Third molar maturity index and legal age in different ethnic populations: Accuracy of Cameriere’s method

Abstract

Molar development is widely accepted as a reliable indicator of chronological age in a forensic context. A quantitative method for age estimation has been proposed by Cameriere et al. based on the relationship between chronological age and the third molar maturity index (I_3M), which is the ratio between the two apical pulp widths and the total tooth length. Cameriere’s cut off value of 0.08 was found to be a reliable tool in assessing the threshold of 18 years of age. The aim of this study was to evaluate the accuracy of the I_3M in different ethnic populations focusing on its specificity (subjects correctly classified as <18 years based on I_3M >0.08) and sensitivity (subjects correctly classified as ≥18 years based on I_3M <0.08). A review of 22 scientific articles was performed, representing data from panoramic radiographs of 12,286 individuals (5723 males and 6563 females) from all over the world, including some ethnical subgroups. The I_3M showed an overall sensitivity for both sexes ranging from 51.8% to 93.4% and a specificity ranging from 86.2% to 100%. The proportion of correctly classified individuals ranged from 74% to 95%. These results can be useful to refer the age estimation with the margin of error of subjects correctly classified as minors (specificity) or adults (sensitivity), according to sex, ethnicity and geographical distribution. The I_3M can be considered a suitable method for estimating adulthood in forensic settings, regardless of sex. However, ethnic group can affect the accuracy.

Keywords

Age estimation third molar dental age Cameriere’s method legal age forensic odontology

Introduction

Forensic age estimation in the living is crucial in criminal and civil proceedings, such as for refugee and asylum seeker procedures, human trafficking, child pornography, old-age pension and competitive sports.¹ The determination of adult or juvenile status is a scientific and legal challenge that forensic professionals need to solve using reliable methods.² According to the recommendations issued by the Study Group on Forensic Age Diagnostic (AGFAD) and Forensic Anthropology Society of Europe (FASE), chronological age estimation of a living person should consist of:

a physical inspection for recording anthropometric data, signs of sexual maturation and any age-relevant developmental disorders;

X-rays of the left hand for the assessment of bone maturity or conventional radiography/computed tomography of the medial epiphysis of the clavicle if the skeletal development of the wrist is completed; and

a dental examination and an orthopantomogram for the assessment of dental maturity.^1,3–5

Regarding dental age estimation, methods commonly applied in adolescence are based on tooth development compared to a set of developmental stages charts,^6,7 while in late adolescence such an estimation is based on the crown and root features of the third molar, such as Demirjian’s system.⁸

The mineralisation of permanent teeth is very helpful until the age of 15–16 years, when second molars have also finished their development.^6,7,9 However, in many countries, 18 years is the age of majority at which the individual is generally defined as an adult.¹⁰ With specific reference to this threshold of adulthood, third molars are the last teeth to complete their development, and therefore they represent the best dental indicators of legal age.^11,12 In fact, the development of third molars takes place between 14 and 23 years of age, during a specific age span from late adolescence to early adulthood, when their mineralisation is completed in most healthy individuals.^6,7,9,13,14

Mincer et al.¹⁵ first recommended the use of the molar development as an estimator of chronological age in forensic context using Demirjian’s system.^8,15 More recently, Cameriere et al.¹⁶ developed a method for estimating age among Caucasian people and discriminating adults and minors based on the relationship between age and the third molar maturity index (I_3M). The I_3M is defined as the ratio between the two apical pulp widths and the total tooth length.¹⁷ According to this approach, a cut-off I_3M value of <0.08 was determined in order to assign an individual as an adult age ≥18 years with a 95% confidence interval. The left inferior third molar is recommended for the measurements and assessment of I_3M due to its longer duration of development.¹⁶ In this regard, it is well known that the root growth in the third upper molar finishes earlier than the lower one.^6,7,9,16 Cameriere’s I_3M showed a better specificity (the proportion of subjects correctly classified as <18 years based on I3M >0.08) compared to Demirjian’s stage G in adolescence and a better sensitivity (the proportion of subjects correctly classified as ≥18 years based on I_3M <0.08) compared to Demirjian’s stage H for adult age.¹⁶ Furthermore, the use of the I_3M method can be particularly suitable when used in conjunction with the other skeletal maturity markers.¹⁸

