After the investigation by the US Department of Justice (DOJ) into the overuse of implantable cardioverter-defibrillators (ICDs) not meeting the Centers for Medicare & Medicaid Services’ 2005 National Coverage Determination (NCD) criteria, the rate of non–NCD-compliant ICD use significantly decreased in the hospital setting, according to a study published in JAMA.1

The DOJ investigated ICDs used that did not meet NCD criteria after a whistle-blower submission in 2008 against 1300 hospital centers under the False Claims Act. By 2015, a total of 457 hospitals reached settlements with the DOJ, representing more than $250 million attributable to non–NCD-compliant ICDs. In the JAMA study, investigators performed a cross-sectional analysis of 300,151 ICDs offered to Medicare beneficiaries with ischemic or nonischemic myopathy across 1809 hospitals in the United States.

The primary outcome was comprised of the proportion of initial primary prevention ICDs that was not compliant with the criteria set forth by the NCD. Secondary outcomes included the number of initial primary and secondary preventive ICDs over time as well as the proportion of non–NCD-compliant primary prevention ICDs used in non-Medicare beneficiaries. The study examined ICD use rates between 2007 and 2015.

Continue Reading

A significantly greater proportion of hospitals who settled with the DOJ in 2015 used initial ICDs not meeting NCD criteria in 2007 compared with non-settlement hospitals (25.8%; 95% CI, 24.7%-26.8% vs 22.8%; 95% CI, 22.1%-23.5%, respectively; P <.001). The relative decrease in non-NCD compliant ICDs was greater during the DOJ investigation among settlement vs non-settlement hospitals (62.7%; 95% CI, 59.2%-66.1% vs 53.2%; 95% CI, 50.4%-56.0%, respectively; P <.001) whereas the absolute decrease was also significantly greater for settlement hospitals (16.1%; 95% CI, 14.8%-17.5% vs 12.1%; 95% CI, 11.2%-13.0%; P <.001).

Between 2007 and 2015, the relative decrease in the volume of implanted ICDs was greater at settlement vs non-settlement hospitals (32.8%; 95% CI, 29.9%-35.7% vs 17.4%; 95% CI, 14.8%-20.0%, respectively; P <.001). Additionally, similar findings were observed for the absolute decrease in the volume of implanted primary prevention ICDs during this time period (1703 ICDs; 95% CI, 1520-1886 vs 1495 ICDs; 95% CI, 1249-1741; P <.001).

A limitation of the analysis was the lack of data on other factors that may have likely influenced ICD practices in the studied hospitals.

Related Articles

According to an accompanying editorial by Paul A. Heidenreich, MD, physicians and hospital leaders can avoid future whistleblower-driven investigations “by developing appropriate use criteria and participating in registries tracking appropriate care.”2


  1. Desai NR, Bourdillon PM, Parzynski CS, et al. Association of the US Department of Justice investigation of implantable cardioverter-defibrillators and devices not meeting the Medicare National Coverage Determination, 2007-2015. JAMA. 2018;320(1):63-71.
  2. Heidenreich PA. US Department of Justice Investigations of implantable cardioverter-defibrillators and quality improvement in health care. JAMA. 2018;320(1):40-42.