Forensic age estimation of adolescents is an important field of research and practice to ensure that unaccompanied, young asylum seekers receive their rights and adults are not treated as minors. In this systematic review, we summarized evidence regarding age estimation on knee and ankle ossification using computed tomography (CT) and magnetic resonance imaging (MRI).
We found no relevant studies using CT, but four using MRI. These studies were from France, Germany and Turkey, and included 1250 participants between 8 to 30 years old. Two different classification methods were presented for knee ossification, and one method for ankle ossification. All three methods showed good intra- and inter-observer reliability. However, most of the studies were conducted with limited number of participants in each ossification stage and/or had uneven number of participants in each age group, which may lead to substantial variation in chronological age distribution in each development stage.
Given the limited and potentially biased results, we decided not to conduct meta-analysis. More studies with sufficient sample size and a uniform age structure are warranted for more accurate and reliable age estimation using these methods.
Age estimation of living adolescents and young adults has become increasingly important in modern society, especially for asylum seekers who come to Norway without legal documentation of their chronological age. It is therefore necessary to assess chronological age in forensic practice to ensure that children receive their entitled rights and that adults are not treated as minors. Age assessments based on skeletal development of left hand-wrist and third molar teeth using radiographs have been used in Norway for years. We have previously published systematic reviews assessing the agreement between chronological age and skeletal development with the Greulich & Pyle atlas for left hand-wrist radiographs, the Demirjian’s stages for the third molar teeth, and computed tomography (CT) and magnetic resonance imaging (MRI) for medial clavicle ossification, respectively. However, methods using knee and ankle ossification for age estimation have not been reviewed. We therefore conducted a systematic review to evaluate the evidence of chronological age distribution using identified methods of knee and ankle ossification by CT and MRI.
We searched for studies in the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase and Google Scholar. This was a joint search conducted for studies using radiographs of left hand-wrist, third molar teeth, CT and MRI of the medial clavicle, knee and ankle in both males and females. An update literature search was conducted in April 2017 for clavicle, knee and ankle. Two of the authors screened the literature independently from the title and abstract first, and subsequently full-text screening. We included studies that presented age distribution according to knee and ankle ossification stages. Two of the authors independently assessed the risk of bias and applicability of the included studies based on the QUADAS-2 checklist. The findings were summarized as forest plots. The mean chronological age and the standard deviation were summarized in each ossification stage for males and females, respectively.
We found 10059 references in the first literature search and 663 in the second. In total 28 potentially relevant publications were forwarded for full-text screening. We did not find any CT studies matching the inclusion criteria. Four MRI studies were included. The included studies were from France (one study), Germany (one study) and Turkey (two studies). These studies were published from 2012 to 2016, involving 1250 participants from 8 to 30 years old. Two different methods were reported for knee ossification (Dedouit’s method and Kramer’s method) and one for ankle ossification (Saint-Martin’s method). Most of the studies showed good intra- and inter-observer reliability (Κ > 0.80). However, all of the included studies showed high risk of age mimicry bias due to uneven number of participants in each age group. Besides, three of the included studies had relatively small sample size in most of the age groups (n<15), leading to large variation in age estimation. Most of the pooled estimates of chronological age and the distribution in knee and ankle ossification showed high heterogeneity (I2 > 75%).
We only found four studies meeting our inclusion criteria, yielding limited evidence on age estimation by knee and ankle ossification methods. Besides, the results from the included studies suffered from relatively small sample size and uneven number of participants in each age group, leading to substantial variation between studies. Nevertheless, these findings are in line with our previous systematic reviews on age estimation based on the development of third molar teeth and medial clavicle, revealing the potentially high risk of age mimicry bias and uncertainty associated with the assessments in age estimation studies.
Three MRI classification methods were used in four studies describing chronological age distribution in knee and ankle ossification, which yields very limited evidence for us to assess the validity of these methods. Furthermore, the majority of the included studies had limited and uneven sample size in each age group, leading to substantial variation on observed age distribution across studies. We did not find any relevant studies using CT classification method on knee or ankle ossification. Future studies with even numbers in each age group, wide age spectrum and sufficient sample size are warranted for a better understanding of the age distribution in the knee and ankle ossification stages.