However, the application of Cameriere’s I_3M is strongly dependent upon the presence of the third molars, which may be congenitally absent or surgically removed.^17,19 Unfortunately, if the root development of the third molar is complete, there is no possibility of measuring the open apices, and therefore the I3M is always <0.08 (adult). Therefore, the possibilities of using this method for assessing the age of majority are limited by the variable duration of the third molar’s development. There are also a wide range of factors associated with the timing of maturation, including genetic, biological and environmental influences, which can affect the time of appearance and progression of maturity indicators.²⁰ Based on the variability in geographical distribution of dental maturity and physical development according to ethnicity, nutrition and socio-economic status, the aim of this study was to evaluate the accuracy of Cameriere’s method as applied to different populations, and the sensitivity and specificity of the I_3M cut-off value (0.08) in discriminating adults and minors.

Methods

A critical literature review was conducted independently by two examiners. Studies evaluating the accuracy of Cameriere’s I_3M method¹⁷ were derived from the following electronic databases: PubMed, Scopus, ISI Web of Science and System for information on Grey Literature in Europe (SIGLE).

Exclusion criteria were studies not having primary data (editorials, literature reviews, commentaries, case reports, letters), studies where I_3M was compared to other age estimation methods, studies not reporting data on sensitivity and specificity of the I_3M cut-off value and studies not in English. No filters were applied for the date of publication.

Two examiners independently provided an initial selection of articles to determine whether they might potentially fit the inclusion criteria. Disagreements were solved between the two examiners; if no agreement could be reached, it was resolved using a third co-author as a reviewer.

A data extraction sheet was developed and for each selected study, and the following items were recorded: first author’s name, year of publication, study’s country, sample size (overall, male and female), age range, sensitivity, specificity and accuracy (overall, male and female when available).

Results

The search of PubMed, Scopus, ISI Web of Science and SIGLE databases provided a total of 973 articles. Of these, 951 studies were discarded because they did not meet the inclusion criteria following review (studies in English with primary data on sensitivity and specificity of the I_3M cut-off value).

A total of 22 research studies were identified and were included in the critical literature review.^{2,12,16,19,21–38} In Table 1, the main findings of the 22 studies are summarised, including country, sample size (male and female), age range, sensitivity, specificity and data accuracy according to sex and 95% confidence interval.

Table 1.

Data from selected articles showing sensitivity, specificity and accuracy.

