Scholarlink

Background <p>The 2024 Phoenix sepsis criteria define sepsis and septic shock in children as life-threatening organ dysfunction, operationalized by a Phoenix Sepsis Score of 2 or more, in presence of suspected or confirmed infection. The Phoenix sepsis criteria have not been validated in children with proven bacterial infection.</p> Objective <p>To determine the diagnostic performance of Phoenix sepsis criteria in predicting mortality, and mortality or prolonged intensive care unit (ICU) stay in a population-based cohort of children with blood culture-proven bacterial infection.</p> Methods <p>We assessed the discriminative power of the scores for the primary outcome, 30-day in-hospital mortality, and the secondary outcome, 30-day in-hospital mortality or ICU length of stay of 3 days or longer, after blood culture sampling, using area under the curve of receiver operating characteristic curves (AUROC) and the area under the precision recall curve (AUPRC). Logistic mixed-effects models were adjusted for age, sex, and presence of chronic medical conditions.</p> Results <p>877 bacteremia episodes in 807 children were analyzed. Mortality was 1% (4/724 episodes) in children without sepsis according to the Phoenix sepsis criteria, 18% (27/153 episodes) in children with sepsis, and 24% (25/104 episodes) in children with septic shock. For 30-day mortality, the Phoenix Sepsis Score had an adjusted AUROC of 0.89 (95% confidence interval 0.83–0.95), and an adjusted AUPRC of 0.49 (95% confidence interval 0.34–0.65), both similar to the Phoenix-8 Score and comparing favorably to IPSCC, pSOFA, PELOD-2, and PODIUM. The positive predictive value was 0.18 (95% confidence interval 0.12–0.24), higher than all other analyzed scores, and the sensitivity 0.71 (95% confidence interval 0.56–0.86) comparing favorably to IPSCC, PELOD-2, and PODIUM. For the secondary outcome, 30-day in-hospital mortality or ICU stay of 3 days or longer, the performance of the Phoenix Sepsis Score and Phoenix-8 Score was best compared to IPSCC, pSOFA, PELOD-2, and PODIUM both in terms of AUROC and AUPRC, with the Phoenix Sepsis Score again having the highest positive predictive value of all analyzed scores.</p> Conclusions <p>In this independent validation of the Phoenix sepsis criteria in a national cohort of children with blood culture-proven bacterial infection, the Phoenix Sepsis Score performed best to predict 30-day mortality compared to other pediatric organ dysfunction scores.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Validation of the Phoenix Sepsis Score in children with blood culture-proven infection – a national prospective cohort study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月21日

查看全文
  • Luregn J. Schlapbach,
  • Sabrina Goertz,
  • Blandine Aubert,
  • Sebastien Papis,
  • Eric Giannoni,
  • Klara M. Posfay-Barbe,
  • Martin Stocker,
  • Ulrich Heininger,
  • Sara Bernhard-Stirnemann,
  • Anita Niederer-Loher,
  • Christian R. Kahlert,
  • Giancarlo Natalucci,
  • Christa Relly,
  • Thomas Riedel,
  • Christoph Aebi,
  • Christoph Berger,
  • Philipp K. A. Agyeman

摘要

Background

The 2024 Phoenix sepsis criteria define sepsis and septic shock in children as life-threatening organ dysfunction, operationalized by a Phoenix Sepsis Score of 2 or more, in presence of suspected or confirmed infection. The Phoenix sepsis criteria have not been validated in children with proven bacterial infection.

Objective

To determine the diagnostic performance of Phoenix sepsis criteria in predicting mortality, and mortality or prolonged intensive care unit (ICU) stay in a population-based cohort of children with blood culture-proven bacterial infection.

Methods

We assessed the discriminative power of the scores for the primary outcome, 30-day in-hospital mortality, and the secondary outcome, 30-day in-hospital mortality or ICU length of stay of 3 days or longer, after blood culture sampling, using area under the curve of receiver operating characteristic curves (AUROC) and the area under the precision recall curve (AUPRC). Logistic mixed-effects models were adjusted for age, sex, and presence of chronic medical conditions.

Results

877 bacteremia episodes in 807 children were analyzed. Mortality was 1% (4/724 episodes) in children without sepsis according to the Phoenix sepsis criteria, 18% (27/153 episodes) in children with sepsis, and 24% (25/104 episodes) in children with septic shock. For 30-day mortality, the Phoenix Sepsis Score had an adjusted AUROC of 0.89 (95% confidence interval 0.83–0.95), and an adjusted AUPRC of 0.49 (95% confidence interval 0.34–0.65), both similar to the Phoenix-8 Score and comparing favorably to IPSCC, pSOFA, PELOD-2, and PODIUM. The positive predictive value was 0.18 (95% confidence interval 0.12–0.24), higher than all other analyzed scores, and the sensitivity 0.71 (95% confidence interval 0.56–0.86) comparing favorably to IPSCC, PELOD-2, and PODIUM. For the secondary outcome, 30-day in-hospital mortality or ICU stay of 3 days or longer, the performance of the Phoenix Sepsis Score and Phoenix-8 Score was best compared to IPSCC, pSOFA, PELOD-2, and PODIUM both in terms of AUROC and AUPRC, with the Phoenix Sepsis Score again having the highest positive predictive value of all analyzed scores.

Conclusions

In this independent validation of the Phoenix sepsis criteria in a national cohort of children with blood culture-proven bacterial infection, the Phoenix Sepsis Score performed best to predict 30-day mortality compared to other pediatric organ dysfunction scores.