A Retrospective Analysis Of Different Contingent Screening Models For Fetal Down Syndrome In Southwestern China

To discuss combinations of traditional screening and noninvasive prenatal screening (NIPS) and to compare which traditional screening is the most suitable first-line screening approach to NIPS, pregnant women were recruited in this retrospective observational study. Pregnant women underwent one of four traditional screening tests. The 9 contingent models were combined by high risk cut-offs of 1:50, 1:100, 1:270 and intermediate risk cut-offs of 1:1000, 1:1500, 1:2000. We analyzed cost and performance of various screening models with contingent screening of different risk cut-offs. Compared with other screening tests, combined first-trimester screening (CFTS) had the lowest proportion of high risk (≥1:270) with the highest detection rate (DR) (78.79%) and the lowest proportion of intermediate risk (1:271~1:1000). When intermediate risk was 1:51 ~1:1500, CFTS as first-line screening had the lowest cost with DR of 93.94%. Other screening tests as the first-line screening with intermediate risk of 1:51~1:1000 had the lowest cost, there DR were 90.91%, 84.62%, 91.67%, respectively. Our study demonstrated if only one traditional screening was allowed to screen pregnant women, CFTS was recommended as the first choice. According to local health and economic conditions, adopting appropriate traditional screening with suitable cut-offs as first-line screening will contributed to a cost-effective screening model.

Performance of traditional screening. The number of patients who reached the three group of the risk based on the screening tests used is shown in Table 2. Following combined first-trimester screening, 370 (1.7%), 732 (3.37%) and 20611(94.92%) patients were classified as high risk, intermediate risk and low risk, respectively. Following quadruple screening, 306(3.78%), 598 (7.39%) and 7183(88.82%) patients were classified as high risk, intermediate risk and low risk, respectively. Following triple screening, 415(3.03%), 943(6.89%) and 12334(90.08%) patients were classified as high risk, intermediate risk and low risk, respectively. Following double screening, 3629 (5.3%), 7753 (11.32%) and 57130(83.39%) patients were classified as high risk, intermediate risk and low risk, respectively. Comparing the proportion of high, intermediate, low risk among the four screening tests, the difference is statistically significant (P < 0.01). The combined first-trimester screening had the lowest proportion of high risk and intermediate risk, followed by triple screening, quadruple screening and double screening.
The high risk detection rate of combined first-trimester screening was 78.79% (26 out of 33 cases), while that for quadruple screening, triple screening and double screening was 72.73% (8 out of 11 cases), 69.23% (9 out of 13 cases) and 70.00% (42 out of 60 case), respectively. There was no significant difference in the detection rate between the four methods (P = 0.829). Table 2 and Table 3.
With high risk cut-off of 1:270, the true negative rate of combined first-trimester screening, quadruple screening, triple screening and double screening were 98.41%(21336 out of 21680 cases), 96.31% (7778 out of 8076

