In our study, we found that NIPT primary to general population could fulfill efficient screening for T21 but more costly. Contingent NIPT with risk threshold as 1/300 (strategy 2 − 1) could effectively reduce unnecessary invasive testing, with best performance in total cost, cost effect and benefit analysis. Setting strategy 2 − 1 as the baseline, current Chinese age-stratified contingent NIPT (strategy 3 − 1) detected one additional case with least incremental cost, which was also less than the costs needed to raise one viable T21. Taking the aging society for sensitivity analysis, strategy 3 − 1 was optimal with increasing proportion of AMA, and vice versa. The primary NIPT demonstrated its screening accuracy among all the scenarios as the most cost effectiveness option when the price dropped to 47 US$.
Overall, our study proposed the advantages of contingent NIPT in “high risk” population in total costs and cost-effect analysis, which was consistent with previous studies21 − 24. Prenatal diagnosis has been directly provided to AMA. T21 incidence in AMA was similar to those identified by serum screening among general populations40,42, which provide rational evidences to provide direct invasive tests for AMA, and therefore we specifically considered the influence of aging society in sensitivity analysis. The advantage of age-based strategy was more prominent when AMA increased. The reverse was also true. Strategy 3 − 1 remained to be the best option if the reproductive age delayed, with pregnancies younger than 35 less than 85%.
In current study, the appropriate threshold for contingent strategy was different from Evans et al study24 although both studies shared similar design on age proportion and NIPT acceptance in total cost and cost effect analysis. In fact, both studies found that contingent strategy with 1/300 was optimal without taking costs for raising T21 into consideration. In incremental cost analysis, the underlying explanation could relate to the NIPT price, which was relative higher in our system, along with the increased cases adopting NIPT when lower the cut-off to 1/1000. When NIPT acceptance changed, the incremental cost tendency was opposite in age-stratified and contingent strategy. The inconsistence also verified the key influence of NIPT price on decision of strategy implementation, that is the advantages of age-based strategy was dominant if NIPT price increased.
When sequencing is affordable, primary NIPT has been launched as the optimal choice with wide application in some countries and regions27. In sensitive analysis of our model, the tendency of incremental cost changed the most in primary NIPT. With decreasing NIPT cost43, primary NIPT becomes costly and dominant. Meanwhile, in other scenarios, the incremental costs changed less due to the costs for testing or miscarriage were far below to make the influence.
Current study provide evidence from public health perspective on how to introduce NIPT into clinics. Using parameters of real world, the proposed strategy verified the current guideline. We raised two key factors influencing the policy making, which should be taken into consideration periodically in the context of aging and NIPT price reduction. There were several strengthens. The widely used clinical indications for NIPT were AMA and high risk identified by serum screening44, hence both contingent and age-stratified strategies were involved in design. Taking 1/300 and 1/1000 into consideration enabled us to identify the more appropriate risk level for NIPT. For areas with different NIPT prices, the sensitive analysis of tendency for incremental costs enable us to cover real-world situations in different regions to figure out the turning point.
There were several limitations for the theoretical model. First, this is a theoretical model. Although the associated publications had been integrated to determine the parameters, with further analysis through sensitivity analysis, the conclusion could be augured on the actual costs, the demographics of the population, the degree of utilization, or all other real-life factors. Second, the value of primary NIPT should extend to other genetic anomalies including the copy number variations14,45. At present, NIPT cannot replace the prenatal diagnosis (CVS or amniocentesis) given the accuracy and wide detectable spectrum with the development of next generation sequencing (NGS)46. The importance of prenatal diagnosis has been emphasized by professional guidelines worldwide, and offered as an option for general pregnancies. T21 detection was the priority in this model. The consideration was that the primary indication of prenatal diagnosis has been the advanced maternal age (AMA) in the context of delayed reproductive age. Third, the model didn’t take religious and cultural factors into consideration. Fourth, due to absence of social investigation, the study focused on public health perspective other than views from all stakeholder. Last, we didn’t calculate quality adjusted life years (QALY), which is standard and desirable indicator in cost-utility analysis47. The controversial and area-specific judgement of parental preferences in prenatal diagnosis48 was the main concern. Some possible collateral effects such as work absence was negligible and we didn’t include it in our analysis.