Fetal Medicine Foundation Stillbirth risk score
Table 1 shows the maternal demographic data that
synthesize the FMF Stillbirth Risk Calculator. In our study
cohort, women after stillbirth compared to matched controls were found
to have a significantly higher BMI (p= 0.002), were more
frequently nicotine consumers (p <0.001) and suffered
more frequently medical conditions, such as hypertension
(p= 0.015), APS (p= 0.017) and diabetes
(p <0.001) with higher need for insulin
(p =0.006).
Figure 1 shows the distribution of the FMFStillbirth risk score in IUFDs and matched controls. The medianFMF Stillbirth risk score in the group of stillbirths was 0.45%
(0.19-5.70%) [1:222 (1:526–1:17)], whilst the median risk score in
the group of matched live births was 0.23% (0.18-1.30%) [1:435
(1:556–1:77); p <0.001]. Figure 2 and
Figure 3 show the FMF Stillbirth risk scores in
stillborn fetuses per cause of death with and without outliers,
respectively.
To evaluate the discriminative power of the FMF Stillbirth Risk
Calculator we performed a ROC analysis and calculated an area under the
curve (AUC) of 0.72 (95% CI 0.67–0.78; p <0.001) to
predict antepartum stillbirth in the total cohort (Figure
4 ). Also after exclusion of all stillborn fetuses with congenital
anomalies, the AUC was 0.74 (95% CI 0.67-0.82;p <0.001). Assessing stillborn fetuses of unknown cause,
AUC was 0.83 (95% CI 0.72-0.93; p <0.001). Assessing
stillborn fetuses due to placental dysfunction, the AUC was 0.64 (95%
CI 0.50-0.78; p =0.053).
Univariate binary logistic regression to examine the FMFStillbirth Risk Calculator´s predictive ability resulted in an OR of
6.22 (95% CI 3.91–9.89; p <0.001) at an optimalFMF Stillbirth risk score cut-off of ≥0.34% for predicting
stillbirth with a specificity of 82% and a sensitivity of 58% in the
total cohort of IUFDs.