Comparison of effects of thiazolidinediones on cardiovascular risk factors: observations from a clinical practice.

OBJECTIVE: To compare short-term glycosylated hemoglobin (HbA(1c)), lipid, weight, tolerability, and hepatic effects after switching patients with type 2 diabetes from troglitazone to either pioglitazone or rosiglitazone treatment. METHODS: This study compared the effects of conversion from maintenance troglitazone therapy to pioglitazone versus rosiglitazone. HbA(1c), lipids, weights, adverse effects, and hepatic status were monitored, providing no other major therapeutic change had been made. Of 163 study candidates, 144 and 125 patients fulfilled the criteria for comparison of HbA(1c) and lipids, respectively. RESULTS: HbA(1c) decreased an absolute mean of 0.08% for each treatment group, after a mean 3.2-month observation. Mean cholesterol, triglyceride, and low-density lipoprotein (LDL) cholesterol levels decreased in the pioglitazone group by 4.7%, 11.3%, and 7.3% but increased 8.4%, 38.4%, and 8.1%, respectively, in the rosiglitazone group. Mean high-density lipoprotein (HDL) increased 2.6% with pioglitazone and decreased 6.3% with rosiglitazone therapy. Patients receiving a statin concomitantly when switched to rosiglitazone treatment had a 51.9% mean triglyceride increase versus a 25.7% increase for those not taking a statin, whereas the patients switched to pioglitazone therapy had respective decreases of 14.2% and 6.2%. Both drugs were generally well tolerated; patients in both groups had similar slight weight increases and no hepatic dysfunction. CONCLUSION: Patients switched from maintenance troglitazone treatment to either pioglitazone or rosiglitazone therapy had similar glycemic control. Conversion to pioglitazone therapy caused a trend toward improvement in all lipid variables, but switching to rosiglitazone therapy caused significantly increased levels of cholesterol, triglycerides, and LDL and a trend toward decreased HDL. Patients already receiving statins when switched to rosiglitazone therapy had particularly notable triglyceride worsening. Whether these effects will lead to changes in cardiovascular outcome or will be maintained over a longer period remains to be established.

Benefits of computerized outcome analysis in diabetes management.

OBJECTIVE: To observe the effects of a computerized relational database designed to measure outcomes and enhance adherence to recommended diabetes management guidelines in a specialty private practice setting. METHODS: DiaTrends, a software program specifically designed to monitor the processes and outcomes of diabetes care, was introduced in a private endocrinology practice in 1996. Information on all patients with diabetes (N = 1,096) was recorded in the database. Reports ("queries") were developed to monitor the adherence to recommended guidelines and to assess the outcomes of care for patients with diabetes treated with glucose-lowering medications. Glycosylated hemoglobin (HbA(1c)) and lipid control were selected from the many outcomes and processes of care monitored as the basis of this report. RESULTS: Data were collected on 1,096 patients between 1996 and 1998. We were able to monitor adherence to published diabetes guidelines and to document the outcomes. Parallel trends toward improvement were noted in both HbA(1c) and lipid variables. The average HbA(1c) achieved for the entire patient population was 7.3%. CONCLUSION: Using a computerized database to monitor glycemic control and lipid management is effective in evaluating the outcomes of care and provides a focused approach that enhances diabetes care. These improvements in metabolic outcomes likely reflect a change in the practice patterns of the physicians as a consequence of using this computerized database system.