Diagnostic accuracy of PCR-based screening methods for congenital cytomegalovirus in newborns: a systematic review of specimen types and testing strategies

Congenital cytomegalovirus (cCMV) is the most common congenital viral infection worldwide and a leading cause of non-genetic sensorineural hearing loss in children. Early identification through universal newborn screening may enable timely intervention and improved long-term outcomes. However, the diagnostic performance of available PCR-based screening methods varies by specimen type. This systematic review sought to evaluate the diagnostic performance of PCR-based screening approaches for cCMV in newborns. A systematic review was conducted in accordance with Cochrane and PRISMA guidelines. Bibliographic databases and grey literature sources were searched for studies evaluating PCR-based screening for cCMV in newborns within the first three weeks of life using a universal screening approach. Study selection, data extraction, and quality assessment (QUADAS-2 and QUADAS-C) were performed independently by two reviewers. Results were synthesized narratively due to substantial heterogeneity. A total of 1,599 citations were identified, of which 1,586 remained after deduplication, and 14 met inclusion criteria (reporting findings from 19 study cohorts published between 2009 and 2024). Saliva-based PCR testing demonstrated consistently high sensitivity (80.0–98.6%) and specificity (≥ 96% in most studies), with high negative predictive values across settings. Dried blood spot (DBS) testing demonstrated consistently high specificity (97.2–100.0%) but more variable sensitivity (39.3–93.0%). Pooled saliva testing, where samples from multiple newborns are combined and individual retesting is triggered only by a positive pool result, showed high sensitivity (87.5–98.6%) and operational advantages. However, positive predictive values varied widely (12–100%), likely reflecting differences in study design rather than true variation in test performance. Sequential testing designs and incomplete follow-up were common and may have introduced bias. Saliva-based PCR screening demonstrates favourable diagnostic performance for universal cCMV screening, although estimates may be influenced by methodological limitations and variability in collection protocols. DBS testing offers high specificity and feasibility within existing screening infrastructure, but may have lower sensitivity. Screening program decisions should consider diagnostic accuracy alongside feasibility, standardization of collection procedures, and confirmatory testing pathways. Future studies using standardized methods and complete follow-up are needed to generate more robust estimates. • Saliva-based PCR testing demonstrates consistently favourable sensitivity and specificity for cCMV screening, although performance depends on standardized collection protocols. • Dried blood spot testing shows high specificity but variable sensitivity, which may limit its effectiveness as a standalone screening approach. • Pooled saliva testing offers potential operational advantages, including reduced laboratory workload, but requires careful confirmatory strategies. • Many included studies used sequential designs, introducing verification bias and limiting reliable estimation of diagnostic accuracy. • Screening program decisions should balance diagnostic performance with feasibility, infrastructure, and the clinical consequences of missed cases.