Publication Date



implantable cardioverter-defibrillator, ICD, morbidity, mortality, risk score


Background: Aurora Health Care, a system of 14 acute care hospitals in eastern Wisconsin, has been a long-time participant in the American College of Cardiology’s National Cardiovascular Data Registries, submitting data to its ICD Registry™ since 2005. Our system’s implantable cardioverter-defibrillator (ICD) procedure volume averages 930 cases annually. During 2012 we experienced an increase in in-hospital mortality/morbidity for ICD cases.

Purpose: A single-center study examining in-hospital mortality/morbidity post-ICD implant before and after changes in practice and patient selection.

Methods: ICD implants and generator changes discharged from January 1, 2009, to December 31, 2012, were included in developing a risk model predicting in-hospital mortality/morbidity. The risk score was shared with physicians for clinical input. A point system was developed, including those factors with highest risk. Using the defined factors, a risk score > 14 was used to indicate those at highest risk for morbidity/mortality. The risk score model was fit on the development group (2009–2012), and then re-run for the intervention cohort from January 1, 2013, to June 30, 2014. Logistic regression was used in the risk model development and validation. Continuous variables were compared using Student’s t-test, and categorical variables were compared using chi-square test.

Results: From 2009 to 2012, 3,417 ICD implants and generator changes were performed and included in risk model development. Of those, 200 (5.9%) patients were indicated as high risk with a score > 14. From January 2013 to June 2014, 1,057 implants and generator changes were performed, with 41 (3.4%) patients indicated as high risk with a score > 14. In the development phase, mean age was 67 years and 70% of patients were male. Post-model development, mean age was 66 years with 72% male. For patients indicated as high risk, in-hospital mortality/morbidity dropped from 20 (10%) to 2 (4.9%), though the decrease was not statistically significant (P = 0.39).

Conclusion: Awareness of high-risk patients and changes in patient selection can lead to improvement in in-hospital mortality/morbidity among those high-risk patients. Although the improvement was not statistically significant, this was most likely due to low volumes and we will continue to monitor outcomes among these patients.




November 9th, 2015


November 16th, 2015


To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.