Department of Medical and Surgical Sciences, Alma Mater Studiorum, University of Bologna, Via Irnerio 48, 40126, Bologna, Italy.
Department of Medical and Surgical Sciences, Alma Mater Studiorum, University of Bologna, Via Irnerio 48, 40126, Bologna, Italy.
Eur J Intern Med. 2020 May;75:60-70. doi: 10.1016/j.ejim.2020.01.009. Epub 2020 Jan 23.
Regulatory agencies warned against fluoroquinolones for the management of minor infections because of the risk of emerging adverse events (collagen-associated adverse events, neuropsychiatric toxicity and long-term disability). We aimed to assess quality and credibility of evidence as well as causality regarding these putative associations.
MEDLINE, Scopus, Web of Science and PROSPERO were searched, from inception to August 2019, for systematic reviews with meta-analyses investigating emerging adverse events. Two investigators extracted data to grade quality (through validated AMSTAR-2 tool), rank credibility of the evidence (convincing, highly suggestive, suggestive, weak) through adapted criteria including E-value calculation, and assess causality (Hill's criteria).
Seven systematic reviews of observational studies providing 16 risk estimates [seven, five and four, respectively, for aortic aneurysm/dissection (AAD), retinal detachment (RD) and any tendon disorders (ATD)] met inclusion criteria. No systematic reviews with meta-analysis investigating the risk of neuropsychiatric toxicity or long-term disability were found. The associations between fluoroquinolones and AAD/ATD showed highly suggestive credibility and were supported by strong evidence of causality (double increased risk, especially within first 2 months of treatment). Conflicting data concerning the emergence of RD were retrieved, resulting in weak evidence of causality. Quality of the evidence ranged from high to low for AAD, from moderate to critically low for RD, and it was moderate for ATD.
Our analysis supports credible, plausible and highly suggestive associations with AAD (rare occurrence but strong causality) and ATD. Limitations of both umbrella reviews and observational evidence should be considered.
由于潜在的不良事件（与胶原相关的不良事件、神经精神毒性和长期残疾）的出现风险，监管机构警告不要将氟喹诺酮类药物用于轻微感染的治疗。我们旨在评估这些潜在关联的证据质量和可信度以及因果关系。
从 2019 年 8 月开始，通过 MEDLINE、Scopus、Web of Science 和 PROSPERO 搜索了系统评价，包括荟萃分析，以调查新出现的不良事件。两名研究人员提取数据，通过验证的 AMSTAR-2 工具对质量进行评分，通过包括 E 值计算在内的改编标准对证据可信度进行排名（令人信服、高度提示、提示、薄弱），并评估因果关系（希尔斯标准）。
符合纳入标准的有 7 项观察性研究的系统评价，提供了 16 项风险估计值[分别为主动脉瘤/夹层（AAD）、视网膜脱离（RD）和任何肌腱疾病（ATD）的 7、5 和 4 项]。未发现针对神经精神毒性或长期残疾风险的系统评价。氟喹诺酮类药物与 AAD/ATD 的关联具有高度提示可信度，并得到了强有力的因果关系证据的支持（风险增加两倍，尤其是在治疗的头 2 个月内）。有关 RD 出现的相互矛盾的数据导致因果关系的证据较弱。AAD 的证据质量从高到低不等，RD 的证据质量从中等到极低不等，ATD 的证据质量为中度。
我们的分析支持与 AAD（罕见但强因果关系）和 ATD 具有可信、合理和高度提示性的关联。应考虑伞式综述和观察性证据的局限性。
