养老院居住者耐甲氧西林金黄色葡萄球菌定植预测模型的构建

耐甲氧西林金黄色葡萄球菌（MRSA）感染增加了患者的病死率和治疗费用，MRSA定植是其感染的重要原因，研究表明清除MRSA定植可降低MRSA感染的风险。养老院居住者是MRSA易感人群，在该人群中使用主动筛查法确定MRSA定植者费时、费力、费用高昂，且需专业人员和相关实验室条件，在多数养老院无法实施，通过科学的定植预测模型确定定植高危者可显著改善养老院环境的MRSA定植管理。本研究拟采用前瞻性队列研究设计，对1500例养老院居住者构成的训练样本进行连续12个月的观察性研究，以主动筛查结果为金标准，采用Cox回归模型构建养老院居住者MRSA定植预测模型，在验证样本中再进行为期12个月的研究以验证其准确性和预测价值并加以改善。通过建立可靠的MRSA定植预测模型，本研究将对制定养老院居住者的MRSA感染预防和干预方案提供帮助。

Methicillin-resistant Staphylococcus aureus (MRSA) infection increases mortality and will be costly to treat. MRSA colonization is an important source of its infection and studies have shown that eradication of MRSA colonization lowered the risk of infection. Nursing home residents are of population mostly susceptible to MRSA infection, however, active surveillance is hardly performed due to laborious and expensive assays and requirement of highly trained laboratory staffs and special lab facilities. So it will greatly benefit the management of MRSA colonization in nursing homes by identifying high-risk individuals with a validated predication model. We thus propose to conduct a 12-month prospective cohort study on a training set consisting of 1500 nursing home residents using active screening test as gold standard and perform Cox regression analysis to establish a MRSA colonization predication model in this population. The initial model will be examined by performing a study in the next consecutive 12 months in validation set to evaluate its accuracy and predictive value and further improvement will be made accordingly. By having a reliable MRSA colonization predication model, the proposed study will facilitate the development of MRSA infection prevention and intervention strategy in nursing home residents.