George Nisha A, Pan Daniel, Silva Luisa, Baggaley Rebecca F, Irizar Patricia, Divall Pip, Al-Oraibi Amani, Khan Durdana P, Martin Christopher A, Nazareth Joshua, Collins Ian, Chew See Lek, Nellums Laura B, Pareek Manish
Development Centre for Population Health, University of Leicester, Leicester, United Kingdom.
Division of Public Health and Epidemiology, School of Medicine, University of Leicester, Leicester, United Kingdom.
EClinicalMedicine. 2025 Aug 21;87:103384. doi: 10.1016/j.eclinm.2025.103384. eCollection 2025 Sep.
Bacterial antimicrobial resistance (AMR) is a leading cause of death globally. However, there has been no data synthesis on whether it influences mortality within hospital settings. We conducted a systematic review and meta-analysis to quantify the prevalence and risk of mortality associated in hospitalised patients with AMR, compared to patients with infections not classified as AMR.
Databases (MEDLINE, EMBASE, and Cochrane library) were searched from inception up to 14th April 2025 for studies that reported the prevalence of AMR in patients who acquired infections in hospitals and mortality (PROSPERO CRD42023420609). We calculated pooled prevalence estimates of AMR as well as unadjusted and adjusted estimates of the effect of AMR on mortality using a random-effects model. Study quality was assessed using the Joanna Briggs Quality Appraisal Tool, risk of bias using DOI plots and LFK index and certainty of evidence of mortality using GRADE criteria.
We identified 34 studies (20,658 patients with resistant organisms) from 18 countries-namely the USA, China, the UK, Canada, Israel, Japan, Malaysia, Korea, Brazil, and Singapore. Of these, 33 were observational studies whilst two studies (one observational study and one purely modelling study) mechanistically modelled risk of mortality in relation to transmission. No studies were conducted in the African subcontinent, the Middle-East, Russia, and India. The prevalence of AMR was high in patients in hospital (pooled prevalence: 36.5%, 95% CI: 29%-44%, = 99%) and associated with higher mortality (unadjusted pooled risk ratio [RR]: 1.64, 95% CI: 1.37-1.97, = 96.22%, τ = 0.20; adjusted pooled RR: 1.58, 95% CI: 1.33-1.87, = 85.9%, τ = 0.13) compared to non-AMR organisms.Sensitivity analyses showed particularly elevated risks for in-hospital mortality and for AMR-associated bacteraemia. Study quality was generally rated to be high, but there was evidence of publication bias in estimates of both prevalence and mortality. Overall certainty of evidence of mortality was graded to be low.
AMR is highly prevalent within hospital settings globally and associated with increased in-hospital mortality. Crucially, no data was identified from the India subcontinent, African subcontinent, the Middle East, and Russia, and only two studies used mechanistic modelling to explore how transmission of AMR affects mortality. Further research is required, particularly in underrepresented regions to inform interventions aimed at reducing both AMR transmission and its related mortality within hospital settings.
Pacific Life.
细菌抗菌药物耐药性（AMR）是全球主要的死亡原因。然而，尚无关于其是否影响医院环境中死亡率的数据综合分析。我们进行了一项系统评价和荟萃分析，以量化与未分类为AMR的感染患者相比，住院的AMR患者的患病率和死亡风险。
检索数据库（MEDLINE、EMBASE和Cochrane图书馆），从数据库建立至2025年4月14日，查找报告医院获得性感染患者中AMR患病率和死亡率的研究（PROSPERO CRD42023420609）。我们使用随机效应模型计算AMR的合并患病率估计值以及AMR对死亡率影响的未调整和调整估计值。使用乔安娜·布里格斯质量评估工具评估研究质量，使用DOI图和LFK指数评估偏倚风险，使用GRADE标准评估死亡率证据的确定性。
我们从18个国家（即美国、中国、英国、加拿大、以色列、日本、马来西亚、韩国、巴西和新加坡）确定了34项研究（20,658例耐药菌患者）。其中，33项为观察性研究，两项研究（一项观察性研究和一项纯建模研究）从机制上模拟了与传播相关的死亡风险。非洲次大陆、中东、俄罗斯和印度未开展相关研究。医院患者中AMR的患病率很高（合并患病率：36.5%，95%CI：29%-44%，I² = 99%），且与较高的死亡率相关（未调整的合并风险比[RR]：1.64，95%CI：1.37-1.97，I² = 96.22%，τ = 0.20；调整后的合并RR：1.58，95%CI：1.33-1.87，I² = 85.9%，τ = 0.13），与非AMR微生物相比。敏感性分析显示，医院内死亡率和AMR相关菌血症的风险尤其升高。研究质量总体评定为高，但在患病率和死亡率估计中均有发表偏倚的证据。死亡率证据的总体确定性等级为低。
AMR在全球医院环境中高度流行，并与医院内死亡率增加相关。至关重要的是，未从印度次大陆、非洲次大陆、中东和俄罗斯获取数据，仅有两项研究使用机制建模来探讨AMR传播如何影响死亡率。需要进一步研究，特别是在代表性不足的地区，为旨在减少医院环境中AMR传播及其相关死亡率的干预措施提供信息。
太平洋人寿保险公司。