Author (year)		Country		Sample size (male/female)		Age range	Sensitivity, male (%)		Specificity, male (%)		Sensitivity, female (%)	Specificity, female (%)			Sensitivity, male and female (%)	Specificity, male and female (%)			Accuracy (%)	95 confidence interval (%)
AlQahtani et al. (2017)²¹		Saudi Arabia		272		14–22	52.3		100		51.3	97			51.8	98.5			m: 75.6
AlQahtani et al. (2017)²¹				(127/145)															f: 72.4
Angelakopoulos et al. (2017)²²		South Africa		833		14 –24	–		–		–	–			80	95			90	87–91
Angelakopoulos et al. (2017)²²				(333/500)
Antunovic et al. (2018)²³		Montenegro		683		13–24	92		94		82	96			87	95			m: 93	m: 90–96
Antunovic et al. (2018)²³				(324/359)															f: 89	f: 85–91
Balla et al. (2017)²		India		150		7 –15	90.2		95.1		83.3	98.3			86.7	96.7			m: 93.1	m: 88.2–98
				(79/71)															f: 91.2	f: 86–96.4
Boyacioglu et al. (2017)¹⁹		Holland		360		14 –22	84		95		72.7	96.3			78.3	95.6			m: 88.9	m: 83.3–91.8
Boyacioglu et al. (2017)¹⁹				(180/180)															f: 83.3	f: 77.7–85.8
Cameriere et al. (2014)²⁴		Albania		286		15 –22	94.1		90.9		75.4	96.6			84.7	93.7			m: 92.5	m: 85.9–96.2
Cameriere et al. (2014)²⁴				(134/152)															f: 87.5	f: 81.2–90.4
Cameriere et al. (2018)²⁵		Chile		822		11 –22	–		–		–	–			70.5	88.4			m & f: 83
Cameriere et al. (2018)²⁵				(350/472)
Cavric et al. (2016)²⁶		Botswana		1294		13 –23	88		94		88	96			88	95			m: 91	m: 88–93
Cavric et al. (2016)²⁶				(582/712)															f: 92	f: 90–93
Chu et al. (2018)²⁷		China		840		12 –25	87.1		95.6		77.2	98.4			82.1	97			m: 90.7	m: 87.9–93.5
Chu et al. (2018)²⁷				(420/420)															f: 87.6	f: 84.5–90
Dardouri et al. (2016)²⁸		Libya		307		14 –22	90.9		100		90.6	100			90.7	100			m: 95.1	m: 91.5–98.7
Dardouri et al. (2016)²⁸				(144/163)															f: 94.5	f: 90.9–98.1
De Luca et al. (2013)¹²	Italy		397		13 –22		–	–		–			–	86.6			95.7	91.4
De Luca et al. (2013)¹²			(205/192)
De Luca et al. (2016)²⁹	Colombia		288		13 –22		91.7	90.6		95.1			93.8	93.4			92.2	m: 91.2
De Luca et al. (2016)²⁹			(125/163)															f: 94.4
Deitos et al. (2015)³⁰	Brazil		444		11–22		87.6	86.9		86.5			67.2	77.4			86.2	87.4
Deitos et al. (2015)³⁰			(205/239)
Franklin et al. (2016)³¹	Australia		143		14 –22		90	85		90			88	90			86.5	m: 87		m: 80–95
Franklin et al. (2016)³¹			(72/71)															f: 88		f: 81–96
Galic et al. (2015)¹⁶	Croatia		1336		14 –23		91.2	91.9		84.3			95.4	87.7			93.5	m: 91.5		m: 89–93.5
Galic et al. (2015)¹⁶			(578/758)															f: 88.8		f: 86.3–90.9
Gulsahi et al. (2016)³²	Turkey		293		14 –22		94.6	100		85.9			100	90.2			100	m: 97.6
Gulsahi et al. (2016)³²			(128/165)															f: 92.7
Kelmendi et al. (2017)³³	Kossovo		1221		12 –23		96.2	97.6		82.6			99.1	89.4			98.3	m: 96.8		m: 92.6–98.5
Kelmendi et al. (2017)³³			(543/678)															f: 90.9		f: 87–91.7
Kumagai et al. (2019)³⁸	Japan		276		14–24		89	96		84			93	86.5			94.5	m: 91		m: 87–96
Kumagai et al. (2019)³⁸			(139/137)															f: 87		f: 81–93
Różyło-Kalinowska et al. (2018)³⁴	Poland		982		15 –24		86.2	91.2		82.6			93	84.4			92.1	m: 87.6		m: 84.8–90.3
Różyło-Kalinowska et al. (2018)³⁴			(547/435)															f: 85.3		f: 82–88.6
Spinas et al. (2018)³⁵	Sardinia		131		16 –23		85	91		79			100	82			95	m: 87		m: 84.8–90.3
Spinas et al. (2018)³⁵	(Italy)		(63/68)															f: 84		f: 82–88.6
Tafrount et al. (2018)³⁷	France		339		14 –22		87.1	95.3		81.3			96.2	84.2			95.7	m: 91.6		m: 82–91
Tafrount et al. (2018)³⁷			(155/184)															f: 89.7		f: 78–89
Zelic et al. (2015)³⁶	Serbia		589		13 –24		96	94		86			98	91			96	m: 95		m: 92–98
Zelic et al. (2015)³⁶			(290/299)															f: 91		f: 87–92

The 22 studies cover a huge population, from five continents of the world, including Africa, Asia, Australia, Europe and South America. The three great human races (white, black and Asian) are also represented, including some ethnical subgroups such as Latino Americans,^25,29,30 black and white Africans,^22,26,28 Saudi Arabians,²¹ Indians,² Chinese²⁷ and Japanese.³⁸

The total number of individuals recruited in the analysed studies was 12,286 (5723 males and 6563 females) mostly with an age range of 14–24 years. The geographical distribution of the individuals was: 6324 from European countries (Albania, Croatia, France, Italy, Sardinia, Kosovo, Montenegro, Netherlands, Poland and Serbia), 1554 from South America (Brazil, Chile and Colombia), 1830 from Asia (China, India, Saudi Arabia, Turkey and Japan), 2434 from Africa (Botswana, Libya and South Africa) and 143 from Australia.

