Related Experiment Video
Updated: Jun 9, 2026

A Computer-Based Platform for Aiding Clinicians in Eating Disorder Analysis and Diagnosis
Published on: May 10, 2022
Development and Validation of Puberty Interpreter, a Mobile Clinical Decision-Support Tool for Pubertal Disorders
Vibha Yadav1, Anurag Bajpai2, Sayan Banerjee3
1Department of Pediatric & Adolescent Endocrinology, Regency Hospital, Kanpur, UP, India.
Objective:
To develop and validate Puberty Interpreter, a mobile application-based point-of-care clinical decision support tool for pubertal disorders.
Methods:
Puberty Interpreter recommendations in 306 children with puberty concerns (199 early and 107 delayed puberty) were compared to the gold standard, a pediatrician, a senior trainee, and a junior trainee.
Results:
Puberty Interpreter concordance with the gold standard was 303/306 (99.0%) for diagnosis, 303/306 (99.0%) for management, and 606/612 (99.0%) overall. This was higher than the pediatrician (476/612, 77.8%), the senior trainee (330/612, 53.9%), and the junior trainee (305/612, 49.8%). The relative risk of error compared with the Puberty Interpreter was 21 for the pediatrician (95% CI 6.7-66), 41 for the senior trainee (95% CI 13-127), and 44 for the junior trainee (95% CI 14-136). Agreement beyond chance for management was highest for the application (κ = 0.978; overall agreement 0.990) and lowest for the junior trainee (κ = 0.233; overall agreement 0.683). Adopting the Puberty Interpreter outputs would have reduced diagnostic discordance by 96 cases (96.9%) for the pediatrician, 197 (98.5%) for the senior trainee, and 212 (98.6%) for the junior trainee. Estimated economic benefit of adhering to Puberty Interpreter guidance was ₹56,700 for the pediatrician ($683; ₹185 [$2.2] per case), ₹236,250 for the senior trainee ($2,846; ₹772 [$9.3] per case), and ₹220,500 for the junior trainee ($2,657; ₹721 [$8.7] per case).
Conclusions:
The Puberty Interpreter delivers accurate guidance, substantially reducing both over- and under-investigation compared to pediatricians and trainees. These findings support implementation and multi-center prospective evaluation.
