TY  - JOUR
T1  - COnsidering selection bias when developing a search strategy
AU  - Nakao YM, Ueshima K, Teramukai S, et al
Y1  - 2011/03/14
N1  - 10.1001/archinternmed.2011.45
JO  - Archives of Internal Medicine
SP  - 471
EP  - 472
VL  - 171
IS  - 5
N2  - 
The Candesartan Antihypertensive Survival Evaluation in Japan (CASE-J) trial, published in 2008,2 assessed cardiovascular outcomes in high-risk hypertensive patients receiving either candesartan or amlodipine. The primary end point of the CASE-J trial was a composite of cardiovascular morbidity and mortality, including heart failure. We think that this trial meets the inclusion criteria in the meta-analysis by Sciarretta et al,1 but it was not included despite a careful search using 2 databases by 2 investigators. Also, neither the KYOTO HEART study3 nor the Heart Institute of Japan Candesartan Randomized Trial for Evaluation in Coronary Artery Disease (HIJ-CREATE)4 was included. The authors also checked in the references of a previous meta-analysis by Verdecchia et al.5 This could not be enough to identify all randomized controlled trials to evaluate a wide range of antihypertensive drugs because this previous meta-analysis by Verdecchia et al5 aimed to compare old antihypertensive drugs (diuretics and β-blockers) or placebo with new drugs (renin-angiotensin system inhibitors or calcium channel blockers). We are afraid that other important clinical trials are not included in the meta-analysis by Sciarretta et al.1

SN  - 0003-9926
M3  - doi: 10.1001/archinternmed.2011.45
UR  - http://dx.doi.org/10.1001/archinternmed.2011.45
ER  -