Regional Variations in Physicians’ Attitudes and Recommendations Surrounding Implantable Cardioverter-Defibrillators

Study Questions:

Are there regional variations in physician attitudes towards the use of implantable cardioverter-defibrillators (ICDs) in the United States?


A web-based survey was sent to a random sample of 9,969 members of the American College of Cardiology. The rate of ICD use in each Hospital Referral Region (HRR) was determined from the National Cardiovascular Data Registry (NCDR®). Surveyed physician attitudes toward general ICD use were determined using a 5-point Likert scale (ranging from “strongly disagree” to “strongly agree”). Then, patient scenarios were given and respondents graded ICD implantation on a 4-point Likert scale (ranging from “not at all likely” to “very likely”). Physicians were categorized into quintiles based on HRR utilization rates. Multivariable logistic regression was then used to assess HRR-level ICD usage rate and physician attitudes on ICD implantation.


There were 1,124 (11% of 9,969 queried) eligible physician survey responders. Over 65% of respondents were from solo or private practices, and 32% were from academic institutions. General cardiologists and electrophysiologists represented 54% and 12% of the group, respectively. Significantly more respondents than nonrespondents were from academic centers, and more respondents were electrophysiologists. Physicians did not differ in age, practice specialty, or years in practice based on HRR utilization quintile. There was no significant association between HRR utilization rates and physician attitudes toward recommending ICD implant in general. Over 86% of physicians felt an ICD would make a general patient live longer and <9% felt they would improve a general patient’s quality of life. There were significant differences in responses to specific patient scenarios. Compared with physicians in low ICD utilization areas, physicians practicing in high utilization areas were significantly more likely to implant a device in a “frail patient” (21% lowest utilization quintile vs. 32% highest utilization quintile) or a patient with a life expectancy <1 year (5% in the lowest utilization quintile vs. 10% in the highest) (both p < 0.05 for trend across ICD utilization quintile). Physicians from high ICD utilization areas demonstrated a nonsignificant trend in being more likely to recommend an ICD to an 80-year-old patient (30% lowest vs. 41% highest utilization quintile, p = 0.08).


The authors concluded that there is no difference in the general physician attitude regarding ICD implant across HRR utilization areas, but physicians in higher utilizations areas are more likely to implant devices in patients who fall outside of guideline indications.


This is an interesting examination of the general physician attitudes toward ICD implant. In general, physicians concur that ICDs provide mortality benefit. The response amongst physicians regarding the benefit and appropriateness of ICD referral for those patients >80 years of age and those who are frail was appropriately lower, as this is where clinical trial data are less supportive of benefit. The reason why physicians practicing in higher utilization centers are more likely to support ICD use outside of guideline recommendations is not clear—it may reflect the relationship between volume and financial incentive, or it could reflect a greater comfort level for implanting devices in a higher-risk patient who is evaluated on a ‘case by case, nonguideline-driven basis.’

Clinical Topics: Arrhythmias and Clinical EP, Heart Failure and Cardiomyopathies, Implantable Devices, SCD/Ventricular Arrhythmias, Acute Heart Failure

Keywords: Physicians, Attitude, Private Practice, Heart Failure, Data Collection, General Practitioners, Defibrillators, Implantable, United States

< Back to Listings