
(E) Curvature of roots
Straight roots	1
Both roots distally curved	2
Distal root distally curved	3
Both roots curved toward each other	4
Mesial root distally curved	5
Distal root mesially curved	6
Mesial root mesially curved	7
Both roots mesially curved	8
Both roots curved away from each other	99

Total score classified into “slightly difficult,” “moderately difficult”

and “difficult.” Minimum score—10 and maximum—33. Roy et al., J. Maxillofac. Oral Surg. (July–Sept 2015) 14(3):745–749

Roy et al. proposed a new index in 2015 and used 29 parameters as reference standards [24], in which factors such as depth from point of elevation, mouth opening, tongue size, angulation of external oblique ridge, cheek flexibility, width of the root and curvature of roots were incorporated along with the Pederson’s index. There is a significant association between this new index and the operation time which reflects on the accuracy of the index [24]. In this index, the authors focused only on the tooth, not taking into account demographic characteristics, and used only time as the outcome variable, without validation [29].

Conti Scale

Conti et al. performed a study involving a complex numerical method with the aim of providing an efficient form of preoperative analysis, enrolled 1000 subjects. Comparison of preoperative difficulty assessment scores with the actual difficulty encountered by an operator confirms the reliability of this protocol [34]. However, this method disregards the relevance of demographic factors and has not been clinically validated [29].

Ribes Lainez N et al. Index

A presurgical radiographic scale was developed, based on ten parameters. Each parameter was scored from 1 to 3. The maximum final score was 27 points, with a minimum final score of 13. Ten parameters were as follows: inclination of the third molar, inclination of the second molar, pericoronal radiolucency, root radiolucency, root shape, Winter’s distance, distance between the ramus and second molar, width of the third molar, coronal area and root length. Extraction was not difficult (10–16) points, average difficulty (17–23 points) and difficult (24–30 points). According to author, this scale is effective, since the highest scores were significantly correlated with longer ostectomy time and total surgical time [35].

Penarrocha et al.

Penarrocha et al. in turn summed the scores corresponding to pericoronal radiolucency, pericoronal space, Winter’s distance and coronal area and subdivided the size and shape of the roots into two separate parameters: the length of the root and the type of root. Each variable was scored from 0 to 2, and the difficulty score was 0–5 (not difficult), 6–10 (average difficult) and over 10 (great difficult). Author claims that this is one of the scales involving the largest number of parameters, and higher scores have been shown to correspond to longer ostectomy times and total surgical times, thereby confirming the efficacy of the classification [36]. In any case, they are mainly based on preoperative studies of panoramic x-ray images, which are unreliable for classification of impacted molars [16].

Pernambuco Index (Table 8)

Table 8.

Pernambuco index

Variable	Classification	Value
Level of the occlusal plane (Pell and Gregory)	A B C	1 2 3
Available retromolar space (Pell and Gregory)	1 2 3	1 2 3
Impaction angle (Winter)	Vertical Mesioangular Horizontal Distoangular	1 2 3 4
Root curvature	Non-dilacerated Dilacerated	1 2
Number of roots	One fused root  ≥ roots Tooth germ	1 2 3
Relationship to the second molar	No contact Contact with crown alone Contact with root	1 2 3
Age (years)	< 25  ≥ 25	1 2
BMI (kg/m2)	18.5–24.9 (ideal weight range)  ≥ 25 (overweight)	1 2
Surgical difficulty
Low Moderate High	8–12 13–17 18–22

BMI—body mass index

de Carvalho RWF et al. Pernambuco index: predictability of the complexity of surgery for impacted lower third molars. International Journal of Oral and Maxillofacial Surgery. 2017; 47(2):234–240

The index presented herein differs from existing indexes in its conception, application and validation. The proposed index was developed on a statistical basis using two outcome variables for difficulty, these being the duration of surgery and the surgical technique, and showed significant evidence for clinical, demographic and radiographic factors. The size of the total sample studied was determined by sampling, not empirically. The index was later verified, showing it to have high sensitivity (93.1%), specificity (87.9%) and accuracy (90.4%). With the use of reference statistics in the development and quality assurance processes, this statistically structured instrument has proven to be an efficient and reliable tool and is better indicators of difficulty than the Pederson index [29].

Sammartino et al. Index (Table 9)

Table 9.

