Physician Experience and the Risk of Procedural Complications
Abstract & Commentary
By John P. DiMarco, MD, PhD, Professor of Medicine, Division of Cardiology, University of Virginia, Charlottesville
Source: Freeman JV, et al. Physician procedure volume and complications of cardioverter-defibrillator implantation. Circulation 2012;125:57-64.
In this study, Freeman and colleagues analyzed data from the National Cardiovascular Data Registry-ICD Registry (NCDR-ICD Registry) to examine the effects of physician procedure volume on in-hospital complications and death. The authors collected data on all patients who had data about their ICD implant recorded in the NCDR-ICD Registry participating hospitals between April 2006 and March 2010. Patients who received an epicardial lead and who had an ICD replacement procedure were not included. In-hospital events including cardiac arrest, cardiac perforation, valve injury, coronary venous dissection, hemothorax, pneumothorax, deep vein thrombosis, TIA or stroke, myocardial infarction, pericardial tamponade, AV fistula, lead dislodgement, peripheral embolus, peripheral nerve injury, and device-related infection were analyzed as early complications. Data on the physician implanter were collected using their National Provider Identifier number. Physician certification was correlated with the databases of the American Board of Internal Medicine, the Society of Thoracic Surgery, and the American College of Surgery. Physicians were characterized as either electrophysiologists, nonelectrophysiologist cardiologists, thoracic surgeons, physicians who met Heart Rhythm Society training standards, or none of the above.
The final dataset included 4011 physicians who performed 356,515 initial ICD implantations at 1463 hospitals. Annualized physician procedure volume varied widely. The median number was only 16 implantations per year. The lowest quartile physicians implanted ≤ 4 ICDs per year! The highest quartile physicians implanted more than 37 ICDs per year. Higher volume implanters treated patients who were slightly older, more likely to have a diagnosis of congestive heart failure, and more likely to have advanced heart failure symptoms. High-volume implanters also were more likely to be board-certified cardiac electrophysiologists. Physician volume was higher in hospitals than in teaching institutions located in urban areas with more patient beds.
The overall in-hospital complication rate after ICD implantation was 3.1%. The rate of adverse events was lower among patients who received a single chamber ICD (1.9%) compared to those patients who received a dual chamber ICD (2.9%) or a biventricular ICD (4.1%). The most common in-hospital adverse events were: lead dislodgement (1%), hematoma (0.9%), pneumothorax (0.4%), and cardiac arrest (0.3%). There was an inverse relationship between physician procedure volume and a rate of adverse events. There was a progressive decline in the frequency of both all complications and major complications with each quartile of physician annual procedure volume. This was particularly striking as the complexity of the devices increased. Physician annual ICD volume was associated with lower event rates even after adjustment for patient characteristics, hospital characteristics, and hospital volume. Since physicians may implant ICDs at more than one hospital, a separate analysis was performed to separate the effects of physician volume from hospital volume. Individual physician ICD procedure volume was still related to better outcomes whether the procedure was performed at higher volume or lower volume hospitals. Board-certified electrophysiologists tended to be the highest volume implanters and also had a lower rate of complications. However, even within each classification of training, procedure volume was still strongly associated with improved results.
The authors conclude that their data are in agreement with previous studies on other cardiac procedures showing that individual physician volume correlates with better outcomes. This effect is most striking for higher risk procedures such as biventricular ICDs. The effect appears to be physician and not hospital-volume specific.
Commentary
This is the largest and most comprehensive analysis of procedure volume and ICD complications yet published. As with other cardiac procedures, operator experience is related to a decreased rate of complications. What is striking is that the number of ICD implants in the lowest quartile of physicians was ≤ 4 per year. Surely these physicians shouldn't be implanting ICDs. Even the next quartile of physicians implanted only 4-16 devices per year and it's questionable if this is an adequate volume. It should also be remembered that the ICD Registry only deals with in-hospital complications. Some ICD complications, like LV lead dislodgements and infections, typically occur later after many ICD recipients have been discharged. It's likely that including these later complications would make the argument for more experienced implanters even stronger.
In this study, Freeman and colleagues analyzed data from the National Cardiovascular Data Registry-ICD Registry (NCDR-ICD Registry) to examine the effects of physician procedure volume on in-hospital complications and death.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content