主页 > 医学前沿 >
【medical-news】临床试验中降压药对糖尿病发病率
DOI:10.1016/S0140-6736(07)60697-7
TITLE: Incident diabetes in clinical trials of antihypertensive drugs
AUHOR: Simon KH Lam a and Andrew Owen a
The network meta-analysis by William Elliott and Peter Meyer (Jan 20, p 201)1 found that use of an angiotensin-receptor blocker (ARB ) is associated with a lower incidence of diabetes, and use of a diuretic is associated with a higher incidence, when compared with placebo. The use of β blockers, angiotensin-converting enzyme (ACE) inhibitors, and calcium-channel blockers did not result in an incidence of diabetes that was different from that associated with placebo.
One potential weakness of this network model is in the appropriate modelling of random effects for trials with multiple treatment groups where a different random effect for each comparison is desired.2 Elliott and Meyer have assumed that all comparisons from the same trial involved the same realisation of the random effects of the treatment comparison.3 This is not ideally suited for meta-analysis with large numbers of multigroup trials.4 Have they considered an extension to their approach to handle this weakness?
We have reanalysed the data using a full Bayesian network approach2,5 with appropriate modelling of random effects for multigroup trials (figure). The results show identical rank order to the original analysis. The conclusion regarding ARB and diuretics is similar, but use of β blockers is now associated with a higher incidence of diabetes than placebo. The reduction in incidence of diabetes after use of ACE inhibitors did not reach significance.
Figure. Results of full Bayesian network meta-analysis of 22 trials with 143153 patients
Analysis was implemented using WinBUGS. Non-informative prior distributions were used, and posterior distributions of treatment effects were constructed from three chains of 30000 simulations. Placebo was used as referent agent. 本人已认领该文编译，48小时后若未提交译文，请其他战友自由认领。 RESOURCE: The Lancet 2007; 369:1513-1514
DOI:10.1016/S0140-6736(07)60697-7
TITLE: Incident diabetes in clinical trials of antihypertensive drugs
临床试验中降压药对糖尿病发病率的影响：ARBs降低、β受体阻滞剂和利尿剂增加糖尿病发病率
AUHOR: Simon KH Lam a and Andrew Owen a
The network meta-analysis by William Elliott and Peter Meyer (Jan 20, p 201)1 found that use of an angiotensin-receptor blocker (ARB ) is associated with a lower incidence of diabetes, and use of a diuretic is associated with a higher incidence, when compared with placebo. The use of β blockers, angiotensin-converting enzyme (ACE) inhibitors, and calcium-channel blockers did not result in an incidence of diabetes that was different from that associated with placebo.
William Elliott 和Peter Meyer通过网络meta分析发现：和安慰剂组相比，应用血管紧张素受体阻滞剂伴随着糖尿病更低的发生率，应用利尿剂发生糖尿病的几率更高。应用β受体阻滞剂，血管紧张素转化酶抑制剂和钙通道阻滞剂后，和安慰剂组相比，糖尿病的发生率没有不同。
One potential weakness of this network model is in the appropriate modelling of random effects for trials with multiple treatment groups where a different random effect for each comparison is desired
...not...that was different from that associated with placebo.和安慰剂组相比，糖尿病的发生率没有不同阅读本文的人还阅读：
作者:admin@医学,生命科学 2010-10-04 17:11
医学,生命科学网