ESTIMATING LIPID MANAGEMENT GUIDELINES' RISK VALUE OF A LIFE YEAR ON TREATMENT

Purpose:  Statins reduce the risk of heart attack and stroke but may have adverse effects. Except for a few surveys there has not been any emphasis on how to quantify these effects to help physicians make treatment decisions. We gauge these adverse effects from a central policy maker's point of view.

Method:  We considered patients who are diagnosed with Type 2 diabetes at age 40. We defined the probabilistic progression of their total cholesterol and high density lipoprotein levels (HDL) by two independent discrete-time Markov chains, and the progression of their triglycerides, systolic blood pressure and HbA1C levels as a function of age. We formulated a dynamic decision model the objective of which is to minimize the occurrence risk of a first major cardiovascular event, i.e., heart attack or stroke. In our model each life year spent on treatment is penalized by a certain factor. The penalty factor represents what the reduction in overall risk of a first major cardiovascular event should be with a year-long use of treatment to make published lipid management guidelines as close as possible to optimal.

Result: Our results demonstrated that the penalty factors ranged from 0.07 % to 0.23 % for males and from 0.04 % to 0.29 % for females. Among all considered guidelines, Adult Treatment Panel (ATP) III* (a modified version of the ATP III guideline specific to diabetes)  had the longest expected treatment durations for both genders: 27 years for males and 32 years for females. It was also the most tolerant guideline to treatment among all. On average, ATP III* was favoring treatment if its annual use reduced the overall risk of a first major cardiovascular event by at least 0.07 % for males and 0.045 % for females.  In terms of reducing the risk of a first major cardiovascular event, almost all guidelines were within 2 % of the optimal performance under their perceived penalty factors. Among all, ATP III* was the closest to being optimal and it was not more than 0.2 % away from the best achievable performance for any gender under their respective penalty factors.

Conclusion: Our analyses showed that guidelines show variation in penalizing a life year on treatment and are close to being optimal under their perceived penalty factors.