SeqFirst: Building equity access to a precise genetic diagnosis in critically ill newborns

Abstract

Access to a precise genetic diagnosis (PrGD) in critically ill newborns is limited and inequitable because the complex inclusion criteria used to prioritize testing eligibility omit many patients at high risk for a genetic condition. SeqFirst-neo is a program to test whether a genotype-driven workflow using simple, broad exclusion criteria to assess eligibility for rapid genome sequencing (rGS) increases access to a PrGD in critically ill newborns. All 408 newborns admitted to a neonatal intensive care unit between January 2021 and February 2022 were assessed, and of 240 eligible infants, 126 were offered rGS (i.e., intervention group [IG]) and compared to 114 infants who received conventional care in parallel (i.e., conventional care group [CCG]). A PrGD was made in 62/126 (49.2%) IG neonates compared to 11/114 (9.7%) CCG infants. The odds of receiving a PrGD were ∼9 times greater in the IG vs. the CCG, and this difference was maintained at 12 months follow-up. Access to a PrGD in the IG vs. CCG differed significantly between infants identified as non-White (34/74, 45.9% vs. 6/29, 20.7%; p = 0.024) and Black (8/10, 80.0% vs. 0/4; p = 0.015). Neonatologists were significantly less successful at predicting a PrGD in non-White than non-Hispanic White infants. The use of a standard workflow in the IG with a PrGD revealed that a PrGD would have been missed in 26/62 (42%) infants. The use of simple, broad exclusion criteria that increase access to genetic testing significantly increases access to a PrGD, improves access equity, and results in fewer missed diagnoses.

Keywords: equity; genetic diagnosis; health disparity; rapid whole-genome sequencing; rare disease.

Figure 1

STROBE diagram SeqFirst-neo cohort ascertainment of interventional group (IG) and conventional care group (CCG) depicted using a STROBE diagram.

Figure 2

Adjudication of variants reported in patients in the intervention cohort Of the 126 participants, 85 had one or more potentially explanatory variants reported by GeneDx based on the American College of Medical Genetics and Genomics (ACMG) guidelines for clinical sequence interpretation (variant of uncertain significance [VUS], likely pathogenic [LP], and pathogenic [P]). We grouped these variants by the genes/regions in which they were reported (i.e., single gene, cytogenetic event, or mitochondrial event) and the inferred inheritance pattern(s) of the condition(s) they underlie (autosomal dominant [AD] or X-linked [XL], autosomal recessive [AR], or unknown mode of inheritance [MOI] for genes that underlie both AD and AR conditions). Variant(s) or event(s) and the MOI were then adjudicated together to determine whether each combination was explanatory of clinical findings (either partly or fully), unrelated to clinical findings and secondary, or of uncertain relationship to clinical findings of each participant. Variants/events in each participant were counted separately so that, for example, the same genotype in two participants would be counted twice.

Figure 3

Diagnostic yield in the intervention group Overall diagnostic yield stratified by test result category in the intervention group (IG) and the conventional care group (CCG) at 90 days and 1 year post-ascertainment.

Figure 4

Diagnostic yield in the intervention group by phenotypic presentation and by racial construct (A) Diagnostic yield in the intervention group (IG) stratified by phenotype presentation, parent- or provider-assigned racial construct (PPARC), and test result category. MCA, multiple congenital anomalies; CHD, congenital heart defects. (B) Access to a PrGD/diagnostic yield in the IG, access to a PrGD in the conventional care group (CCG), and the diagnostic yield in the CCG who received genetic testing (CCG-tested) stratified by PPARC. PrGD, precise genetic diagnosis. Asterisks denote statistically significant differences between groups.

Figure 5

Impact of PPARC on the prediction of precise genetic diagnosis in newborns Prediction of precise genetic diagnosis (PrGD) status for non-White and non-Hispanic White infants in the intervention group (IG) as assessed by patterns of referral for a genetics consult by their neonatologists. We defined successful prediction of PrGD status as requesting a consultation on a patient who was later found to have a PrGD and not requesting consultation on a patient who was not found to have a PrGD. Neonatologists missed requesting a genetics consult significantly more often (2.4-fold) for non-White infants who would eventually receive a PrGD and also referred non-White infants for consults who were 0.17-fold less likely to go on to receive a PrGD. This translates to higher precision (75%) and recall (86%) for predicting PrGD status for non-Hispanic White infants than for non-White infants (55% vs. 68%, respectively).

Figure 6

Impact of a precise genetic diagnosis on changes in management The impact of a precise genetic diagnosis on changes in management stratified by parent- or provider-assigned racial construct (PPARC). Management options are reflected on the bottom left, with blackened circles representing assessment by clinical team. Multiple options could be selected. The vertical black lines connecting blackened circles represent a unique combination of changes in management options, and all the unique changes in management combinations reported more than once are shown. Bar plots above each set of responses indicate the number of families with each change in management combination stratified by PPARC. Guidance, anticipatory guidance; counseling, genetic counseling; consults, request for additional consults; labs, ordering additional laboratory tests; cascade evaluation, evaluation of other family members; imaging, ordering additional imaging studies; meds, addition of medications; cancel orders, canceling laboratory tests, imaging studies, or medications.