Scholarlink

Background <p>While guidelines recommend severity scores for community-acquired pneumonia (CAP), their predictive accuracy across diverse clinical scenarios—from acute instability to resource utilization—remains incompletely understood.</p> Objectives <p>To evaluate the performance of five severity scores (CURB-65, PSI, qSOFA, APUA, BAR) in predicting acute deterioration, complications, and hospital readmission in CAP.</p> Methods <p>This retrospective cohort study included 332 adults hospitalized with CAP (January 2020–June 2025). Primary outcomes were ICU admission, in-hospital mortality, and sepsis. Secondary outcomes included pleural effusion, prolonged hospitalization (≥ 7 days), and 30-day revisits. Discrimination was assessed using area under the curve (AUC) and pairwise DeLong tests.</p> Results <p>Among 332 patients (median age 73.0 years; 60.2% male), 14.5% required ICU admission, 8.4% died, and 42.2% required prolonged hospitalization. Demonstrating a clear separation in performance, qSOFA showed the highest discrimination for in-hospital mortality (AUC 0.781) and sepsis (AUC 0.762), outperforming CURB-65 (<i>p</i> &lt; 0.05). For resource utilization, APUA significantly outperformed all other scores in predicting prolonged hospitalization (AUC 0.719; <i>p</i> &lt; 0.05) and, alongside PSI, showed superior discrimination for pleural effusion (AUC &gt; 0.70). None of the scores successfully predicted 30-day respiratory-related revisits (AUC range 0.433–0.488).</p> Conclusions <p>Pneumonia severity scores demonstrate a domain-specific performance profile. qSOFA is the superior tool for predicting acute physiological instability (sepsis/mortality), whereas the multidimensional APUA score is the most robust predictor for intermediate-term complications and length of stay. Post-discharge revisits appear to be driven by factors beyond initial biological severity.</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

Outcome-Specific Prognostic Performance of Pneumonia Severity Assessment Tools for Predicting a Spectrum of Multiple Clinical Events

  • 期刊论文

  • 发表日期：

    2026年7月4日

查看全文
  • Abuzer Özkan,
  • Ahmet Taha Özatak,
  • Soner Yeşilyurt

摘要

Background

While guidelines recommend severity scores for community-acquired pneumonia (CAP), their predictive accuracy across diverse clinical scenarios—from acute instability to resource utilization—remains incompletely understood.

Objectives

To evaluate the performance of five severity scores (CURB-65, PSI, qSOFA, APUA, BAR) in predicting acute deterioration, complications, and hospital readmission in CAP.

Methods

This retrospective cohort study included 332 adults hospitalized with CAP (January 2020–June 2025). Primary outcomes were ICU admission, in-hospital mortality, and sepsis. Secondary outcomes included pleural effusion, prolonged hospitalization (≥ 7 days), and 30-day revisits. Discrimination was assessed using area under the curve (AUC) and pairwise DeLong tests.

Results

Among 332 patients (median age 73.0 years; 60.2% male), 14.5% required ICU admission, 8.4% died, and 42.2% required prolonged hospitalization. Demonstrating a clear separation in performance, qSOFA showed the highest discrimination for in-hospital mortality (AUC 0.781) and sepsis (AUC 0.762), outperforming CURB-65 (p < 0.05). For resource utilization, APUA significantly outperformed all other scores in predicting prolonged hospitalization (AUC 0.719; p < 0.05) and, alongside PSI, showed superior discrimination for pleural effusion (AUC > 0.70). None of the scores successfully predicted 30-day respiratory-related revisits (AUC range 0.433–0.488).

Conclusions

Pneumonia severity scores demonstrate a domain-specific performance profile. qSOFA is the superior tool for predicting acute physiological instability (sepsis/mortality), whereas the multidimensional APUA score is the most robust predictor for intermediate-term complications and length of stay. Post-discharge revisits appear to be driven by factors beyond initial biological severity.