作者: Reint Meursinge Reynders
关键词: Funnel plot 、 Clinical trial 、 Cochrane Library 、 Data mining 、 Confidence interval 、 MEDLINE 、 Medicine 、 Dentistry 、 Prospective cohort study 、 Implant 、 Publication bias 、 General Dentistry
摘要: Medline, Embase, LILACS, Cochrane Library, Google Scholar, Web of Science, African Journals Online, Digital Dissertations. Randomised controlled trials (RCTs), prospective clinical trials(CCTs) and cohort studies were included. Studies on implants with a diameter greater than 2 mm excluded. Data extracted independently by two authors. Bias in RCTs was assessed using the risk bias tool, non-randomised being Newcastle-Ottawa Scale. Failures mini-screw expressed as event rates 95% confidence intervals. Risk factors random effects model sub-group meta-regression analyses also conducted. Fifty-two studies; five RCTs, eight CCTs, twenty-seven twelve unclear designs that to be A funnel plot suggested there may publication bias. Analysis 4987 miniscrew used 2281 patients indicated an overall failure rate 13.5% (95% interval, 11.5%-15.8%). not associated patient sex or age implant insertion side, whereas they significantly jaw insertion. Some trends identified through exploratory analysis but no definite conclusions could drawn. Orthodontic have modest small mean rate, indicating their usefulness practice. Although many seem affect rates, majority them still need additional evidence support any possible associations.