Volume-outcome relationship (VOR) of complications in cardiac implantable electronic device (CIED) implantation. Evaluation of 250,000 patients from external quality assurance from 2010 to 2020

The number of CIEDs implantations has steadily increased over the past decade and is one of the most common procedures in cardiology. There are approximately 230 actively implanting hospitals in the German federal state North Rhine-Westphalia (NRW) with 17.9 million inhabitants. A large variation in implantation numbers exists between hospitals, from < 10 to 1,234 cases per year. Peri-/post-operative complications (C) such as pneumothorax, hemothorax, pericardial effusion, pocket hematoma, wound infection and lead problems (dislocation and dysfunction) are not uncommon after pacemaker (PM) and defibrillator (ICD) implantation. We analyzed whether the C rate depends on the number of implantations in the hospital which has not been clarified statistically for such a wide variability of implantation procedures.

The analysis is based on the data of all first implantation of CIEDs (including CRT systems) performed between January 1st, 2010 and December 31st, 2020 in any hospital in NRW. All peri-/post-operative C documented as part of the mandatory quality assurance were recorded. A potential association between implantation volume and quality (VOR) was evaluated using the statistical software R, based on the report on "Data evaluation on minimum quantities" (IQTIG 2020). The method includes risk adjustment and takes into account the high statistical uncertainty with small case numbers.

A total of 249,615 cases were examined (PM 73.4%, ICD 26.6%, male 61.3%). ICD patients (pts) were younger (66 y) than PM pts (76 y, p < 0.001). The total number of C was 2.4% (PM 2.6%, ICD 1.6, p<0.001). The most C was lead related at 1.5% (PM 1.7%, ICD 0.7%, p<0.001). Pts with C were older (median 78 y vs 76 y, p<0.001). The forest plot analysis of the general risk factors resulted in an odds ratio (OR) in PM/ICD pts for the ASA class 3 of 1.7/2.1 and the class 4 of 4.2/6.1 (reference ASA class 1). ORs for the type of PM-system were 1.5/2.2 for DDD and 1.0/2.1 for CRT (reference VVI). The VOR revealed a statistically significant relationship between the number of implantations and the occurrence of C, especially for PM patients (p<0.001, AUC 0.7) (see figure). Further evaluation showed that, for example, with a cut-off of 50 annual implantations ~59 of 377 (~16%) C could be avoided and ~2,500 of 16,073 (~16%) pts would have to be redirected to other hospitals. In case of cut-off of 25 annual implantations ~30 of 377 (~8%) C could be avoided and ~850 of 16,073 (~5%) pts. would have to be redirected to other hospitals.

In addition to the risk factors that have been analyzed in the literature, such as age, high ASA class or type of system, the VOR demonstrated a significant influence of the implantation volume. With a relevant number of possibly avoidable C, minimum quantities can be justified and consequences can be estimated.

VOR for PM implantations in 2010-2020