The dental maturity of the above study group assessed by the I_3M showed an overall sensitivity for both the male and female samples ranging from 51.8% (Saudi Arabia)²¹ to 93.4% (Colombia)²⁹ and a specificity ranging from 86.2% (Brazil)³⁰ to 100% (Libya and Turkey).^28,32 Regarding the proportion of individuals correctly classified as ≥18 years of age (I_3M <0.08), it is worth mentioning that the second lowest value of sensitivity after the one reported in the Saudi Arabia population (51.8%)²¹ was found in Chile (70.5%).²⁵ The overall proportion of correctly classified individuals was between 74% (Saudi Arabia)²¹ and 95% (Turkey).³²

In Figure 1, the distribution of sensitivity, specificity and accuracy calculated for each country is represented by histograms.

Figure 1.

Distribution of sensitivity, specificity and accuracy by country.

In terms of sex, for the male sample only, the sensitivity ranged from 52.3% (Saudi Arabia)²¹ to 96.2% (Kossovo)³³ and the specificity from 85% (Australia)³¹ to 100% (Saudi Arabia, Libya and Turkey).^21,28,32 For the female sample, the sensitivity was between 51.3% (Saudi Arabia)²¹ and 95.1% (Colombia),²⁹ and the specificity was between 67.2% (Brazil)³⁰ and 100% (Libya and Turkey).^28,32 Unfortunately, three studies did not provide data according to sex analysis, and they were from South Africa,²² Chile²⁵ and Italy.^12,35

In terms of geographical distribution, for the European population, the lowest sensitivity value was 78.3% found in Holland,¹⁹ while the highest was 91% reported in Serbia.³⁶ The minimum and maximum values of specificity were 92.1% in Poland³⁴ and 98.3% in Kossovo,³³ respectively. The overall proportion of individuals correctly classified (accuracy) ranged between 85.1% (Sardinia)³⁵ and 93% (Serbia).³⁶ For the African countries, the minimum and maximum values of sensitivity were 80% in South Africa²² and 90.7% in Libya, respectively.²⁸ The specificity range was between 95% (Botswana and South Africa)^22,26 and 100% (Libya),²⁸ while the accuracy ranged between 90% (South Africa)²² and 94.8% (Libya).²⁸ For the Asian population, the sensitivity ranged from 51.8% (Saudi Arabia)²¹ to 90.2% (Turkey),³² the specificity from 94.5% (Japan)³⁸ to 100% (Turkey)³² and the proportion of individuals correctly classified from 74% (Saudi Arabia)²¹ to 95% (Turkey).³² In Latin American populations, the sensitivity ranged from 70.5% (Chile)²⁵ to 93.4% (Colombia),²⁹ the specificity from 86.2% (Brazil)³⁰ to 92.2% (Colombia)²⁹ and the accuracy from 83% (Chile)²⁵ to 92.4% (Colombia).²⁹

