A study of 1,106 children and adolescents in Kosovo found that a modified version of Demirjian's dental age estimation method provided more accurate results than the original French-Canadian standards. Researchers said population-specific adaptations are needed to improve reliability in clinical and forensic settings.
"The present study demonstrates the need for population specific adaptations of the Demirjian method and proves that modification of the method provides more reliable results when compared to the Kosovar population," said Jeta Kelmendi, DDS, PhD, of the Department of Orthodontics, Alma Mater Europaea, Campus Rezonanca, Kosova, and colleagues.
Digital panoramic radiographs from 563 girls and 543 boys aged 6 to 16 years were collected at the University Dentistry Clinical Center of Kosovo between 2011 and 2015. Radiographs were included only if all seven left mandibular teeth were visible. Exclusions were made for missing teeth, orthodontic treatment, fractures, pathologic findings, or poor image quality.
Teeth were staged using Demirjian's A–H classification. Stages were converted into sex-specific maturity scores for the Kosovar population and compared with French-Canadian standards. Two trained evaluators, blinded to patient age and sex, independently scored the radiographs to assess reliability.
Using Kosovar-specific maturity scores reduced discrepancies between dental and chronological age. On average, dental age was underestimated by 0.1 years in girls and 0.2 years in boys. By comparison, French-Canadian standards significantly overestimated age, especially in younger patients.
Dental maturity in the Kosovar population started at about 6 years and peaked at 7 years, earlier than in the French-Canadian reference group. Accuracy was highest in girls aged 7.5 to 13 years and boys aged 7.5 to 12 years.
The correlation between dental and chronological age was very strong, with coefficients of 0.997 for girls and 0.988 for boys. Most deviations were within 1 year, and discrepancies greater than 3 years were rare. Inter-rater and intra-rater reliability were high, with kappa values ranging from 0.62 to 0.88.
Third molars were not assessed, which limited analysis in older adolescents. Only left mandibular teeth were evaluated, which, while standard practice, narrowed comprehensiveness. The study relied on a single population, and further validation is needed in other Kosovar and regional groups.
The modified Demirjian method improved precision and remains a reliable tool in dentistry, orthodontics, and forensic identification. Researchers noted that such population-specific datasets could serve as valuable training inputs for AI-driven systems and automated dental age estimation models.
The authors reported no conflicts of interest.
