Home | News | Find a Doctor | Ask a Question | Free

Validation of scalp coverage scoring methods for scalp hair loss in AGA


Validation of scalp coverage scoring methods for scalp hair loss in male pattern hair loss (androgenetic alopecia)

Skin Research and Technology
Volume 12 Issue 2 Page 89-93, May 2006

Global evaluation of hair loss in male subjects affected by androgenetic alopecia has been proposed as a means for monitoring changes over time, including placebo-controlled drug efficacy studies. Because of the potential impact of subjectivity (e.g. placebo effect) of clinical investigators, global photographs (GPs) have been introduced as a more objective record. Examination of paired before and after pictures and rating on a seven-point scale (from greatly decreased −3 to greatly increased +3) have been historically introduced by United States of America (US) experts.

Based on published GPs and original GPs obtained at our clinical research facility, we developed a training set in order to allow European Union (EU) observers to practice and compare with ratings by the US experts.

After training with the seven-point scale, there was a positive correlation between three US and three EU ratings (n=52 paired images from 35 different subjects, r=0.795). The results of a test–retest evaluation was performed on 18 paired images from the initial image collection by the three EU experts. Correlation r=0.806 and identical scores in 78% of cases documents a reproducibility similar to the single one US expert published data (119 subjects, retest correlation 0.76 with 75% identical duplicate ratings).

Seventeen subjects taken from a placebo-controlled trial had GPs at 6 and 12 months. The average difference between an efficacious drug treatment and the placebo were almost similar in the US (0.833) as in the EU (0.689) expert panels. We also trained the EU experts in performing the scalp coverage scoring (SCS), a novel system for the global evaluation of scalp hair in vivo and on GP. SCS was performed on single images (randomised as to time and treatment) taken from the same set of 17 paired GPs. This showed a between-group difference of 0.055 at 6 months and 0.201 at 12 months, i.e. 5% improved coverage in favour of the active group.

After completion of our study, US and trained - calibrated EU experts seem equally valuable in comparing before–after GPs. SCS can also be used on GPs and may support the clinical investigator during inclusion of test subjects and for real-time efficacy evaluation during the trial.