S1 File -

Background
Evidence-based empirical antibiotic prescribing requires knowledge of local antimicrobial resistance patterns. The spectrum of pathogens and their susceptibility strongly influences guidelines for empirical therapies for urinary tract infections (UTI) management.
Objective
This study aimed to determine the prevalence of UTI causative bacteria and their corresponding antibiotic resistance profiles in three counties of Kenya. Such data could be used to determine the optimal empirical therapy.
Methods
In this cross-sectional study, urine samples were collected from patients who presented with symptoms suggestive of UTI in the following healthcare facilities; Kenyatta National Hospital, Kiambu Hospital, Mbagathi, Makueni, Nanyuki, Centre for Microbiology Research, and Mukuru Health Centres. Urine cultures were done on Cystine Lactose Electrolyte Deficient (CLED) to isolate UTI bacterial etiologies, while antibiotic sensitivity testing was done using the Kirby-Bauer disk diffusion using CLSI guidelines and interpretive criteria.
Results
A total of 1,027(54%) uropathogens were isolated from the urine samples of 1898 participants. Staphylococcus spp. and Escherichia coli were the main uropathogens at 37.6% and 30.9%, respectively. The percentage resistance to commonly used drugs for the treatment of UTI were as follows: trimethoprim (64%), sulfamethoxazole (57%), nalidixic acid(57%), ciprofloxacin (27%), amoxicillin-clavulanic acid (5%), and nitrofurantoin (9%) and cefixime (9%). Resistance rates to broad-spectrum antimicrobials, such as ceftazidime, gentamicin, and ceftriaxone, were 15%, 14%, and 11%, respectively. Additionally, the proportion of Multidrug-resistant (MDR) bacteria was 66%.
Conclusion
High resistance rates toward fluoroquinolones, sulfamethoxazole, and trimethoprim were reported. These antibiotics are commonly used drugs as they are inexpensive and readily available. Based on these findings, more robust standardised surveillance is needed to confirm the patterns observed while recognising the potential impact of sampling biases on observed resistance rates.

背景
基于循证依据的经验性抗菌药物处方，需掌握当地的抗菌药物耐药谱。病原体种类及其药敏特性，对尿路感染（urinary tract infection, UTI）诊疗的经验性治疗指南具有显著影响。

研究目的
本研究旨在探明肯尼亚三个郡内尿路感染致病菌株的流行情况及其抗菌药物耐药特征，所得数据可用于确定最优经验性治疗方案。

研究方法
本研究为横断面研究，研究对象为在以下医疗机构就诊的疑似尿路感染患者，采集其尿液样本：肯雅塔国家医院、基安布医院、姆巴加蒂医院、马库埃尼医院、南尤基医院、微生物研究中心以及穆库鲁医疗中心。采用胱氨酸乳糖电解质缺乏（Cystine Lactose Electrolyte Deficient, CLED）培养基进行尿液培养，以分离尿路感染致病菌；抗菌药物敏感性试验采用柯比-鲍尔（Kirby-Bauer）纸片扩散法，并参照临床和实验室标准协会（Clinical and Laboratory Standards Institute, CLSI）的指南与判读标准进行结果判读。

研究结果
本研究共从1898名受试者的尿液样本中分离得到1027株尿路致病菌，分离率为54%。葡萄球菌属（Staphylococcus spp.）和大肠埃希菌（Escherichia coli）为主要致病菌，占比分别为37.6%和30.9%。尿路感染常用治疗药物的耐药率如下：甲氧苄啶64%、磺胺甲噁唑57%、萘啶酸57%、环丙沙星27%、阿莫西林克拉维酸5%、呋喃妥因9%及头孢克肟9%。广谱抗菌药物如头孢他啶、庆大霉素和头孢曲松的耐药率分别为15%、14%和11%。此外，多重耐药（Multidrug-resistant, MDR）菌株占比达66%。

结论
本研究显示氟喹诺酮类、磺胺甲噁唑及甲氧苄啶的耐药率较高。此类药物因价格低廉且易于获取，为临床常用抗菌药物。基于本研究结果，需开展更为规范的强化监测以验证本次观察到的耐药模式，同时应认识到抽样偏倚对观测到的耐药率可能产生的影响。