Discussion

The demand for forensic age determination in living individuals is growing all over the world. According to the 2017 Global Trends Report, by the end of the 2017, 68.5 million individuals had been identified as having been forcibly displaced throughout the world as a result of feared persecution, internal conflicts or generalised violence.³⁹ About 3.1 million people asylum seekers are still waiting for a decision on their application, and 173,800 are unaccompanied minors.³⁹ Moreover, in developing countries, only half of children younger than five years of age have birth certificates.⁴⁰ For individuals with no valid proof of birth date (also known as legal age), age estimation is therefore crucial, and it represents one of the main requests made by police and judicial authorities.¹⁹ There is wide agreement in the literature that the most appropriate approach to legal age estimation of a living person must include the assessment of somatic and sexual maturity as well as of skeletal and dental maturity by radiographic examinations.^3–5,41,42 Dental growth and emergence have long been recognised as the most useful criteria, especially in adolescents and young adults, and several methods are available.^8,15,16 Unfortunately, the margin of error can be substantial, even when using combined methods.^3,43 In fact, no method is able to estimate age 100% precisely.³² An error in estimating legal age can result in a child being treated as an adult, or an adult as a child.⁴⁴ According to the FASE and AGFAD recommendations, the estimates should necessarily refer to the margin of error or the possible range of variation around the mean associated with the particular method adopted.³ Cameriere’s method provides such a margin of error for males and females according to their geographical distribution. In this regard, the I_3M can provide reliable support for chronological age assessment.^14,19 The proportion of individuals correctly classified as ≥18 years of age (I_3M <0.08) is >70.5% in all countries, except for the Saudi Arabia sample (51.8%).²¹ This can be considered as poor sensitivity if compared to the higher specificity of the I_3M cut-off value ranging from 86.2% (Brazil)³⁰ to 100% (Libya and Turkey).^28,32 However, this is a relevant result considering that, based on the present critical review, the high proportion of subjects correctly classified as <18 years of age (based on I_3M >0.08) is fully consistent with the principles of medical ethics. It has been emphasised¹⁶ that from a forensic perspective, the discrimination performance of any age estimation method should have better specificity than sensitivity, reducing the risk of overestimating the chronological age of adolescence. Based on inaccurate estimates, authorities could incarcerate minors in adult detention centres, which are unsafe and inappropriate for children, not protecting children from child labour or early marriage.^10,41 Minors must be treated as children, with the full benefit of safeguarding through being children in care.¹⁸ Therefore, within a forensic context, it is ethically better to underestimate the age of an individual because of the judicial implications involving a possible adult.

The overall results show a lower sensitivity of 84.3% and higher specificity of 94.4% compared to a previous meta-analysis (94.6% and 88.8%, respectively) that did not include nine recently published research studies^{19,22,23,25,33–35,37,38} dealing with the application of the I_3M to other ethnical subgroups (e.g. Chinese, Japanese, South African and Chilean people). Of course, the I_3M has some limitations mainly related to its variation in position, morphology and development.^6,7,9 Such limitations should be underlined in any expert opinion in order to protect the person from the risk of overestimation.

The results of this meta-analysis show that Cameriere’s I_3M is easy to use and practical for forensic practitioners, with a slight risk of under-estimation in both sexes. The I_3M method also shows good reliability and reproducibility in the main ethnic groups and different subgroups according to their geographical distribution. In fact, the cut-off value (0.08) is a reliable and valuable threshold for discriminating adults and minors (high specificity in both sexes for every population studied), even in geographically isolated populations. It is worth mentioning that positive results of reliability and reproducibility were also found in a unique population such as Sardinians, known as one of the most extreme in its relative lack of heterogeneity.³⁵

Actually, it is not fully understood how ethnic origin can influence tooth development. A comparative study of wisdom tooth mineralisation has been carried in white, Asian and African populations on the basis of Demirjian’s stages.⁴⁵ Based on their results, the authors recommended the use of population-specific standards in order to enhance the accuracy of forensic age estimation.⁴⁵ In this regard, the I_3M method can be applied all over the world, according to the margin of error of living subjects correctly classified as minor (specificity) or adults (sensitivity) found in each country for males and females.

Conclusion

Forensic science evidence is closely linked to the rates at which examiners make errors.^46,47 The I_3M provides such error rates for males and females according to the geographical distribution of population spread across five continents. The risk of overestimating the legal age of adolescence is pretty low using the I_3M method, but inaccurate estimates of chronological age cannot be excluded according to the overall results of sensibility and specificity. Although the I_3M is a suitable method in forensic settings, examiners must be careful in any legal age assessment, keeping in mind the known rate of error of such evidence and its probative value.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Federica Martino

Mariano Cingolani

References

Schmeling

Forensic age estimation. In: Houck

(ed) Forensic anthropology. London: Academic Press, 2017, pp.120–145.

Balla

Galic

Karunaker

, et al. Validation of third molar maturity index (I_3M) for discrimination of juvenile/adult status in South Indian population. J Forensic Leg Med 2017; 49: 2–7.

Introna

Campobasso

CP.

