Quality control for intravascular intrauterine transfusion using cumulative sum (CUSUM) analysis for the monitoring of individual performance.

INTRODUCTION: Intravascular intrauterine transfusion (IUT) is an effective and relatively safe method for the treatment of fetal anemia. Although implemented in centers all over the world in the 1980s, the length and strength of the learning curve for this procedure has never been studied. Cumulative sum (CUSUM) analysis has been increasingly used as a graphical and statistical tool for quality control and learning curve assessment in clinical medicine. We aimed to test the feasibility of CUSUM analysis for quality control in fetal therapy by using this method to monitor individual performance of IUT in the learning phase and over the long term.
METHODS: IUTs performed in the Dutch referral center for fetal therapy from 1987 to 2009 were retrospectively classified as successful or failed. Failed was defined as no net transfusion or the occurrence of life-threatening procedure-related complications. The CUSUM statistical method was used to estimate individual learning curves and to monitor long-term performance. Four operators who each performed at least 200 procedures were included.
RESULTS: Individual CUSUM graphs were easily assessed. Both operators pioneering IUT in the late 1980s had long learning phases. The 2 operators learning IUT in later years in an experienced team performed acceptably from the start and reached a level of competence after 34 and 49 procedures. DISCUSSION: CUSUM analysis is a feasible method for quality control in fetal therapy. In an experienced setting, individual competence may be reached after 30 to 50 IUTs. Our data suggest that operators need at least 10 procedures per year to keep a level of competence.