A total of 470 of the 502 publically named hospitals that settled with the DOJ were matched to facilities in the ICD registry. Those that settled were found to be larger, private facilities with high case volumes, and were more likely to be found in southern and western regions of the U.S. Two facilities were confirmed to have closed while 16 failed to submit data or did not meet NCDR quality standards.

Prior to the announcement of the DOJ investigation, the proportion of ICD placements not meeting NCD standards was relatively stable and experienced a modest and similar decline rate among all hospitals, according to the interrupted time series evaluation.


Significant declines were found in the proportions of ICDs not meeting the NCD at all hospitals, though the rate was much larger and rapid at ones that moved on to reach settlements. The amount of ICDs not meeting these guidelines was similar among both groups at the end of the study.

Researchers also incorporated non-Medicare patient information to evaluate the impact of the investigation on overall patient population undergoing ICD implants, finding declines in the rates of ICD implants not meeting the NCD for non-Medicare patients at both hospitals that did and did not settle.

Despite there being declines, less than 3.8 percent of ICDs at top performing institutes did not meet NCD criteria, while more than 14.3 percent were in the same category at lower performing ones.

Limiting the study is the fact that not all non-Medicare patients may have been included due to hospitals only being required to submit data on Medicare patients.

The DOJ also did not release a public list identifying all hospitals. In addition, interrupted time series analyses assess change on a single factor, and researchers were unable to determine if the DOJ investigations created barriers in the placement of clinically indicated ICDs.

Average age of ICD patients at hospitals that did and did not settle was 75.1. Of the participants, 71.5 were male and 41.8 percent were diabetic. Heart attacks were incurred by 54.4 percent.

ICDs are considered safe and efficient in the prevention of primary and secondary sudden cardiac arrest. Findings of the study are available in the Journal of the American Medical Association.

Back to HCB News

Cardiology Homepage