The Impact of Outcomes Reporting on Access to Health Care of High-Risk
Patients to Interventional Cardiologists in the United States

	8/10/99 - Pressure is being placed on medical sub-specialists to
lower procedure costs while improving outcomes. Outcomes data reporting has
been used to try to improve procedural results; however, some negative
effects of this reporting are being seen.
	We surveyed 5,229 interventional cardiologists practicing in the
USA about the effect of outcomes reporting on their approach to high-risk
patients who needed interventional procedures. The results were analyzed one
month after mailing the survey.
	1,444 cardiologists responded. 93% of them said that outcomes
reporting would have some effect on their approach to high-risk patients.
Only 7% said outcomes reporting would have no effect on their willingness to
do interventions on high-risk patients. The majority of the respondents did
51-100 interventions per year. Those who did more than 500 interventions per
year said they would be less affected if outcomes reported from their results.
	Conclusions - Our survey indicates that outcomes reporting would have
a significant effect on the willingness of cardiologists to do procedures on
high-risk patients. This might limit access to needed care for the high-risk
population.

Long version:

	With increasing pressure to lower costs and improve outcomes for
procedures done in the USA, outcomes reporting has been increasingly used. In
New York state, risk-adjusted mortality data for bypass surgery has been
published in the newspaper for every heart surgeon in the state. The impact
of this reporting is unknown. In a recent report, Omoigui found that the
public reporting of outcomes data may be associated with increased referral
of high-risk patients from New York to out of state medical centers.
	As HMOs try to get better outcomes at lower costs, they may use
individual physician outcomes data to select their group members. So doctors
may not be want to do procedures on high-risk patients so they can keep good
outcomes statistics. If this happens, high-risk patients might have limited
access to the procedures that would benefit them the most.
	There is little data on how outcomes reporting affects the willingness
of interventional cardiologists to do procedures on high-risk patients. We
took the theory that fear of outcomes reporting has already limited high-risk
patients' access to needed procedures. To study the impact of such reporting,
we sent an anonymous survey along with a letter of explanation to all
interventional cardiologists in the United States.

Survey

	We sent surveys to 5,229 interventional cardiologists in the USA.
These doctors were actively practicing. The questions were carefully phrased
with the help of a psychology department to get useful information using non-
offensive questions. We asked about their practice as well as the volume of
reperfusion procedures they did each year. We also asked how outcomes data
reporting would affect their willingness to perform interventional procedures
on high-risk patients.
	The survey was mailed in December of 1996. Respondents were allowed
to stay anonymous. After one month, responses were analyzed. 1,444
cardiologists (28%) responded to the survey. The surveys sent to HMO
physicians yielded the fewest responses! Most respondents did 101-500
procedures a year. 85% knew their own outcomes data. Solo practitioners did
less procedures than other groups. Of all respondents, only 26% said that
outcomes reporting had little or no effect on their willingness to procede
with high-risk patients. 85% said that if outcomes were reported, they would
be somewhat or much less likely to perform interventions on high-risk
patients.
	Of all respondents, only 97 (7%) answered that outcomes reporting
would have "no effect" on their decision to do procedures on high-risk
patients. 1,345 respondents (93%) answered that outcomes reporting would have
some effect on their decision. Interventional cardiologists with very low
volume and those with very large volume were less likely to say that outcomes
reporting would have "any effect" on their approach to these patients.
Physicians in a teaching hospital had fewer doctors who were "much less
likely" to perform interventions. This group also had the highest volume.

Discussion

	Our results suggest that outcomes reporting would limit the access of
high-risk patients to interventional cardiology procedures in the USA. Since
outcomes data for bypass surgery was first released in 1986, there has been
debate about its use. At first, healthcare consumers, employers, and
administrators felt that reporting it would improve quality of care.
	The Consumer Guide to Coronary Artery Bypass Graft Surgery lists
annual risk-adjusted mortality rates for all hospitals, and which surgeons do
such procedures in Pennsylvania. One study surveyed cardiologists and heart
surgeons to test the influence of this guide. Most respondents said the guide
had inadequate risk adjustment and unreliable data. Most cardiologists
reported increased difficulty finding surgeons willing to do bypass surgery
on severely ill patients who required it, and most heart surgeons responding
to the survey said they were less willing to operate on such patients.
	Preliminary data showed that risk-adjusted death rates declined in
New York after the New York Department of Health started the Cardiac Surgery
Reporting System. However, further review shows that the improvement may have
been due to increased reporting of risk factors by the surgeons rather than
an improvement in care.
	It has also been reported that many high-risk patients from New York
have been referred to out of state centers because of the publishing of
clinical outcomes. In the city of New York, some surgeons refused to do
high-risk bypass on the father of a journalist after the public release of
outcomes data was introduced in newspapers.
	High-risk patients are often those for whom bypass surgery can result
in the greatest benefit. Likewise, high-risk patients needing angioplasty are
often those refused for surgery because of their high operative risk. These
patients, who have large amounts of ischemia and reduced left ventricular
function, have the most to gain by having a successful angioplasty. With the
use of stents and improvements in angioplasty equipment, cardiologists are
better able to unblock difficult arteries. However, the ability to better help
more high-risk patients also results in higher complication rates. Outcomes
reporting may make cardiologists unwilling to do procedures on these patients.
	Only 7% of cardiologists reported that outcomes reporting would have
no effect on their treatment decisions. On the other hand, our survey showed
that outcomes reporting was less likely to affect the practice of
cardiologists in high-volume practices. High-volume operators may be more
willing to accept high-risk patients. They may be able to give better care
since higher volume is associated with lower complication rates.
	The collection of data is time consuming and expensive. The risk-
adjusted model is very complex and has many variables. It is difficult to
include all risk factors and weigh their importance to the overall model.
Also, mortality may not be the best endpoint. Other endpoints, such as
functional capacity, feeling of wellness, heart attack, length of hospital
stay, and overall cost may be important outcomes.
	An important finding was that outcomes reporting would not affect the
practice of high-volume operators as much as average or low-volume operators.
This underscores one of the major goals of reporting outcomes, which is to
encourage more experienced, high-volume operators to accept the more difficult
cases, and discourage the lower volume operators from attempting them.
	
Limitations

	The main limitation of our study was that only 28% of those surveyed
responded. Other studies have sent reminders, resulting in higher response
rates. We preferred to allow the survey to remain anonymous since some of the
questions were about avoiding high-risk patients.
	There may be some bias because of the limited response in that those
doctors already aware of their own outcomes and whose outcomes are favorable
might be more willing to respond than those who do not have such high success
rates.

J Invas Cardiol 11(3):111-115, 1999