Tuesday, September 25, 2007

Three for the Road

1 - Eulogy for a Quality Measure

Dr. Thomas Lee writes in the New England Journal of Medicine:

On May 8, 2007, one of the best-known quality measures in health care was put to rest. The percentage of patients with acute myocardial infarction who receive a prescription for beta-blockers within 7 days of hospital discharge has been used to evaluate U.S. managed care plans since 1996. This measure will no longer be reported by the National Committee for Quality Assurance (NCQA) because it is simply no longer needed — a development that offers encouragement and important lessons.

The data in the graph show why the NCQA Committee on Performance Measurement voted unanimously to retire the beta-blocker measure. A . . .

Sorry, extract only without paying a fee -- but here are the next couple of sentences:

. . . decade ago, only two thirds of US patients who survived acute myocardial infarctions recieved beta-blockers; today, nearly all do. As the curve representing the 10th percentile crept above 90%, the NCQA found little variation among health plans. At least when it comes to this intervention, the U.S. health care system has become reliable.

2 - Mashup Request

Bob Coffield's excellent Health Care Law Blog had an interesting piece this past weekend citing NetDoc's mashup of HHS hospital data -- heart attacks, heart failure, pneumonia, surgical infection prevention -- with Google maps. Note the following thoughtful comment from Bonnie on this entry:

This is a very smart idea! Next, I want to see a mashup that shows hospitals within a certain region that exhibit the best infection control rates.

Well, maybe not Bonnie! After all, the site itself warns:

Important: This tool should not be used to make medical decisions - check the original data source (HHS Hospital Search) and discuss hospital options with your physician to select the best hospital for you. Neither hospital locations nor the accuracy of the rankings/data shown is guaranteed, and there may be errors and/or ommissions (sic).

3 - Does Disclosure Hurt Minorities?

Finally, a 2004 article in Circulation entitled "Racial Profiling,The Unintended Consequences of Coronary Artery Bypass Graft Report Cards," by Rachel M. Werner, MD, PhD; David A. Asch, MD, MBA; Daniel Polsky, PhD suggests:

Although public release of quality information through report cards is intended to improve health care, there may be unintended consequences of report cards, such as physicians avoiding high-risk patients to improve their ratings. If physicians believe that racial and ethnic minorities are at higher risk for poor outcomes, report cards could worsen existing racial and ethnic disparities in health care.

A similar conclusion was reiterated by a couple of the authors the next year:

Public reporting of quality information promotes a spirit of openness that may be valuable for enhancing trust of the health professions, but its ability to improve health remains undemonstrated, and public reporting may inadvertently reduce, rather than improve, quality. Given these limitations, it may be necessary to reassess the role of public quality reporting in quality improvement.

As arguments about disclosure start to be more vigorous here in Massachusetts, the first article is finding its way around the halls of government. Does anyone know if these same authors have updated their findings since 2005, or if there have been further articles by others on this subject?


Anonymous said...


I don't know the answer to your question, but you could search the NIH Library of Medicine database (Pubmed) here:


I gave it a quick look with negative results, but have to run off somewhere right now.

Anonymous said...

Oops, revised comment! Enter Rachel m werner into the Pubmed site search box I mentioned above, and you get a whole bunch of stuff. That should help you.

Anonymous said...

Would it be Easily Possible for a transparent Hospital system in Boston to evade? if a High Risk Destitute stumbeled in and gasped! I think that it would be difficult to get him off the record and equally difficult to let him leave unattended. Even if he is uncovered under any healthcare plan & regardless of his minority status he is still on the record of somebody in the hospital. Is it this way?

Anonymous said...

The study by Werner et al published in the March 15, 2005 issue of Circulation (2005;111:1257-63) is consistent with other literature which it cited about there being unintended negative consequences of report cards. However, the Werner study defined growing racial disparities in cardiac care by an increase in the difference between races in CABG for AMI usage rates, rather than, as is more usual, an increase in the ratio between races in the CABG usage rates. For example, from the period in New York before the report card to after the report card (see table 2 in Werner et al) the use of CABG for AMI in Whites grew from 3.6% to 8.0%, while that in blacks grew from 0.9% to 3.0%. Thus the NY difference in rates between whites and blacks grew from 2.7% to 5.0%, but the W/B ratio decreased from 4.0 to 2.7. In the same period in comparison states, the CABG for AMI usage in whites increased from 5.9% to 8.8%, while that in blacks increased from 2.5% to 5.2%. The difference between whites and blacks in CABG for AMI grew from 3.4% to 3.7%, while the W/B ratio decreased from 2.4 to 1.7. The “ratio of ratios,” that being the W/B ratio in NY compared to the W/B ratio in comparison states, was 1.7 before the cardiac report card and 1.6 after. I am not aware that the authors have addressed this criticism. This makes the New York CABG for AMI evidence for growing disparities, for me, unconvincing.

Jon Olson, DPM, DrPH, Epidemiologist, Connecticut Department of Public Health, Hartford.

P.S. I enjoy your blog, I agree that there are often racial differences in the delivery of health care in America, and I agree that there are often unintended negative consequences of cardiac report cards.

Unknown said...

Disclosure of hospital quality and safety statistics is of limited value to this consumer. I pick a physician first and then s/he gives me a choice of hospitals. Disclosure about complications, infections, costs, length of stay, and other relevant indicators about the PHYSICIAN'S practice would be more meaningful.