Discussion
Our study presents that because of the lowest proportion (1.7%) of high risk (cut-off 1:270) with the highest detection rate (78.79%) and the highest positive predictive value (7.03%), the performance of combined first-trimester screening is better than second trimester screening tests. Offering combined first-trimester screening led to the lowest number of invasive tests compared with other traditional screenings. And this consequence is consistent with Lan's research report 21 . It is also an advantageous to offer combined first-trimester screening to reduce the risk of iatrogenic fetal loss. Influenced by economic situation, educational background and regional health level, pregnant women have low compliance with the second screening in integrated screening or sequential screening and most pregnant women have received only one traditional screening test in China. If only one traditional screening test is allowed to screen pregnant women regardless of first trimester or second trimester, combined first-trimester screening is recommended as the first choice.    1000) were reassured that fetal trisomies were unlikely and no further testing was necessary. Using this strategy, our study shows contingent model that NIPS contingent on result of combined first-trimester screening has a higher detection rate compared with the second trimester screening tests. And this contingent model can reduce subsequent the number of NIPS, thus reducing the cost of health economics.
The cut-off value of traditional screening needs to consider detection rate, invasive prenatal diagnosis rate and health economic costs and cut-off values of traditional screening in different countries and regions are different. However, with the contingent strategy of traditional screening and NIPS, a more appropriate cut-off should be found. In a study by Gil et al., the authors suggested that NIPS could be offered as a contingent screen following combined first trimester screening results. They proposed that women with intermediate risk (defined as 1:11-2500 in that study) could also be offered NIPS 22 . Some researchers suggested that the introduction of NIPS as a second line screening test, conditional to a risk ≥1:1000 from Standard of Care screening, showed a 3% increase in the detection of trisomies, with a 71% decrease in the number of invasive tests performed 23 . Other previous study suggest that the use of risk scores between 1:251 and 1:1000 may be a more cost-effective threshold 24 . While international experience provides some insights, it is very difficult to forecast how the availability and accessibility of NIPS will affect screening for fetal Down syndrome in China, particularly across models with varied definitions of high and intermediate risk.
When adopting quadruple screening or triple screening or double screening as the first-line screening, with high risk cut-off was 1:50 and intermediate risk cut-off was 1:1000, the proportion of women receiving prenatal diagnosis and NIPS were smaller than other cut-off models. For the second trimester screening tests, after adjusting the different risks, with the increase number of invasive diagnostic test and NIPS, the detection rate increases with the increase of cost. The overall cost and average cost per DS detected of combined first-trimester screening as a first-line screening were both the lowest when intermediate risk defined as 1:51~1500. Moreover, the detection rate of 93.94% remained higher than other cut-off models of combined first-trimester screening. Using this intermediate risk of combined first-trimester screening, the number of missed cases, the number of prenatal diagnosis and NIPS were relatively lower.
The cost of these models is sensitive to variation in uptake in invasive testing and NIPS. The high risk cut-off and intermediate risk cut-off used in the contingent models both had an impact on overall cost and effectiveness. It is helpful to compare economic evaluations on different contingent screening models, costs and assumptions. Consequently, our study found that contingent screening with different cut-offs could offer more cost-effective  www.nature.com/scientificreports www.nature.com/scientificreports/ models, which is consistent with international studies in Belgium, the Netherlands and the UK [25][26][27] . In line with the Italian report, we consider that contingent screening using conventional CFTS and second trimester screening tests is effective 28 . The cost of each policy as a function of Down syndrome case diagnosed seems the best criterion to substantiate this hypothesis: CFTS is the best policy for the selection of patients to provide NIPS.

conclusions
We describe the performance of contingent screening models using a range of increasingly sensitive traditional screening risk cut-offs from the retrospective of the Southwestern China public health system. Our study demonstrated if only one traditional screening test is allowed to screen pregnant women, combined first-trimester screening is recommended as the first choice. The combined first-trimester screening can reduce subsequent the number of NIPS, thus reducing the cost of health economics. The findings of the present study confirm that NIPS contingent on the results of combined first-trimester screening with intermediate risk of 1:51~1:1500 is a cost-effective means of screening model for fetal Down syndrome in Chinese populations.
To the best of our knowledge, the present study is the first in China to assess the combinations of traditional screening and NIPS and has compared which traditional screening using an appropriate intermediate risk the most suitable first-line screening test approach to NIPS. In view of the fact that the whole nation couldn't be covered by only one screening method because of the different detection ability and economic situation among different regions of China, our study offer clinicians clues to take a right choice.

Limitation
This study was limited to a public health system perspective across the duration of strategies of screening and diagnosis for fetal Down syndrome. The incidence of Down syndrome was biased among the groups, because an unknown proportion of fetuses might miscarry without fetal tissue diagnosis.

