Illustration of the experimental steps for digital RMD analysis. For each maternal plasma DNA sample, both the mutant DNA proportion (P_r) and the fractional fetal DNA concentration are determined by digital PCRs. (A) P_r is determined by the use of the real-time PCR assay targeting the mutation carried by the mother, whereas fetal DNA proportion is determined by the use of the real-time PCR assay for the homologous ZFY and ZFX gene region. (B) Digital PCR is carried out on a microfluidics Digital Array (Fluidigm), which consists of 12 panels with each panel further partitioned into 765 reaction chambers. Each DNA sample is analyzed with 6 panels (ie, 765 × 6 = 4590 chambers). The PCR mixture is first manually added into the sample inlet of each panel. The mixture is next aliquoted into 765 chambers in each panel automatically by an Integrated Microfluidics Circuit Controller (Fluidigm). Each chamber contains a final reaction volume of 6 nL. The cell-free DNA concentration in maternal plasma is typically very low such that there is less than one template molecule per chamber on average. Hence, the distribution of template molecules to the chambers follows the Poisson distribution. (C) Real-time PCR is performed on the BioMark System (Fluidigm). Most of the chambers contain zero or one template DNA molecule, and the molecules are amplified and detected individually. (D) After PCR, the concentrations of mutant and wild-type DNA, as well as ZFY and ZFX DNA, are calculated. The number of chambers showing positive amplifications for the corresponding allele is counted. The concentration is then Poisson-corrected using the equation [ − ln((N − P)/N)]*N, where N is the total number of reaction chambers analyzed, P is the number of chambers positive for the allele, and ln is the natural logarithm. (E) P_r is calculated from the Poisson-corrected mutant and wild-type DNA concentrations. Fractional fetal DNA concentration is calculated from the Poisson-corrected ZFY and ZFX DNA concentrations. (F) SPRT is used to compare the experimental P_r with the expected P_r at the observed fractional fetal DNA concentration and average reference template concentration (m_r). SPRT classification will be made if the mutant allele is over- or underrepresented when compared to the wild-type allele, which implies an affected or nonaffected fetus, respectively.