Clinical Warning Signs in Detecting Inborn Errors of Immunity in Children: A Diagnostic Accuracy Systematic Review.

Abstract

Early recognition of inborn errors of immunity (IEI), formerly primary immunodeficiencies, is crucial, yet the diagnostic accuracy of widely promoted clinical "warning signs" in children is uncertain. To assess the diagnostic accuracy of clinical warning signs and established criteria for identifying pediatric IEI. We searched MEDLINE and EMBASE (inception-May 2024) for studies reporting sensitivity and specificity of individual or combined warning signs in children (0-18 years) with suspected IEI. The reference standard was physician-confirmed IEI. Data were synthesized descriptively and, where appropriate, using bivariate random-effects models. Risk of bias was assessed with the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2) tool. Twelve studies (∼7000 participants, 8 countries) met inclusion criteria. Diagnostic performance was highly heterogeneous, and pooled estimates are exploratory. Overall, most individual warning signs showed high specificity but low sensitivity; family history of IEI and signs such as persistent thrush, deep-seated abscesses, and failure to thrive were generally highly specific but insensitive. Infection-based signs such as recurrent pneumonia or need for intravenous antibiotics provided a more balanced but still imperfect profile. Combinations of signs, including the Jeffrey Modell Foundation threshold of 2 or more warning signs, improved sensitivity in some settings but with variable specificity, and up to one-third of children with IEI did not meet any Jeffrey Modell Foundation warning-sign criteria. Traditional warning signs have good rule-in but poor rule-out value for pediatric IEI. Reliance on these signs alone risks delayed or missed diagnoses, particularly for noninfectious IEI phenotypes. Context-specific criteria and decision-support tools are needed to better capture the breadth of IEI presentations and support timely referral.