Improved Febrile Child Referral: sTREM1 & Pulse Oximetry vs WHO Criteria

New prediction models combining standard clinical assessments with either pulse oximetry or a blood test measuring soluble triggering receptor expressed on myeloid cells-1 (sTREM1) show improved accuracy in identifying febrile children in low-resource settings who require urgent hospital referral, according to a study published April 29, 2026, in Nature Medicine.

The multicountry cohort study, led by researchers from Médecins Sans Frontières (MSF) and the University of Oxford’s Mahidol Oxford Tropical Medicine Research Unit (MORU), found that the new models significantly outperformed current World Health Organization (WHO) criteria for identifying children at risk of severe disease. The research involved data from 3,405 children aged 1–59 months presenting with community-acquired acute febrile illnesses across Bangladesh, Cambodia, Indonesia, Laos, and Vietnam. Cambodian data were used for external validation.

Improved Accuracy in Identifying High-Risk Children

Current WHO danger signs missed nearly half (44.5%) of the children who ultimately died or required organ support, the study revealed. In contrast, the new models—integrating clinical signs with pulse oximetry or sTREM1 levels—identified approximately 89% of these high-risk children.

The model utilizing simple clinical parameters alone demonstrated a sensitivity of 74.7% (95% confidence interval: 59.4–88.1%) and a specificity of 99.1% (95% confidence interval: 97.7–99.7%) for identifying children at risk of severe disease. Incorporating either pulse oximetry or sTREM1 further increased sensitivity to 88.9% (95% confidence interval: 76.7–97.8%) and 89.2% (95% confidence interval: 76.9–97.5%), respectively.

Cost-Effectiveness and Potential Impact

The pulse oximetry-based model not only improved accuracy but also led to a threefold reduction in referral rates. Researchers found these approaches to be cost-effective, with an incremental cost-effectiveness ratio (ICER) of $26.28 for pulse oximetry and $196.46 for sTREM1.

“MSF operates in remote and conflict-affected areas where reaching higher-level care is often irregular and fraught with difficulty. A simple, deployable tool that enables health workers in decentralised care settings to correctly identify children at high risk of sepsis would be invaluable,”

Dr. Sakib Burza, MSF Principal Investigator of the study

The findings address a critical gap in pediatric emergency care, where existing WHO protocols often fail to identify children who develop life-threatening complications or lead to unnecessary referrals of stable patients to overburdened hospitals. The study suggests that the new models could transform referral practices for febrile children in resource-constrained community settings.

sTREM1 as a Promising Biomarker

The biomarker sTREM1, which reflects immune system activation during severe infections, emerged as a particularly promising tool for frontline healthcare workers operating without access to advanced diagnostic equipment. This is especially relevant in settings where laboratory infrastructure is limited.

The research validates practical diagnostic approaches that could be implemented immediately in community health settings across resource-constrained regions. Researchers recommend further evaluation of these models in randomized controlled trials to confirm their effectiveness and guide widespread adoption.

The study builds on previous research highlighting the limitations of current WHO danger signs in accurately identifying children needing referral. The improved predictive ability of the new models offers the potential to significantly reduce both preventable deaths and unnecessary healthcare burdens in vulnerable populations.