[Exploration of factors for failure of non-invasive prenatal testing based on Logistic regression analysis].

OBJECTIVE

To explore the factors for the failure of non-invasive prenatal testing (NIPT) through multifactorial unconditional Logistic regression analysis.

METHODS

A total of 3 410 pregnant women who had visited Dalian Women and Children Medical Group from July 2019 to June 2020 were selected as the study subjects and divided into first success NIPT group (n = 3 350) and first failed group (n = 60). Clinical data including age, weight, body mass index (BMI), gestational week, type of pregnancy (singleton/twin), history of delivery, heparin treatment, and conception method [natural conception/assisted reproductive technology (ART)] were collected. Independent sample t-test and Chi-square test were carried out for comparing the two groups, and multi-factorial unconditional Logistic regression analysis was carried out to explore the factors for the failure of NIPT, and receiver operating characteristic curve (ROC) analysis was used to evaluate the diagnosis and predictive effects.

RESULTS

Among the 3 410 pregnant women, 3 350 were assigned to the first success NIPT group, and 60 were assigned to the first failed group, and the first-time failure rate was 1.76% (60/3 410). No significant difference was found in age, weight, BMI and method of conception between the two groups (P > 0.05). Compared with first success group, first failed group had lower sampling gestational weeks, lower proportion of women with previous history of delivery, and higher proportion of twin pregnancies and heparin treatment (P < 0.05). Multi-factorial unconditional Logistic regression analysis indicated that sampling gestational week (OR = 0.931, 95%CI: 0.845 ~ 1.026, P < 0.001) and history of heparin use (OR = 8.771, 95%CI: 2.708 ~ 28.409, P < 0.001) are independent factors for first failed NIPT. One-way unconditional Logistic regression analysis for sampling gestational weeks indicated that the regression equation for NIPT screening failure was Logit (P) = -9.867 + 0.319 × sampling gestational week, with the area under the ROC curve being 0.742, a Jordan index of 0.427, and a cutoff value of 16.36 weeks.

CONCLUSION

Gestational week and heparin treatment are independent factors for the first failed NIPT. A regression equation has been established and determined the optimal sampling gestational week to be 16.36 weeks, which may provide a reference for the timing of NIPT screening.