CHICAGO – Statin therapy in individuals who had a low risk of cardiovascular disease was not cost effective when the therapy’s potential to increase the risk of diabetes was taken into account, according to an analysis by researchers at the Centers for Disease Control and Prevention.
Statins were cost effective among patients with a high risk of cardiovascular disease (CVD), but the therapy’s cost effectiveness varied among patients with medium and low CVD risk who were at different levels of diabetes risk, the study investigators found.
"We think that physicians should check patients’ blood glucose levels when prescribing statins for preventing CVD among persons with low or medium CVD risk," explained Xiaohui Zhou, Ph.D., a health economist at the CDC and the lead author of the study.
Dr. Xiaohui Zhou
A recent meta-analysis of existing data suggests that while statins reduce the risk of cardiovascular disease by as much as 38%, they could also increase the risk of diabetes by 8%-25%, said Ping Zhang, Ph.D., a senior health economist at the CDC, who presented the unpublished abstract at the annual meeting of the American Diabetes Association.
To evaluate the cost-effectiveness of statins for the prevention of CVD while accounting for the elevated risk of diabetes, Dr. Zhang and colleagues used a simulation model and assessed the 30-year health outcomes of a low-cost statin therapy among patients at different risk levels for CVD and diabetes. The study population included a combined sample of nondiabetic participants in five large-scale trials: ASCOT-LLA, JUPITER, WOSCOPS, MEGA, and AFCAPS/TexCAPS.
The primary outcomes were the incidences of diabetes and CVD, quality-adjusted life-years (QALYs), and cost per QALY.
The investigators stratified the analysis according to baseline risk, categorizing the CVD risk into low (5-year risk less than 5%), medium (5%-10%), and high risk (more than 10%). The diabetes risk was categorized by normal glucose tolerance, impaired glucose tolerance (IGT), and IGT plus impaired fasting glucose.
When measuring the effect of statin therapy, the researchers made two assumptions: a constant relative CVD risk reduction from statin use across the three levels of baseline CVD risk, and a constant relative diabetes risk increase from statin use across the three levels of baseline diabetes risk.
Statin use reduced CVD events, but it increased diabetes events over a period of 30 years to varying degrees, based on the diabetes risk level.
For instance, for individuals with a 5-year CVD risk greater than 10% and normal glucose tolerance, the risk of diabetes increased by nearly 4%. In those with impaired glucose tolerance, that risk increased by roughly 5%, and in patients with IGT and impaired fasting glucose, the risk rose to more than 6%.
In addition, although the cost-effectiveness of statin use was largely dependent on CVD risk, it was also affected by diabetes risk.
Statin use was not cost effective when the 5-year CVD risk was less than 5%: $101,800/QALY. In patients with a 5-year CVD risk of 5%-10%, the cost-effectiveness of statins varied based on baseline diabetes risk: $8,900-$16,400/QALY for normal glucose tolerance, $16,300-$73,300/QALY for impaired glucose tolerance, and $326,700/QALY for IGT and impaired fasting glucose.
However, in individuals who had a 5-year CVD risk of greater than 10%, statin use was cost effective regardless of the diabetes risk, the researchers found: $4,500-$6,300 cost/QALY for normal glucose tolerance, $6,500-$11,800/QALY for impaired glucose tolerance, and $14,400-$45,800/QALY for IGT and impaired fasting glucose.
So, does the increase in diabetes risk matter when statins are used?
It depends, Dr. Zhang said.
For patients at high risk of CVD, it doesn’t matter, Dr. Zhang noted, and statin use is cost effective across all three levels of diabetes risk. For patients at medium risk of CVD, statin use is cost effective for those who have normal glucose tolerance or impaired glucose tolerance. However, it’s not cost effective for those who have IGT plus impaired fasting glucose.
The increased risk of diabetes does matter in individuals at a low CVD risk, as statin use in those patients was no longer cost effective.
The study had some limitations, Dr. Zhang cautioned. The model did not capture all the possible beneficial or harmful effects of statin use. Also, the results apply to the trial population only, who were mostly middle aged or older, and the authors assumed homogeneity of statin effect in the study population at various baseline risks.
The findings were limited to the population used in the analysis, and the true long-term clinical benefits of statin therapy still remain largely unknown, the authors noted.
Dr. Zhou and Dr. Zhang said they had no relevant financial disclosures.