Materials and Methods
Study design. This was a retrospective analysis of singleton pregnant women underwent combined first-trimester screening and quadruple screening, triple screening, double screening of second trimester. We collected and analyzed the results of combined first-trimester screening and the three screening tests of second trimester and there pregnancy outcomes, then we analyzed the performance of the four traditional screening tests. On this basis, it was assumed that the detection rate and false positive rate of NIPS were consistent with those reported, that meant detection rate of 99.5% and false positive of 0.5% 4 . The uptake rates of NIPS and invasive prenatal diagnosis were assumed to be 100%. The 9 contingent models were combined by high risk cut-offs of 1:50, 1:100, 1:270 and intermediate risk cut-offs of 1:1000, 1:1500 and 1:2000. And then, analyzed the cost and performance of various screening models when adopted to contingent screening model at different cut-offs, Fig. 1.
This study was conducted at Prenatal Diagnosis Center of West China Second University Hospital from January 2011 to December 2017. The study has been approved by the Institutional Ethics Committee of Sichuan University and all participants signed written informed consent prior to the test. The research was conducted in accordance with the relevant guidelines and clinical norms 29,30 , the details were as flows. The inclusion criteria were a maternal age of 16 years or older, pregnancy with a singleton live fetus, and the gestational age of 11 weeks through 13 weeks 6 days in the first trimester, 15 weeks through 20 weeks 6 days in the second trimester. Women were excluded from the study if they had a family history of chromosomal abnormalities or congenital malformations, medical and surgical diseases during pregnancy, twins or multiple pregnancies, and one of the twins disappearing. Fetal chromosome status was determined by amniocentesis; or by tissue sampling in cases of spontaneous pregnancy loss, pregnancy termination, or stillbirth; or by telephone follow-up results of pregnant women who did not receive amniocentesis 30 . The pregnant women were followed up by telephone six months after the pre delivery period to inquire about the pregnancy outcome and whether the fetus or newborn was normal 30 , cross-linked with the Sichuan Prenatal Diagnosis Information Network.
Traditional screening. The combined first-trimester screening risk was calculated from measurements of nuchal translucency and two serum markers, pregnancy-associated plasma protein A (PAPP-A) and the free beta subunit of human chorionic gonadotropin (fβhCG), together with maternal age. The second trimester risk of quadruple screening was calculated from measurements of serum alpha-fetoprotein(AFP), the free beta subunit of human chorionic gonadotropin (fβhCG), unconjugated estriol and inhibin A, together with maternal age. The www.nature.com/scientificreports www.nature.com/scientificreports/ second trimester risk of triple screening was calculated from measurements of serum alpha-fetoprotein(AFP), the free beta subunit of human chorionic gonadotropin (fβhCG) and inhibin A, together with maternal age. The second-trimester risk of double screening was calculated from measurements of serum alpha-fetoprotein(AFP) and the free beta subunit of human chorionic gonadotropin (fβhCG), together with maternal age 31 .
Measurements of biochemical markers were converted into multiples of the median (MoM) for gestational age, adjusted for maternal weight, insulin-dependent diabetes mellitus, status of smoking and race. Biochemical markers MoM values were center-specific. The risk of fetal Down syndrome was estimated by multiplying the maternal age-specific odds of the live birth of an infant affected by Down syndrome by the likelihood ratio obtained from the overlapping Gaussian distributions of affected and unaffected pregnancies.
Detection of serum alpha-fetoprotein (AFP), the free beta subunit of human chorionic gonadotropin (fbhCG), unconjugated estriol and pregnancy-associated plasma protein A(PAPP-A) by using reagents and instruments of Perkin Elmer (USA). Detection of inhibin A by using reagents and instruments of Beckman Coulter (USA). Using Lifecycle (Perkin Elmer, USA) to calculate risk of combined first-trimester screening and double screening and use Prenatal Screening Software (TCSoft,China) to calculate risk of quadruple screening and triple screening.
Cost of health service. A cost-effectiveness analysis was designed to evaluate contingent screening model in a retrospective traditional screening tests. In this contingent strategy, the follow-up to high risk further screening using amniocentesis and the follow-up to intermediate risk further screening using NIPS. As a result, the capability of intermediate risk will be reduced by 0.5% since the detection rate of NIPS screening is not 100% but rather 99.5% 4 . In China, the price of health services was obtained based on the charges set by the local government. This study treated the price of health service as cost because it was paid by the medical insurance and pregnant women. The cost comes from actual charge price and references [32][33][34] .
Direct medical costs included the costs of NIPS, traditional screening tests, amniocentesis, and social costs of missed detection for DS. Direct nonmedical costs and non-direct costs were not included in this study. Costs in Chinese yuan were converted into USD at the average of 2015-2017 exchange rate of 6.50 yuan= USD 1.00, Table 5.
Statistical analysis. Descriptive data were presented as median (interquartile range (IQR)) for continuous variables and as n (%) for categorical variables. Comparisons between groups were performed using Kruskal-Wallis H test or Fisher's exact test for categorical variables. The statistical software package SPSS 23.0 (SPSS Inc., Chicago, IL, USA) was used for data analyses.
invasive diagnostic test b 385 missed detection for DS 184615 Table 5. The cost of health service. a The cost of combined first-trimester screening included measurements of PAPPA, freeβhCG and nuchal translucency; b The cost of invasive diagnostic test included amniocentesis and analysis of fetal karyotype.