Arthroscopic visualization of the posterior compartments of the knee

The purpose of this study is to test the hypothesis that the efficacy of routine arthroscopic evaluation of the posterior compartments of the knee outweighs the morbidity or inefficiency.

Prospective cohort analysis.

One hundred consecutive knees underwent posteromedial and posterolateral evaluation according to an algorithm designed to maximize efficiency of technique. Loose bodies were defined as expected when preoperative imaging or arthroscopy of the rest of the knee revealed loose bodies. The number of attempts required for visualization or visualization not accomplished, morbidity (subjective femoral condylar scuffing or complications associated with the technique), and findings were recorded.

With regard to posteromedial, directly inserting the camera (as opposed to an obdurator) resulted in instrument breakage. In this study, 82% of posteromedial compartments were visualized on the first attempt, and 3% were not visualized. We found that 66% of femoral condyles had no damage, 28% mild, 3% moderate, and 3% severe. Loose bodies were found in 4 of 11 (36%) when expected and 0 of 86 when not expected, a significant difference. Five other posteromedial positive findings were observed. With regard to posterolateral, 93% of compartments were visualized on the first attempt, directly with the camera via the anterolateral portal. All were visualized. We found that 96% of femoral condyles had no damage, 4% mild, and 0% moderate or severe. Damage occurred significantly less frequently than with posteromedial. No findings were noted.

Despite inefficiency, we recommend an obdurator for posteromedial visualization via the anterolateral portal. Posteromedial visualization was associated with morbidity but was efficacious in some cases and is strongly recommended when loose bodies are expected. Posterolateral visualization directly with the camera and via the anterolateral portal is efficient and associated with minimal morbidity but was not efficacious in the cohort evaluated.

Level II.