Biological vs legal age of living individuals. In: Schmitt

Cunha

Pinheiro

(eds) Forensic anthropology and medicine. Complementary sciences from recovery to cause of death. New York: Humana Press, 2006, pp.57–82.

Cunha

Baccino

Martrille

, et al. The problem of aging human remains and living individuals: a review. Forensic Sci Int 2009; 193: 1–13.

Schmeling

Geserick

Reisinger

, et al. Age estimation. Forensic Sci Int 2007; 165: 178–181.

AlQahtani

Hector

Liversidge

HM.

Brief communication: the London atlas of human tooth development and eruption.

Am J Phys Anthropol 2010; 142: 481–490.

Ubelaker

DH.

Human skeletal remains: excavation, analysis, interpretation. 2nd ed. Washington, DC: Taraxacum, 1989.

Demirjian

Goldstein

Tanner

JM.

A new system of dental age assessment. Hum Biol 1973; 45: 211–227.

Anderson

Thompson

Popovich

Age of attainment of mineralization stages of the permanent dentition.

J Forensic Sci 1976; 21: 191–200.

10.

Black

Aggrawal

Payne-James

Age estimation in the living: the practitioner’s guide. New York: Wiley-Blackwell, 2010.

11.

Liversidge

HM.

Timing of human mandibular third molar formation.

Ann Hum Biol 2008; 35: 294–321.

12.

De Luca

Pacifici

, et al. Third molar development by measurements of open apices in an Italian sample of living subjects. J Forensic Leg Med 2016; 38: 36–42.

13.

Thevissen

Fieuws

Willems

Human third molars development: comparison of 9 country specific populations.

Forensic Sci Int 2010; 201: 102–105.

14.

Thevissen

Alqerban

Asaumi

, et al. Human dental age estimation using third molar developmental stages: accuracy of age predictions not using country specific information. Forensic Sci Int 2010; 201: 106–111.

15.

Mincer

Harris

Berryman

HE.

The ABFO study of third molar development and its use as an estimator of chronological age. J Forensic Sci 1993; 38: 379–390.

16.

Cameriere

Ferrante

De Angelis

, et al. The comparison between measurement of open apices of third molars and Demirjian stages to test chronological age of over 18 year olds in living subjects. Int J Legal Med 2008; 122: 493–497.

17.

Galić

Lauc

Brkić

, et al. Cameriere’s third molar maturity index in assessing age of majority. Forensic Sci Int 2015; 252: 191.e1–5.

18.

De Luca

Biagi

Begnoni

, et al. Accuracy of Cameriere’s cut-off value for third molar in assessing 18 years of age. Forensic Sci Int 2014; 235: 102.e1–6.

19.

Boyacıoğlu Doğru

Gulsahi

Çehreli

, et al. Age of majority assessment in Dutch individuals based on Cameriere’s third molar maturity index. Forensic Sci Int 2018; 282: 23.

20.

Cameron

Jones

LL.

Growth maturation and age. In: Black

Aggrawal

Payne-James

(eds) Age estimation in the living: the practitioner’s guide. New York: Wiley-Blackwell, 2010, pp.95–129.

21.

AlQahtani

Kawthar

AlAraik

, et al. Third molar cut-off value in assessing the legal age of 18 in Saudi population. Forensic Sci Int 2017; 272: 64–67.

22.

Angelakopoulos

De Luca

Velandia Palacio

, et al. Third molar maturity index (I_3M) for assessing age of majority: study of a black South African sample. Int J Legal Med 2018; 132: 1457–1464.

23.

Antunovic

Galic

Zelic

, et al. The third molars for indicating legal adult age in Montenegro. Leg Med 2018; 33: 55–61.

24.

Cameriere

Santoro

Roca

, et al. Assessment of legal adult age of 18 by measurement of open apices of the third molars: study on the Albanian sample. Forensic Sci Int 2014; 245: 205.e1–5.

25.

Cameriere

Velandia Palacio

Pinares

, et al. Assessment of second (I_2M) and third (I_3M) molar indices for establishing 14 and 16 legal ages and validation of the Cameriere’s I_3M cut-off for 18 years old in Chilean population. Forensic Sci Int 2018; 285: 205.e1–5.

