Decision making in the multiligament-injured knee: an evidence-based systematic review.

PURPOSE The purpose of this systematic review was to address the treatment of multiligament knee injuries, specifically (1) surgical versus nonoperative treatment, (2) repair versus reconstruction of injured ligamentous structures, and (3) early versus late surgery of damaged ligaments. METHODS Two independent reviewers performed a search on PubMed from 1966 to August 2007 using the terms "knee dislocation," "multiple ligament-injured knee," and "multiligament knee reconstruction." Study inclusion criteria were (1) levels I to IV evidence, (2) "multiligament" defined as disruption of at least 2 of the 4 major knee ligaments, (3) measures of functional and clinical outcome, and (4) minimum of 12 months' follow-up, with a mean of at least 24 months. RESULTS Four studies compared surgical treatment with nonoperative treatment. There was a higher percentage of excellent/good International Knee Documentation Committee (IKDC) scores (58% v 20%) in surgically treated patients, as well as higher rates for return to work (72% v 52%) and return to full sport (29% v 10%). Two studies compared repair with reconstruction of damaged structures, with similar mean Lysholm scores (88 v 87) and excellent/good IKDC scores (51% v 48%). However, repair of the posterolateral corner had a higher failure rate (37% v 9%). Similarly, repair of the cruciates yielded decreased stability and range of motion and a lower return to preinjury activity levels (0% v 33%). There were 5 studies comparing early surgery (<or=3 weeks) with late surgery. Early treatment resulted in higher mean Lysholm scores (90 v 82) and a higher percentage of excellent/good IKDC scores (47% v 31%), as well as higher sports activity scores (89 v 69) on the Knee Outcome Survey. CONCLUSIONS Our review suggests that early operative treatment of the multiligament-injured knee yields improved functional and clinical outcomes compared with nonoperative management or delayed surgery. Repair of the posterolateral corner may yield higher revision rates compared with reconstruction.