OBJECTIVES: Severe combined immunodeficiency (SCID) is a rare, inherited inborn error of immunity that is typically fatal in the absence of definitive treatment. This systematic review aimed to summarise international evidence on the cost effectiveness of newborn screening for SCID compared with detection based on family history or symptomatic presentation.

METHODS: A systematic search was carried out in three databases from January 1, 2010 to June 1, 2022, supplemented by a grey literature search. The primary outcome of interest was the incremental cost-effectiveness ratio (ICER) of newborn screening for SCID compared with no screening. All costs were converted to 2021 Irish Euro. A protocol detailing the methods was published prior to the review. The CHEC-list and ISPOR questionnaire were used to assess methodological quality and transferability of the economic evaluations. The reporting adhered to Preferred Reporting Items for Systematic Reviews and Meta-Analyses criteria.

RESULTS: Eleven studies, representing ten unique models, were included in the review. Of these, three performed a cost-utility analysis (CUA), three performed a cost-effectiveness analysis (CEA), and four presented results for both a CUA and a CEA. Eight studies adopted a healthcare system perspective, one a societal perspective, and one adopted both perspectives in separate analyses. For studies reporting costs per life-years (LYs) gained, adjusted ICERs ranged from €14,027-€217,657/LY. For the studies reporting costs per quality-adjusted LYs (QALYs) gained, the adjusted ICERs ranged from €14,549-€83,670/QALY. Most studies reported that the models were sensitive to variations in a number of key variables, including: test specificity, incidence of SCID, screening and diagnostic test costs, cost of treatment, and survival post-treatment.

CONCLUSIONS: Results indicate that newborn screening for SCID may be a cost-effective intervention, compared with no screening, at willingness-to-pay thresholds between €20,000-€45,000/QALY. However, there is a large amount of uncertainty in terms of the model inputs and, thus, outcomes.