OBJECTIVES: Screening for SCID is possible through the quantification of TRECs, whereby samples below a given cut-off represent T-cell lymphopenia (TCL) and are referred for confirmatory testing. In this way, TREC testing identifies both SCID and other forms of TCL. This study aimed to quantify the test accuracy of TREC-based screening for SCID to inform population-level screening decision-making.

METHODS: A systematic search up to 1 November 2021 was conducted in electronic databases, supplemented with grey literature and hand-searching. A preference was given to studies of real-world screening programmes. Primary outcomes of interest were positive predictive value (PPV) and false positivity rates. Secondary outcomes included operational measures such as rates of retest, resampling and referral.

RESULTS: The review identified 27 unique cohorts reported across 19 studies. There was notable heterogeneity in terms of screening algorithms, test methodologies, TREC cut-offs, and diagnostic criteria used. PPV for SCID alone ranged from 0.80% to 20.00%. PPV for all TCL (including SCID) ranged from 20.29% to 100%. As a proportion of the total population screened, the highest false positivity rate reported was 0.09% (for all TCL [including SCID]). Rates of retest (range 0.24 to 2.03%), repeat sample requests (range 0.02 to 0.61%) and onward referral (range 0.02 to 0.11%) varied. Incidence of SCID and non-SCID TCLs varied considerably across the included cohorts.

CONCLUSIONS: International screening programmes for SCID are heterogeneous, with noted variability in the thresholds, methods and screening algorithms used. As a proportion of the total population screened, the false positivity rates for all TCLs (including SCID) are low, but are considerably higher when restricted to SCID. High false positivity rates increase the number of samples requiring confirmatory testing, since both SCID and non-SCID conditions will require additional work-up once identified by the TREC test, thereby impacting organisational and budget impact considerations.