Infrapopliteal Percutaneous Transluminal Angioplasty Versus Bypass Surgery as First-Line Strategies in Critical Leg Ischemia: A Propensity Score Analysis Soderstrom MI, Arvela EM, Korhonen M et al. Ann Surg 2010;252:765-73.Conclusion: Infrapopliteal percutaneous angioplasty, when feasible, can be an effective first-line strategy achieving similar long-term results as bypass surgery in patients with critical limb ischemia (CLI) when redo procedures are actively used.Summary: Retrospective studies of patient characteristics in endovascular and surgical trials often differ significantly. Propensity score analysis can be used to adjust for important differences between treatment groups. In this study, propensity score analysis was used to compare outcomes in patients who underwent percutaneous transluminal angioplasty (PTA) or bypass surgery of infrapopliteal arteries. The study cohort consisted of 1023 nonrandomized patients treated for CLI, with 262 endovascular and 761 surgical revascularization procedures performed on tibial or pedal arteries. A propensity score was used for stratification and for adjustment in multivariable analysis. Choice of treatment was based on physician preference and treatment feasibility. PTA and bypass surgery achieved similar 5-year limb salvage (75.3% vs 76.0%), similar survival (47.5% vs 43.3%), and amputationfree survival (37.7% vs 32.7%). Rates of freedom from further revascularization were also similar (73.3% vs 74.4%). Freedom from surgical revascularization was higher after bypass surgery (94.3% vs 86.2%, P Ͻ .001). Outcomes were similar with propensity-score matched pairs except for freedom from surgical revascularization, which was higher in the bypass surgery group (91.4% vs 85.3% at 5 years, P ϭ .045). In patients who underwent isolated infrapopliteal revascularization, PTA was associated with better limb salvage (75.5% vs 68.0%, P ϭ .042). Differences in freedom from surgical revascularization were not significant (78.8% vs 85.2%, P ϭ .17). The difference in the limb salvage rate was not maintained in propensity score-matched pairs (P ϭ .12).Comment: Propensity score analysis is an interesting mathematical exercise. However, as pointed out by Dr J. F. Hamming in a discussion after the article, the authors are still comparing "apples with oranges," in that endovascular treatment was reserved for patients with shorter lesions and mostly stenosis and not occlusions. The underlying problem of trying to compare treatment modalities in patients with differing extent of disease still exists. What the paper really demonstrates is that it is possible to be reasonably good at individualizing treatment in patients with CLI. The overall effectiveness of the different treatment modalities in patients with equal burden of disease is not answered by this study.