Deeny Sarah R, Worby Colin J, Tosas Auguet Olga, Cooper Ben S, Edgeworth Jonathan, Cookson Barry, Robotham Julie V
Modelling and Economics Unit, Centre for Infectious Disease Surveillance and Control, Public Health England and Health Protection Research Unit in Modelling Methodology, London, UK
Center for Communicable Disease Dynamics, Harvard School of Public Health, Boston, MA, USA.
J Antimicrob Chemother. 2015 Dec;70(12):3366-78. doi: 10.1093/jac/dkv249. Epub 2015 Sep 3.
The objectives of this study were to estimate the relative transmissibility of mupirocin-resistant (MupR) and mupirocin-susceptible (MupS) MRSA strains and evaluate the long-term impact of MupR on MRSA control policies.
Parameters describing MupR and MupS strains were estimated using Markov chain Monte Carlo methods applied to data from two London teaching hospitals. These estimates parameterized a model used to evaluate the long-term impact of MupR on three mupirocin usage policies: 'clinical cases', 'screen and treat' and 'universal'. Strategies were assessed in terms of colonized and infected patient days and scenario and sensitivity analyses were performed.
The transmission probability of a MupS strain was 2.16 (95% CI 1.38-2.94) times that of a MupR strain in the absence of mupirocin usage. The total prevalence of MupR in colonized and infected MRSA patients after 5 years of simulation was 9.1% (95% CI 8.7%-9.6%) with the 'screen and treat' mupirocin policy, increasing to 21.3% (95% CI 20.9%-21.7%) with 'universal' mupirocin use. The prevalence of MupR increased in 50%-75% of simulations with 'universal' usage and >10% of simulations with 'screen and treat' usage in scenarios where MupS had a higher transmission probability than MupR.
Our results provide evidence from a clinical setting of a fitness cost associated with MupR in MRSA strains. This provides a plausible explanation for the low levels of mupirocin resistance seen following 'screen and treat' mupirocin usage. From our simulations, even under conservative estimates of relative transmissibility, we see long-term increases in the prevalence of MupR given 'universal' use.
本研究的目的是估计耐莫匹罗星（MupR）和对莫匹罗星敏感（MupS）的耐甲氧西林金黄色葡萄球菌（MRSA）菌株的相对传播性，并评估MupR对MRSA控制策略的长期影响。
使用马尔可夫链蒙特卡罗方法，根据来自两家伦敦教学医院的数据，估计描述MupR和MupS菌株的参数。这些估计为一个模型设定了参数，该模型用于评估MupR对三种莫匹罗星使用策略的长期影响：“临床病例”、“筛查并治疗”和“普遍使用”。根据定植和感染患者天数对策略进行评估，并进行情景分析和敏感性分析。
在不使用莫匹罗星的情况下，MupS菌株的传播概率是MupR菌株的2.16倍（95%可信区间1.38 - 2.94）。模拟5年后，采用“筛查并治疗”莫匹罗星策略时，定植和感染的MRSA患者中MupR的总患病率为9.1%（95%可信区间8.7% - 9.6%），采用“普遍使用”莫匹罗星策略时，患病率增至21.3%（95%可信区间20.9% - 21.7%）。在MupS传播概率高于MupR的情景中，采用“普遍使用”策略时，50% - 75%的模拟中MupR患病率增加，采用“筛查并治疗”策略时，超过10%的模拟中MupR患病率增加。
我们的结果提供了来自临床环境的证据，表明MRSA菌株中MupR存在适应性成本。这为“筛查并治疗”使用莫匹罗星后所见的低水平莫匹罗星耐药性提供了合理的解释。从我们的模拟结果来看，即使在相对传播性的保守估计下，“普遍使用”莫匹罗星也会导致MupR患病率长期上升。
