The Clinical Respiratory Score: Diagnostic Performance for Acute Chest Syndrome in Sickle Cell Disease

Acute chest syndrome (ACS) is a leading cause of morbidity in pediatric sickle cell disease (SCD). The Clinical Respiratory Score (CRS) is used prospectively at the bedside for ACS, whereas the NHLBI Clinical Severity Index (CSI) classifies ACS severity retrospectively. We evaluated the diagnostic performance of peak CRS to identify severe ACS.

We conducted a retrospective cohort study of patients with ACS (one admission each) aged 3-21 years at a tertiary children's hospital (2015-2021). Peak CRS was abstracted from bedside documentation; CSI (1 = mild, 2 = moderate, 3 = severe) was assigned from the chart. The primary analysis estimated sensitivity, specificity, positive/negative predictive values (PPV/NPV), and accuracy of peak CRS ≥ 4 for identifying CSI = 3, with exact 95% confidence intervals. We prespecified a CRS 0-1 subgroup to assess NPV and de-escalation outcomes [respiratory support, transfusion, length of stay (LOS)].

Among 213 patients, peak CRS ≥ 4 predicted CSI = 3 with sensitivity 1.00 (95% CI 0.863-1.000), specificity 0.888 (0.835-0.926), PPV 0.543 (0.402-0.678), NPV 1.000 (0.978-1.000), and accuracy 0.901 (0.854-0.935). No severe events occurred with CRS < 4 (95% upper bound ~1.8%). In CRS 0-1 (n = 74), none had severe ACS (95% upper bound ~4.1%) nor required significant respiratory support; 75.7% were not transfused; and median LOS was 2.3 days (IQR 1.4-3.1).

Peak CRS ≥ 4 identifies a high-risk cohort for severe ACS; CRS 0-1 identifies a low-risk cohort requiring minimal intervention. CRS offers a bidirectional bedside tool to guide escalation and de-escalation of care; prospective validation is warranted.