The purpose of the present study was to quantify the learning curve for arthroscopic os trigonum excision using the log-linear model. Twenty-three consecutive feet underwent arthroscopic os trigonum excision and release of the flexor hallucis longus. The required time from the beginning of shaving of the soft tissue until completion of os trigonum excision and release of the flexor hallucis longus (van Dijk time) was recorded. Regression analysis was applied to predict the required time on the basis of the cumulative case volume after logarithmic transformation of both statistics. The mean required time was 35.2 (range 9 to 90) minutes. After logarithmic transformation, a significant linear correlation was observed between the required time and the cumulative case volume (p = .0043). The best-fit linear equation was calculated as log (y, estimated required time) = −0.41 log (x, case volume) + 1.86, resulting in an estimated learning rate of 75.3% (= 2−0.41). The results showed an overall time reduction in arthroscopic os trigonum excision in support of a learning curve effect with an ˜75% learning rate, indicating that the required time for arthroscopic os trigonum excision can decrease by ≤25% when the cumulative volume of cases has doubled.
Level of Clinical Evidence: 2