Performance monitoring of the arterial switch operation: a moving target.

OBJECTIVES Twenty years ago our institution published an analysis of a cluster of failures associated with the arterial switch operation (ASO). The concept of 'near miss' was explored to detect warning signs of suboptimal performance. The aim of the current study was to review the parameters best suited to monitoring early ASO outcomes in the modern setting and re-examine the concept of near misses as failure equivalents. METHODS All ASOs performed in our institution between 1983 and 2012 were reviewed. The experience was divided into three eras (1983-92, 1993-2002 and 2003-12). The cumulative sum graphic for sequential monitoring was used for early mortality. The need to re-establish cardiopulmonary bypass (CPB), CPB time >240 min and extracorporeal membrane oxygenation (ECMO) were explored as variables of near misses. RESULTS The cohort consisted of 606 patients. The 30-day mortality rate was 23% (n = 29) in Era 1, 6% (n = 14) in Era 2 and 1% (n = 3) in Era 3. There were further 4, 8 and 6 deaths between 30 and 90 days in the three eras, respectively. In Era 3, the majority of deaths occurred between 30 and 90 days. In the current era, ECMO and CPB time >240 min as a marker of near miss was associated with an increased risk of death both within 30 days and 90 days after ASO. CONCLUSIONS The 30-day outcomes of ASO have significantly improved over the last 30 years. As life-saving mechanical support after surgery has been implemented more often, an extended 90-day window of reporting can offer a more realistic outcome indicator of performance.