|
Cardiovascular outcomes in trials of oral diabetes medications: a systematic review |
Selvin E, Bolen S, Yeh H C, Wiley C, Wilson L M, Marinopoulos S S, Feldman L, Vassy J, Wilson R, Bass E B, Brancati F L |
|
|
CRD summary This review evaluated cardiovascular risk associated with taking oral diabetes medications and concluded that metformin is moderately protective, and rosiglitazone may be harmful, but a lack of power prohibited firmer conclusions. There were some limitations with the available evidence, but the review was generally well-conducted and the conclusions are likely to be reliable. Authors' objectives To evaluate the cardiovascular risks in adults associated with taking oral diabetes medications. Searching MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials were searched from inception to February 2006. Reference lists of included studies were also examined and 15 journals were handsearched from December 2005 to February 2006. Study selection Trials recruiting at least 40 adults with type 2 diabetes, of at least 3 months duration, which assessed FDA (Food and Drug Administration) approved oral diabetic agents and reported a mortality outcome were eligible for inclusion. Combinations of commonly-used therapies were included. Trials evaluating combinations of any three oral diabetes agents, or first-generation sulfonylureas, were excluded.
In the included trials, the mean participant age ranged from 52 to 69 years, and the mean HbA1c (haemoglobin level) at baseline ranged from 6.2% to 10.2%. Medications used included: pioglitazone hydrochloride, metformin hydrochloride, rosiglitazone, glyburide, glimepiride, repaglinide, glipizide, nateglinide, gliclazide, either individually or in combination. Many trials were placebo-controlled, sometimes along with dietary care and/or another oral diabetes agent. Two-thirds of the trials were of less than one year in duration. For most trials cardiovascular outcomes were recorded as adverse events. Most trials recorded all-cause mortality as an outcome.
Two reviewers independently selected studies for inclusion in the review, with disagreements resolved by consensus. Assessment of study quality Study quality was assessed using the Jadad scale (5=high quality and 0=low quality) which assessed randomisation, blinding, and withdrawals and drop-outs. If cardiovascular disease was not a primary endpoint, the quality of adverse event reporting for cardiovascular outcomes was also rated using a 4-point scale based on FDA and the Consolidated Standards of Reporting Trials guidelines for adverse event reporting. This scale assessed adverse events, withdrawals, drop-outs and adverse event definitions.
Two reviewers assessed study quality, although as there was 85% consistency between reviewers, results of only one were reported. Data extraction Mortality rates were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs). For multi-arm trials, dosing arms were combined providing doses were consistent with current clinical practice; trials with more than one comparison group had groups combined when appropriate. Intention-to-treat population data was used.
Two reviewers independently extracted data, with disagreements resolved by consensus. Methods of synthesis For outcome comparisons where there were at least four relatively homogeneous trials, meta-analyses pooling ORs were performed using a fixed-effect model. Statistical heterogeneity was assessed using the I2 statistic (another assessment was also made, but method details were not reported). Sensitivity analyses were conducted to examine the effect of influential studies and different dosing and control group arms. Results of the review Forty randomised controlled trials (RCTs) were included in the review (n=29,734). Sample sizes ranged from 44 to 8,732 and duration of follow up ranged from 3 to 48 months. The mean Jadad quality score was 3. The quality of adverse event reporting (32 trials) was fair, with a mean of 3; 12 trials scored 4 out of 4.
Treatment with metformin hydrochloride was associated with a statistically significant decreased risk of cardiovascular mortality compared with any other medication or placebo, OR 0.74, (95% CI: 0.62, 0.89, 6 trials). For cardiovascular morbidity and all-cause mortality there were similar, but non-significant, results: cardiovascular morbidity OR 0.85, (95% CI: 0.69, 1.05, 7 trials) and all-cause mortality OR 0.81, (95% CI: 0.60, 1.08, 9 trials). No other statistically significant associations were found for any oral agent in any outcome, although rosiglitazone use showed a possible increase in risk of cardiovascular morbidity OR 1.68 (95% CI: 0.92, 3.06, 5 trials). No statistically significant heterogeneity was found. When very large studies were removed, the analyses revealed similar results but with wider CIs. Authors' conclusions Metformin is moderately protective against cardiovascular effects, and rosiglitazone is possibly harmful, but a lack of power prohibits firmer conclusions. CRD commentary The review addressed a clear question and was supported by appropriate inclusion criteria. Attempts to identify relevant studies were undertaken by searching electronic databases, checking references and other methods (although language restrictions were not reported). However, the restriction of using only peer-reviewed papers means the review could be prone to publication bias. Study quality was assessed, but was not used in interpreting the results of the review. Suitable methods were employed to minimise the risks of reviewer error and bias for the processes of study selection and data extraction. Quality assessment was conducted in duplicate, but disagreements were not discussed, so errors and bias could have been introduced. Sufficient study details were provided. The appropriateness of the synthesis was uncertain, as a fixed-effect model was used despite the possibility of clinical heterogeneity. The authors also did not state why they did not pool results when there were fewer than four trials, and they did not report the I2 statistics (although p-values were reported). Although many of the included trials were of a short duration, the authors' conclusions are still likely to be reliable. Implications of the review for practice and research Practice: The authors did not state any implications for practice.
Research: The authors stated that there is a critical need for studies of oral diabetes medications with long-term outcomes, along with clearer reporting of adverse events and withdrawals. Funding Agency for Healthcare Research and Quality, Johns Hopkins Evidence-Based Practice, contract number 290-02-0018; National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases, grant numbers K01DK076595, K24DK62222. Bibliographic details Selvin E, Bolen S, Yeh H C, Wiley C, Wilson L M, Marinopoulos S S, Feldman L, Vassy J, Wilson R, Bass E B, Brancati F L. Cardiovascular outcomes in trials of oral diabetes medications: a systematic review. Archives of Internal Medicine 2008; 168(19): 2070-2080 Other publications of related interest Bolen S, Wilson L, Vassy J, et al. Systematic review: Comparative effectiveness and safety of oral medications for type 2 diabetes mellitus. Ann Intern Med. 2007;147(6):386-99.
Bolen S, Wilson L, Vassy J, et al. Comparative effectiveness and safety of oral diabetes medications for adults with type 2 diabetes. Agency for Healthcare Research and Quality. Comparative Effectiveness Review; 8. 2007. Indexing Status Subject indexing assigned by NLM MeSH Cardiovascular Diseases /etiology /mortality /prevention & Diabetes Complications /mortality /prevention & Humans; Hypoglycemic Agents /therapeutic use; control; control AccessionNumber 12008106508 Date bibliographic record published 01/12/2008 Date abstract record published 15/04/2009 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|