Physicians who bill for both technical and professional services with nuclear stress imaging and stress echocardiography order such testing at double the rate of those who do not bill for such services, according to a study published online Nov. 8 in JAMA.
Overall, 12.2% of studied patients had a nuclear stress or echocardiography test within 30 days of a postrevascularization outpatient visit. Cardiologists were more likely than primary care physicians to conduct stress tests, at about 12%, compared with 5%, said Dr. Bimal R. Shah of Duke University, Durham, N.C., and colleagues.
The more physicians stood to receive from imaging, the more often they ordered the tests. Physicians who billed for both professional and technical fees ordered stress or echo tests about 13% of the time, compared to a 9% incidence among those who only billed for professional fees, and 5% for those who billed for neither.
According to Dr. Shah, 80% of the nuclear imaging studies would be considered self-referrals by physicians who billed for both technical and professional fees; 63% were self-referrals by those who billed for professional fees only. Rates were similar for echo tests, with 85% considered self-referrals by physicians who billed for both technical and professional fees, and 67% considered self-referrals by those who billed for professional fees only.
"The association between physician billing status for stress tests and testing frequency persisted after adjusting to the extent possible for patient and physician factors that influence testing," Dr. Shah and his colleagues said.
The authors focused on testing that was ordered following coronary revascularization. According to American College of Cardiology Foundation appropriateness utilization criteria, neither nuclear stress testing nor echocardiography is recommended following percutaneous coronary intervention or coronary artery bypass graft. Thus, stress testing in these cases is likely to be more discretionary based on the criteria.
The researchers analyzed United Healthcare data on enrollees aged 18-64 years during the period of 2004-2007. They excluded testing and outpatient visits during the first 90 days after revascularization, as those could be potentially deemed necessary. To determine how many tests might be self-referrals, the authors looked at the proportion of stress tests for which the testing physician’s tax identification number matched the outpatient visit tax ID; the final study population was 17,847 patients (JAMA 2011;306:1993-2000).
Only 14% had symptoms that were given as the indication for that index outpatient visit; 86% had no billing diagnosis code.
Among the physicians studied, 70% (2,111) of cardiologists billed for both technical and professional fees for nuclear stress imaging studies. In all, 14% (416) billed for professional fees only, and 16% (486) did not bill for either. The proportions were similar for cardiology practices, with 50% billing for both professional and technical services, and 19% for professional fees. A total of 31% did not bill for the services.
Of primary care physicians who conducted both nuclear stress testing and echocardiography, 5% (162) billed for both professional and technical fees for stress testing, and 3% (88) billed both fees for echocardiography. In all, 2% (44) and 1% (28) billed for professional fees only. More than 90% did not bill at all for stress testing or echocardiography.
Additionally, analysis determined that nuclear perfusion imaging in cardiologists’ offices increased 215% from 1998 to 2006, and increased by 181% in other physician offices during the same period, the authors said.
"Discretionary stress testing after revascularization has potential financial and clinical disadvantages for patients, including the costs of the tests, the exposure to ionizing radiation as well as potential downstream costs, and consequences from following up false-positive test results," they wrote.
The authors acknowledged that the study period was before, and contemporaneous to, the publication of the American College of Cardiology Foundation appropriateness criteria. Thus, they could not say whether those guidelines might have had any impact on patterns of testing.
The authors noted that they could not analyze the appropriateness of physicians’ decision-making in ordering tests, since it was based largely on administrative data. But they said that previous studies have shown that using ICD-9 coding, as they did, accurately reflects diagnoses.
But this absence of data "prevents understanding physician intent for performing the imaging study," Dr. Brent K. Hollenbeck and Dr. Brahmajee K. Nallamothu of the University of Michigan, Ann Arbor, said in an accompanying editorial (JAMA 2011;306:2028-30).
They also note that the data was observational, and that it is possible that the study could be founded by selective referral, for example, that the patients who needed imaging were referred to physicians who specialized in such testing.