OBJECTIVE:To establish a new semi-automatic drug dispensing mode,with proper cost,which falls in between full-automatic drug dispensing mode and manual drug dispensing mode,good work efficiency,standard and simple operation meth-od and meets national laws and regulations. METHODS:A semi-automatic drug dispensing system was designed,in which the in-formation in the drug dispensing sheet could be automatically printed on the drug bag,and automatic bagging,packaging and deliv-ery of drugs were realized. Such drug dispensing system included hardware(mechanical structures such as drug turntable and drug funnel,transmission device,etc.)and software control systems(the program of interface with hospital information system,micro control unit software,computer software,etc.). Through commissioning,formal operation and statistics,based on 18 oral drug dis-pensing sheets with the same contents,calculated the time of drug dispensing and the number of drug dispensing errors by 3 phar-macists respectively in manual drug bag dispensing mode and semi-automatic drug dispensing mode,to evaluate the effect of the semi-automatic drug dispensing system. RESULTS:From commissioning in May 2012 to formal operation in September 2012,the system operated normally and utility model patents were obtained. In the two modes,the total time of drug dispensing was 481 and 397 min (t=6.82,P<0.001),the numbers of drug dispensing errors were 25 and 7 (χ2=9.353 8,P=0.002 2),respectively. There was statistical significance. CONCLUSIONS:The semi-automatic drug dispensing system has higher efficiency and less num-ber of drug dispensing errors compared with manual drug bag dispensing mode and lower cost compared to full-automatic drug dis-pensing system. It deserves promotion.

BACKGROUND: The predictive value of hemoglobin A1c (HbA1c) levels in non-diabetic patients with myocardial infarction undergoing percutaneous coronary intervention (PCI) is still controversial. This study aimed to evaluate whether HbA1c levels were independently associated with adverse clinical outcomes in non-diabetic patients with coronary artery disease (CAD) who had undergone PCI by performing a meta-analysis of cohort studies. METHODS: This meta-analysis included non-diabetic patients with CAD who had undergone PCI. A systematic search for publications listed in the PubMed, Embase, and Cochrane Library databases from commencement to December 2018 was conducted. Studies evaluating the adverse clinical outcomes according to abnormal HbA1c levels in non-diabetic patients diagnosed with CAD who had undergone PCI were eligible. The primary outcomes were long-term all-cause deaths and long-term major adverse cardiac events, and the secondary outcome was short-term all-cause deaths. The meta-analysis was conducted with RevMan 5.3 and Stata software 14.0. Odds ratios (ORs) were pooled using a random or fixed-effects model, depending on the heterogeneity of the included studies. Sub-group analysis or sensitivity analysis was conducted to explore potential sources of heterogeneity, when necessary. RESULTS: Six prospective cohort studies involving 10,721 patients met the inclusion criteria. From the pooled analysis, abnormal HbA1c levels were associated with increased risk for long-term all-cause death (OR 1.39, 95% confidence interval [CI] 1.16-1.68, P = 0.001, I = 45%). Sub-group analysis suggested that abnormal HbA1c levels between 6.0% and 6.5% predicted higher long-term major adverse cardiac event (including all-cause deaths, non-fatal myocardial infarction, target lesion revascularization, target vessel revascularization, recurrent acute myocardial infarction, heart failure requiring hospitalization, and stent thrombosis) risk (OR 2.05, 95% CI 1.46-2.87, P < 0.001, I = 0). Contrarily, elevated HbA1c levels were not associated with increased risk of short-term all-cause death (OR 1.16, 95% CI 0.88-1.54, P = 0.300, I = 0). CONCLUSIONS An abnormal HbA1c level is an independent risk factor for long-term adverse clinical events in non-diabetic patients with CAD after PCI. Strict control of HbA1c levels may improve patient survival. Further studies in different countries and prospective cohort studies with a large sample size are required to verify the association.