Error rates for unvalidated medical age assessment procedures

Area of Science:

• Forensic medicine
• Biostatistics
• Demography

Background:

• Sweden faced a surge in asylum applications, including many unaccompanied minors, leading to distrust in age claims by authorities.
• The Swedish National Board of Forensic Medicine (RMV) developed a medical age assessment procedure using dental and femur imaging.
• A lack of validation studies for RMV's procedure prompted an investigation into its accuracy with 2017 data.

Purpose of the Study:

• To develop a stochastic model to evaluate the consistency of RMV's age assessment procedure with observed data.
• To analyze the relationship between dental and femur maturity indicators in male asylum seekers.
• To estimate the accuracy and potential misclassification rates of the RMV age assessment system.

Main Methods:

• Development of a general stochastic model to analyze age indicator parameters and population profiles.
• Utilized 2017 data from 9617 males and 337 females subjected to RMV's age assessment procedure (dental and femur imaging).
• Statistical analysis to estimate parameters, assess maturity sequences, and calculate misclassification risks.

Main Results:

• Contrary to RMV claims, femur maturity was observed to occur significantly before dental maturity on average.
• An estimated 15% of tested males were children, with a 33% risk of being misclassified as adults.
• An estimated 7% risk of adults being misclassified as children was observed.

Conclusions:

• The RMV's age assessment procedure shows potential biases, with femur maturity occurring earlier than dental maturity.
• Significant misclassification risks exist for both children and adults, questioning the reliability of the current system.
• Further validation and refinement of medical age assessment methods are crucial for accurate determination of age in asylum seekers.