Comparison between Two Surgical Techniques to Repair Total Anomalous Pulmonary Venous Connection Using propensity Scoreanalysis / 中山大学学报(医学科学版)

ABSTRACT

[Objective]To evaluate the effects of sutureless technique in comparison to conventional techniques for repair of total anomalous pulmonary venous connection(TAPVC)with the method of propensity score analysis.[Methods]From October 2007 to December 2013,179 consecutive patients were enrolled in this study. Patients were operated with sutureless technique(n = 81)or Conventional technique(n=98),and followed up at an interval of 1 month,3 months,6 months and then once a year post-operation. During analysis,three type of propensity-score matching methods,including nearest neighbor caliper matching,Mahalanobis metric matching with propensity score ,optimal full matching were used to create balanced groups of patients receiving each treatment. Surgeons’performance difference was assessed with random frailty proportional hazards models with gamma. Composite endpoints was defined by postoperative death or late death or postoperative pulmonary venous obstruction(PVO),which was evaluated with Kaplan-Meier curve and multivariable Cox proportional hazard model,adjusted by Preoperative-PVO,age,gender,weight and TAPVC type.[Results]Nearest neighbor caliper matching method was the best choice during propensity score analysis. After matching ,sutureless group included 73 patients and Conventional group73 patients. In sutureless group,cardiopulmonary bypass(CPB)time(Z=2.18, P=0.030),cross-clamp time(Z=3.63,P<0.001),rate of composite endpoints(HR 95%CI=0.20(0.06~0.61),P=0.005),late death(HR 95%CI=0.03(0.01~0.55),P=0.017)were significantly better than that in Conventional group. In subgroup analysis ,for patients with pre-PVO,decreased composite endpoints was seen in sutureless group.[Conclusion]Comparison using thepropensity score analysis demonstrated that sutureless strategy for primary repair of TAPVC may associate with decreased mortality rate of post-PVO and CPB time and cross-clamp time.