文献速递临床微生物质量管理与危急值

血培养生长模式对鉴别污染与真病原体的诊断价值

Diagnostic value of blood culture growth patterns in distinguishing contaminants from pathogens.
PubMed 文献 · J Clin Microbiol · 2026-01-22

该研究回顾性分析2019年和2024年外周血培养组,比较单瓶生长(不一致)与双瓶生长(一致)对污染与真病原体的鉴别价值。共分析38,216组血培养,其中不一致组1,491组(3.9%),一致组1,938组(5.1%)。不一致组中污染占71.1%,真病原体占28.9%;一致组中污染占11.4%,真病原体占88.5%(P < 0.001)。

要点

  • 不一致生长组中污染比例显著高于一致生长组(71.1% vs 11.4%)。
  • 凝固酶阴性葡萄球菌不一致生长对排除真菌血症的阴性预测值为98.1%。
  • 需氧瓶更常检出污染菌(37.9% vs 20.5%),可能因分流效应。

核心结论

  • 不一致生长的凝固酶阴性葡萄球菌培养与污染强相关,可辅助血培养结果解读。
  • 不一致生长模式分析可在24小时内为临床提供及时信息,支持抗菌药物管理。

研究设计与样本

  • 回顾性分析2019年和2024年外周血培养组,共38,216组,其中不一致组1,491组(3.9%),一致组1,938组(5.1%)。
  • 根据临床意义将生长模式分为污染与真病原体进行分析。

主要发现

  • 不一致组培养出污染菌1,060/1,491(71.1%),真病原体431/1,491(28.9%)。
  • 一致组培养出污染菌222/1,938(11.4%),真病原体1,716/1,938(88.5%)(P < 0.001)。
  • 仅分析凝固酶阴性葡萄球菌(2019年数据),不一致组中629/642(98.0%)为污染,13/642(2.0%)为真病原体。
  • 金黄色葡萄球菌仅82/270(30.4%)在不一致组中生长。
  • 在858例患者首次凝固酶阴性葡萄球菌阳性培养中,不一致组624/636(98.1%)为污染,12/636(1.9%)为真病原体。

关键数值与推荐

  • 不一致首次凝固酶阴性葡萄球菌培养排除真菌血症的阴性预测值为98.1%(95%CI 96.7%-98.9%)。
  • 在356组不一致培养中,污染菌更常见于需氧瓶(135/356,37.9% vs 73/356,20.5%,P = 0.04)。
  • 真病原体在需氧瓶和厌氧瓶中的比例相似(79/356,22.2% vs 69/356,19.4%,P = 0.4)。

临床怎么用

  • 利用生长模式(单瓶 vs 双瓶)辅助早期鉴别血培养污染与真病原体。
  • 不一致生长的凝固酶阴性葡萄球菌培养结果可高概率排除真菌血症,支持抗菌药物合理使用。

局限性

  • 回顾性研究设计。
  • 部分分析仅基于2019年数据。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

Blood culture contamination is common, causing diagnostic uncertainty and unnecessary antibiotic use. Analyzing growth patterns within culture sets might offer diagnostic value. We retrospectively analyzed peripheral blood culture sets from 2019 and 2024. Growth pattern (one bottle [discordant] vs both bottles [concordant]) was analyzed according to clinical significance (contaminant vs true pathogen). Overall, 38,216 blood culture sets were analyzed, including 1,491 (3.9%) discordant and 1,938 (5.1%) concordant sets (remaining 34,787 [91.0%] sets were sterile). Discordant sets grew 1,060/1,491 (71.1%) contaminants and 431/1,491 (28.9%) true pathogens. Concordant sets grew 222/1,938 (11.4%) contaminants and 1,716/1,938 (88.5%) true pathogens (P < 0.001). Examining coagulase-negative staphylococci (CoNS) only (2019 data set), 629/642 (98.0%) discordant sets grew contaminants, while 13/642 (2.0%) grew true pathogens. In contrast, Staphylococcus aureus grew in only 82/270 (30.4%) discordant sets. Among 858 first CoNS-positive cultures per patient, 624/636 (98.1%) discordant sets grew contaminants, and 12/636 (1.9%) grew CoNS defined as a true pathogen. The negative predictive value of a discordant first CoNS set to exclude true CoNS bacteremia was 98.1% (95% confidence interval 96.7%-98.9%). Examining aerobic vs anaerobic bottles in 356 discordant sets, contaminants were found more frequently in aerobic bottles (135/356, 37.9% vs 73/356, 20.5%, P = 0.04). The proportion of true pathogens was similar in both (79/356, 22.2% vs 69/356, 19.4%, P = 0.4). Discordant CoNS-positive cultures were strongly associated with contamination. This could assist in interpreting blood culture results and supporting antimicrobial stewardship. Discordance might result from a diversion effect, the aerobic bottle acting as a diversion device for the anaerobic bottle.IMPORTANCERapidly distinguishing blood culture contaminants from true pathogens is essential for optimizing antimicrobial stewardship and avoiding unnecessary antibiotic therapy. In this large, two-period study, we demonstrate that discordant growth of coagulase-negative staphylococci in a two-bottle set has a negative predictive value of 98.1% for true bacteremia. This finding remained robust across both study years and when restricted to first positive cultures, highlighting its reliability. Incorporating simple growth pattern analysis into early blood culture interpretation can provide clinicians with reliable and timely information within 24 hours, supporting more targeted and judicious antibiotic use.

来源:J Clin Microbiol,查看原文。版权归原作者所有。

引用

  1. Ben-Chetrit E, Helviz Y, Levin PD, et al. Diagnostic value of blood culture growth patterns in distinguishing contaminants from pathogens. J Clin Microbiol. 2026. PMID: 41568951. doi:10.1128/jcm.01210-25

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。