Thoracoscopic Versus Open Mitral Valve Repair: A Propensity Score Analysis of Early Outcomes

Background: The very low risk of mitral valve repair performed through median sternotomy must be reproducible when using a port-access approach to justify early repair employing minimally invasive platforms. We compared the outcomes of mitral valve repair performed through port access using thoracoscopic assistance (port) versus median sternotomy (open). Methods: The early results after mitral valve repair performed by two different surgeons at two separate institutions were analyzed. Between January 1999 and December 2006, isolated mitral valve repair was performed with a port approach in 350 patients and an open approach in 365 patients. Results: The mean age was similar between the two groups; however, port patients were more frequently female (148 [42%] versus 119 [33%], p = 0.007), and had a higher likelihood of having New York Heart Association class III to IV symptoms (100 [29%] versus 48 [13%], p < 0.001), diabetes mellitus (19 [5%] versus 8 [2%], p = 0.023), congestive heart failure (90 [26%] versus 26 [7%], p < 0.001), and a lower ejection fraction (53% versus 64%, p < 0.001) preoperatively. Cross-clamp time (104 versus 24 minutes, p < 0.001) and bypass time (140 versus 33 minutes, p = 0.001) were significantly lower for the open group. On univariate analysis, the duration of postoperative ventilatory support was significantly lower in the port group (5.0 versus 11.0 hours, p < 0.001); however, the length of hospital stay was longer (6.95 versus 6.19 days, p < 0.001). There were 2 early deaths (2 port versus 0 open). A propensity score factor was calculated and utilized to account for differences between groups. After adjusting for propensity score and significant factors identified in multivariate models, port mitral repair independently predicted a diminished duration of postoperative ventilatory support (p = 0.045), but there were no significant differences in other outcomes including postoperative blood transfusion, reoperation for hemorrhage, or length of stay in hospital. Conclusions: Despite longer cross-clamp and bypass times, early outcomes using a thoracoscopic port-access approach were similar to those for mitral valve repair performed through median sternotomy. Minimally invasive mitral valve repair was associated with a shorter time to extubation, but that did not translate into a diminished duration of postoperative hospitalization.