Debate Continues Over Screening Asymptomatic Diabetics for CAD

May 7, 2007 (Washington, DC) - Using what they call a "back-of-the-envelope" calculation, Drs George Diamond, Sanjay Kaul, and Prediman Shah (Cedars-Sinai Medical Center, Los Angeles, CA) predict that unconditionally treating all asymptomatic diabetics with statins would be cheaper and prevent more cardiovascular events than routinely screening the same subjects for subclinical atherosclerosis and treating only those with a positive test [1].

By their analysis, treating all 14 million asymptomatic diabetics in the US with statins would prevent 84 000 events (30% of the expected total events) at a cost of $10.1 billion, gaining 1 092 000 life-years, for a cost-effectiveness ratio of $9249 per life-year. By contrast, the alternative strategy of screening all 14 million asymptomatic diabetic subjects using myocardial perfusion scintigraphy at an estimated cost of $809 per person would cost $11.3 billion and identify the 20% of subjects among whom 80% of events would occur. If these test-positive patients were then treated with statins, to the tune of an additional $2 billion, the life-years gained would be 873 600 and the cost-effectiveness ratio would be $15 224 per life-year. This number is likely an underestimation of costs, given that a positive perfusion imaging test would also like to lead to further tests, Diamond et al write.

"The belief among test advocates is that we should throw more technology at patients," Diamond argued in an interview with heartwire. "Fine-tuning these selection processes is not going to solve the problem fundamentally, because no test can save a life. Only treatments save lives, and testing excludes treatment."

But in a counterpoint article, Dr George Beller (University of Virginia, Charlottesville, VA) argues that the question is not whether or not diabetics should be universally screened to make decisions about statin treatment because all asymptomatic diabetics should have their LDL cholesterol below 100 mg/dL, and most would require statin therapy to reach this goal [2].

"The issue is really, what do you do to identify those type 2 diabetics who might require even more aggressive medical therapy and even further testing for silent extensive coronary disease?" Beller told heartwire.

In response, Diamond told heartwire, "That would be fine if there were evidence to support an outcomes benefit of aggressive management among diabetics. There isn't."

The trials of trials for screening asymptomatic subjects

Diamond et al also point out that calls for randomized trials, while "well-intentioned," are misguided.

"It would be rather daunting to conduct a clinical trial of these [screening] strategies," they write. "Given the small difference in projected outcome (7.0% for unconditional treatment vs 7.6% for conditional testing over five years), a trial to prove the superiority of the former over the latter would require the randomization of 80 000 subjects followed for five years."

Beller believes a trial is essential and that these hurdles could be overcome by focusing only on high-risk asymptomatic diabetics previously identified through clinical and laboratory criteria. "The problem is that we do not yet know how to select this high-risk asymptomatic group," he acknowledges.

To heartwire, Beller points out that this is not a rationale for abandoning screening in asymptomatic diabetics. Instead, pharmaceutical innovations need to keep pace with screening technology such that information gleaned through screening tests could in fact be paired with the different therapies or different degrees of aggressiveness with therapies.

"This is the population of patients that is going to contribute to an increased prevalence of CAD in the next 20 to 50 years," Beller said. "I do advocate a clinical trial, [and] I really think that if we want to test the feasibility of screening asymptomatic cohorts, I think diabetics would be a good population."

Diamond, however, insists that Beller and others who have argued for a strategy of screening high-risk asymptomatic diabetics need to first answer three "critical" questions: how much does it cost, how much is it going to benefit, and where is the money going to come from?

"I think what these two papers illustrate is the remarkably different manner in which people like me, and people like the test advocates, approach these problems," Diamond said.

The screening debate, Diamond argues, detracts from the simple fact that diabetics are not getting the proven medical therapies they need.

"There should be a global strategy to minimize their risk across the board, and that would include predominantly unrestricted use of proven medications like statins and ACE inhibitors. If we're not going to treat diabetics with all these medications, we better have a real good reason why we're not, and the reason shouldn't be, I gave a test and it wasn't positive."

1. Diamond GA, Kaul S, Shah PK, et al. Screen testing: Cardiovascular prevention in asymptomatic diabetic patients. J Am Coll Cardiol 2007; 49:1915-1917.
2. Beller GA. Noninvasive screening for coronary atherosclerosis and silent ischemia in asymptomatic type diabetic patients: Is it appropriate and cost-effective? J Am Coll Cardiol 2007; 49:1918-1923.

Source: http://www.medscape.com/viewarticle/556175?src=mp