Sammartino G et al. index

Variables	Score
Angulation
Mesiovestibular inclination	1
Mesiolingual inclination	2
Horizontal	2
Transverse with crown vestibular version	2
Vertical	3
Distovestibular inclination	3
Distolingual inclination	5
Transverse with crown lingual version	5
Available space:
Class I	1
Class II	2
Class III	3
Depth
Position A	1
Position B	2
Position C	3
Relation with mandibular canal
Sufficient space (> 3 mm)	1
Reduced space (1–3 mm)	2
Space absent or direct contact	4
Bone density
D1	2
D2	1.5
D3	1
D4	0.5
Buccolingual position
Vestibular	1
Middle of crest	2
Lingual	3
Dental morphology
No abnormalities	1
Low degree of abnormalities	2
The degrees of surgical difficulty
Low (L)	6.5–12.5	Plus	L plus
Medium (M)	13–17.5	Plus	M plus
High (H)	18–22	Plus	H plus

Sammartino G et al. Extraction of mandibular third molars: proposal of a new scale of difficulty. Br J Oral Maxillofac Surg. 2017;55(9):952–957

Sammartino et al. propose and validate a new scale of surgical difficulty based on variables not previously considered (relating to morphological abnormalities, the type of undercut and the transverse position of the tooth); result of the kappa test was 0.73, which indicated a substantial concordance between the preoperative and postoperative assessments of difficulty indicated by the new scale. The linear weight of kappa was 0.8 and the quadratic weight 0.87. In light of the final score, we can distinguish three groups of difficulty: low, medium and high. A high degree of abnormality includes any factor that has not been included in the score on the grounds that, in itself, it is a major difficulty even if all the other variables are not difficult, in which case author added the term “Plus” to the evaluation score. Such Plus groups are considered to be the most difficult [37].

Zhang et al. Index (Table 10)

Table 10.

Zang scale

Degree of bone impaction
None	0
Partial	2
Full	3
Shape of roots
Normal	0
Swollen root	1
Crooked root	2
Impaction angle
< 30	0
> 30	1
Relation to IAC
None	0
Touching	0.5
Crossing	1
Roots, n
1	0
2	1
Age (year)
< 25	0
25–35	1
> 35	2
Difficulty score
Low	0–5.4
Moderate	5.5–7.4
High	7.5–10

IAC inferior alveolar canal

Zhang et al. New Index for Impacted Third Molar Removal. J Oral Maxillofac Surg 2019

In this index, a mathematical model and regression analysis were performed to explore six main factors (age, number of roots, degree of bone impaction, shape of roots, impaction angle and its relation). Comparison of the Pederson index and new index with operating time showed k agreements of 65.30 and 77.9%, respectively (P < 0.01), suggesting that the prediction results of the new index are more objective and accurate. For statistical analysis, the total score was classified as “low,” “moderate” or “high” based on quartiles of the scores (i.e., quartile ≤ 1, low; quartiles 1 to 3, moderate; quartile ≥ 3, high). The minimum score was 15 and the maximum score was 30. Though variations of the roots, maximal mouth opening and cheek dispensability were not considered, this study used a panoramic radiograph, which has limitations of two-dimensional imaging and the overlap of structures [38]. Compared with the study conducted by Roy et al. [24], the proposed index is more effective, concise and straightforward [38].

Kim et al. Modified Index

For difficulty classification, the authors propose one more difficult category beyond the existing three-step difficulty index. In this study, author added reverse and level D, which are not included in the existing three-step difficulty index [25]. Thus, one more advanced category (category IV: extremely difficult) was added in the difficulty index. When the entire crown of the impacted third molar was located below the CEJ of the adjacent second molar, it was defined as either level C or level D. Level C was defined as a condition when more than half of the impacted third molar crown was located superior to the mid-level of the adjacent second molar root. Third molar crown levels inferior to the aforementioned reference were categorized as level D. According to the difficulty, the spatial relationship was scored on a scale of 1–5 as follows: Mes (1), Hor (2), Ver (3), Dis (4) and Rev (5). For depth, 1–4 points were given to levels A to D, respectively. Similarly, 1–3 points were given to classes I to III, respectively. Finally, the difficulty index was defined as slightly difficult (3–4 points), moderately difficult (5–7 points), very difficult (8–10 points) and extremely difficult (11–12 points). The strength of agreement of the total points score and difficulty index was 0.855 and 0.746, respectively [39].

Conclusion

Several classification systems have been established to estimate the surgical difficulty of removing mandibular third molars, but they have proven to be of little clinical use. To the best of our knowledge, this review including all indices till date and classification proposed is the first on this subject. There is a definite need to derive and validate a user-friendly index that could be used for preoperative prediction of difficulty; however, it is much more important that any index so employed should provide accurate information as the consequence of wrong judgments could be quite deleterious to the patients and embarrassing to the surgeon.

Funding

None.

Compliance with Ethical Standards

Conflict of interest

None.