Understanding NIPT Accuracy

Footnotes & caveats

PrenatalSafe (Eurofins Genoma)

‖Microdeletion performance data is pooled across all microdeletions, not per-condition. Pooled sensitivity 83.33% has a very wide 95% confidence interval (35.88–99.58%) reflecting a small validation cohort. Per-condition PPV and NPV are not published by the manufacturer.

‖Rare autosomal aneuploidy data (T15, T16, T22) is pooled: sensitivity 100% (89.42–100%), specificity 99.92% (99.89–99.95%). Per-condition PPV/NPV not published.

Per-condition PPV values shown for main aneuploidies (T21 90.9%, T18 97.6%, T13 73.3%, etc.) are derived by Bayesian calculation from the manufacturer's published sensitivity/specificity and an assumed population prevalence. Only sensitivity and specificity are directly published by Eurofins Genoma.

Performance data comes from Eurofins Genoma Italy's internal validation on 71,740 pregnancies using the CE-IVD NIPT FLOW™ algorithm. PrenatalSafe 3 UK and 5 UK are processed in the UK using Illumina's VeriSeq NIPT Solution v2, which has its own clinical validation (Illumina package insert, document 200006957-00).

PrenatalSafe Complete Plus also covers inherited genetic diseases via direct fetal DNA testing: Cystic Fibrosis (CFTR), Sickle Cell Disease and Beta Thalassemia (HBB), and autosomal recessive deafness (GJB2/GJB6). This uses a different methodology to Unity Complete's maternal carrier workflow and is not shown in the Carrier Screening rows above.

KNOVA (Fulgent)

Every aneuploidy in the KNOVA column is shown with its own per-condition values from Fulgent's analytical-performance appendix: analytical sensitivity is >99.9% across the trisomies and sex-chromosome aneuploidies, specificity is 99.9% or higher, and positive predictive value ranges from 83.3% (Jacob's, 47,XYY) to >99.9% (Trisomy 21, Trisomy 13, Turner and Klinefelter) according to each condition's background prevalence. Individual aneuploidy performance may vary.

Every microdeletion in the KNOVA column shares the same appendix values: sensitivity >99.9%, specificity 99.8%, PPV 69.2% and NPV >99.9%. The PPV is modest because each microdeletion is individually rare; the 95% confidence interval is wide (PPV: 38.6-90.9%), reflecting small validation cohorts.

Per-gene monogenic sensitivities range from 43.7% (NSD1, Sotos) to >99% with wide confidence intervals and small validation cohorts. The shared specificity and NPV are >99.9% and the shared PPV is 97.6%, pooled across the gene panel.

KNOVA does not detect: triploidy, tetraploidy, uniparental disomy, translocations, inversions, ring chromosomes, or partial chromosomal aberrations within target regions. It reports only pathogenic or likely pathogenic variants per ACMG criteria - variants of uncertain significance are not reported.

KNOVA is a CLIA-certified Laboratory Developed Test and has not been cleared or approved by the FDA.

KNOVA screens additional conditions not listed above as individual rows due to their rarity and the shared performance profile. These include: Campomelic dysplasia (SOX9), BNAR syndrome (FREM1), Progeria (LMNA), C syndrome (CD96), Schmid metaphyseal chondrodysplasia (COL10A1), Conradi-Hünermann syndrome (EBP), Pfeiffer/Kallmann (FGFR1), Larsen syndrome (FLNB), Greig cephalopolysyndactyly (GLI3), Au-Kline syndrome (HNRNPK), Say-Barber-Biesecker-Young-Simpson (KAT6B), CHILD syndrome (NSDHL), Shprintzen-Goldberg (SKI), Fontaine progeroid (SLC25A24), Cerebrocostomandibular (SNRPB), Opitz GBBB/Teebi (SPECC1L), Hyper-IgE syndrome (STAT3), TRAF7-related syndrome, ZIC1-related craniosynostosis type 6. Per-gene sensitivity ranges from 43.7% to >99%; see Fulgent performance appendix for details.

Panorama (Natera)

*For 22q11.2 deletion (DiGeorge) NPV: ongoing clinical follow-up is performed to ensure the NPV does not fall below the quoted value, but follow-up is not obtained for all low-risk calls.

‡Performance depends on fetal fraction (FF). For 22q11.2 deletion syndrome, only the paternal allele is evaluated at FF ≤6.5%. For 1p36 deletion syndrome and Cri-du-chat syndrome, only the paternal allele is evaluated at FF <7%.

§For Angelman syndrome, no risk assessment is reported at FF <7%. For Prader-Willi syndrome, no risk assessment is reported at FF ≤2.8%.

†Per-condition PPVs for Turner, Klinefelter, Triple X and Jacob's syndromes are not published in Natera's validated Panorama performance table. The source reports fetal sex identification (Female/Male) Sens/Spec but does not quantify per-condition PPVs for sex chromosome aneuploidies.

Performance values frequency-weighted from published validation studies; per-condition confidence intervals omitted for readability. Full CIs available from source references.