OBJECTIVETo identify the potential predictors of restenosis after bare mental stent (BMS) deployment in diabetic patients in Chinese diabetic patients.

METHODSWe retrospectively analyzed all patients implanted with BMS (n = 1126 with 2376 lesions) in our department from 2002 to 2004. The multivariate logistic regression analysis was made to compare the clinical and angiographic characteristics between diabetic patients with and without restenosis. Restenosis was defined as > or = 50% diameter stenosis within the stent and 5 mm in adjacent.

RESULTSThe 6-month follow-up angiograms were available in 889 out of 1126 patients (78.9%) and 151 out of 889 patients (17%) were diabetic patients. Restenosis rate in nondiabetic patients group was 21.2% and 35.9% in diabetic patients (P < 0.001). The predictors of restenosis in diabetics were reference vessel diameter (< or = 3.0 mm), length of lesion (> 15 mm) and insulin use (P < 0.05). The restenosis predicting model showed that reference vessel caliber was the paramount predictor for restenosis in diabetic patients.

CONCLUSIONSRestenosis rate post BMS implantation is significantly higher in diabetic patients compared to non-diabetic patients. Vessel caliber, lesion length and insulin use are predictors of restenosis in diabetic patients. Diabetic patients with reference vessel diameter of > 3.0 mm combined with lesion length < 15 mm and non-diabetic patients with lesion length < 15 mm regardless of the vessel caliber could be treated with BMS since the predicted restenosis rate is lower than 15% in these patients, otherwise DES would be a better choice.