26.

Cavrić

Galić

Vodanović

, et al. Third molar maturity index (I_3M) for assessing age of majority in a black African population in Botswana. Int J Legal Med 2016; 130: 1109–1120.

27.

Chu

Wang

, et al. Third molar maturity index (I_3M) for assessing age of majority in northern Chinese population. Int J Legal Med 2018; 132: 1759–1768.

28.

Dardouri

AAK

Cameriere

De Luca

, et al. Third molar maturity index by measurements of open apices in a Libyan sample of living subjects. Forensic Sci Int 2016; 267: 230.e1–6.

29.

De Luca

Aguilar

Rivera

, et al. Accuracy of cut-off value by measurement of third molar index: study of a Colombian sample. Forensic Sci Int 2016; 261: 160.e1–5.

30.

Deitos

Costa

Michel-Crosato

, et al. Age estimation among Brazilians: younger or older than 18. J Forensic Leg Med 2015; 33: 111–115.

31.

Franklin

Karkhanis

Flavel

, et al. Accuracy of a cut-off value based on the third molar index: validation in an Australian population. Forensic Sci Int 2016; 266: 575.e1–6.

32.

Gulsahi

De Luca

Cehreli

, et al. Accuracy of the third molar index for assessing the legal majority of 18 years in Turkish population. Forensic Sci Int 2016; 266: 584.e1–6.

33.

Kelmendi

Cameriere

Koçani

, et al. The third molar maturity index in indicating the legal adult age in Kosovar population. Int J Legal Med 2018; 132: 1151–1159.

34.

Różyło-Kalinowska

Kalinowski

, et al. Validity of the third molar maturity index I_3M for indicating the adult age in the Polish population. Forensic Sci Int 2018; 290: 352.

35.

Spinas

De Luca

Lampis

, et al. Is the third molar maturity index (I_3M) useful for a genetic isolate population? Study of a Sardinian sample of children and young adults. Int J Legal Med 2018; 132: 1787–1794.

36.

Zelic

Galic

Nedeljkovic

, et al. Accuracy of Cameriere’s third molar maturity index in assessing legal adulthood on Serbian population. Forensic Sci Int 2016; 259: 127–132.

37.

Tafrount

Galić

Franchi

, et al. Third molar maturity index for indicating the legal adult age in southeastern France. Forensic Sci Int 2018; 294: 218.e1–6.

38.

Kumagai

Takahashi

Palacio

LAV

, et al. Accuracy of the third molar index cut-off value for estimating 18 years of age: validation in a Japanese samples. Leg Med (Tokyo) 2019; 38: 5–9.

39.

The United Nations Refugee Agency. Global Trends. Forced Displacement in 2017, www.unhcr.org/globaltrends2017/ (2017, accessed 30 July 2019).

40.

UNICEF. Progress for Children: Achieving the MDGs with Equity, www.unicef.org/publications/files/Progress_for_ChildrenNo.9_EN_081710.pdf (2010, accessed 30 July 2019).

41.

Schmeling

Grundmann

Fuhrmann

, et al. Criteria for age estimation in living individuals. Int J Legal Med 2008; 122: 457–460.

42.

Santoro

De Donno

Marrone

, et al. Forensic age estimation of living individuals: a retrospective analysis. Forensic Sci Int 2009; 193: 129.e1–4.

43.

Schmeling

Olze

Reisinger

, et al. Statistical analysis and verification of forensic age estimation of living persons in the Institute of Legal Medicine of the Berlin University Hospital Charité. Legal Med 2003; 5: S367–S371.

44.

Pinchi

Norelli

Bartolini

Ethical implications of Italian legislation on ‘epilepsy and driving’.

J Med Ethics 2014; 40: 552–527.

45.

Olze

Schmeling

Taniguchi

, et al. Forensic age estimation in living subjects: the ethnic factor in wisdom tooth mineralization. Int J Legal Med 2004; 118: 170–173.

46.

Analysis of errors in forensic science. J Forensic Sci Med 2017; 3: 139–143.

47.

Ribeiro

Tangen

McKimmie

BM.

Beliefs about error rates and human judgment in forensic science. Forensic Sci Int 2019; 297: 138–147.