Second Trimester Screening of Preeclampsia Using Maternal Characteristics and Uterine and Ophthalmic Artery Doppler

Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review

PURPOSE OF REVIEW

Machine learning (ML) approaches are an emerging alternative for healthcare risk prediction. We aimed to synthesise the literature on ML and classical regression studies exploring potential prognostic factors and to compare prediction performance for pre-eclampsia.

RECENT FINDINGS

From 9382 studies retrieved, 82 were included. Sixty-six publications exclusively reported eighty-four classical regression models to predict variable timing of onset of pre-eclampsia. Another six publications reported purely ML algorithms, whilst another 10 publications reported ML algorithms and classical regression models in the same sample with 8 of 10 findings that ML algorithms outperformed classical regression models. The most frequent prognostic factors were age, pre-pregnancy body mass index, chronic medical conditions, parity, prior history of pre-eclampsia, mean arterial pressure, uterine artery pulsatility index, placental growth factor, and pregnancy-associated plasma protein A. Top performing ML algorithms were random forest (area under the curve (AUC) = 0.94, 95% confidence interval (CI) 0.91-0.96) and extreme gradient boosting (AUC = 0.92, 95% CI 0.90-0.94). The competing risk model had similar performance (AUC = 0.92, 95% CI 0.91-0.92) compared with a neural network. Calibration performance was not reported in the majority of publications. ML algorithms had better performance compared to classical regression models in pre-eclampsia prediction. Random forest and boosting-type algorithms had the best prediction performance. Further research should focus on comparing ML algorithms to classical regression models using the same samples and evaluation metrics to gain insight into their performance. External validation of ML algorithms is warranted to gain insights into their generalisability.

Performance of HDP-Gestosis Score and Ophthalmic Artery Doppler in Prediction of Pre-eclampsia

Purpose of Study

To study the predictive accuracy of maternal characteristics, mean arterial pressure, uterine artery doppler and maternal ophthalmic artery doppler in second trimester for subsequent development of pre-eclampsia.

Methods

A prospective cohort study of 440 women at 19-24 weeks' gestation. It included recording of maternal demographic characteristics and calculation of HDP Gestosis Score, measurement of MAP and ultrasound evaluation for fetal anatomy, Uterine artery doppler and maternal ophthalmic artery doppler. The mean Peak Systolic Velocity ratio was obtained by taking two sets of reading from both eyes. The statistical analysis was done using independent t-test for quantitative variables and chi-square test, Fisher's exact test for qualitative variables and Area Under Curve was obtained at 10% False Positive Rate.

Results

Among 440 pregnant women, 43(10.8%) developed Hypertensive Disorder of Pregnancy (8 early onset PE, 16 late onset PE and 19 GHTN) and 42 were lost to follow up, with an incidence rate of 10.8%. Combination of Gestosis Score + OAD PSV Ratio had highest AUC of 0.73. Whereas combination of Gestosis Score + MAP + UtAPI + OAD-PSV Ratio had highest sensitivity of 97.67%. The OAD PSV ratio improved detection rate of Gestosis Score (from 90 to 100%) for prediction of development of PE, especially Preterm PE at 10% FPR.

Conclusion

Maternal Ophthalmic artery doppler study significantly improved prediction of development of PE at 19-24 weeks' gestation, both independently as well as in combination with HDP-Gestosis Score, MAP and uterine artery doppler.

Alternatives of Risk Prediction Models for Preeclampsia in a Low Middle-Income Setting

Abstract Objectives:  To  develop prediction models for the first-trimester prediction of PE (PE) using the established biomarkers including maternal characteristics and history, mean arterial pressure (MAP), uterine artery Doppler pulsatility index (UtA-PI ), and Placental Growth Factor (PlGF)) in combination with Ophthalmic artery Doppler peak ratio (PR). Methods:  This was a  prospective observational study in women attending a first-trimester screening at 11-14 weeks’  gestation. Maternal characteristics and history, measurement of MAP, ultrasound examination for UtA-PI measurement, maternal ophthalmic PR Doppler measurement, and serum  PlGF collection were performed during the visit. Logistic regression  analysis was used to determine if the maternal factor had a significant contribution in predicting PE. The Receiving Operator Curve (ROC) analysis was used to determine the area under the curve  (AUC), positive predictive value  (PPV), negative prefictive value (NPV) and positive screening cut-off in predicting the occurrence of PE at any gestational age.   Results: Of the 946 eligible participants, 71 (7,49%) subjects were affected by PE. Based on the ROC curves, optimal high-risk cutoff value for prediction of preeclampsia at any gestational age for model 2 (primary care model) in this Indonesia study population were 63% with the sensitivity and specificity of 71.8% and 71.2%, respectively. Both sensitivity and specificity for model 3 (complete model) were 70.4% and 74.9%, respectively for the cutoff value 58%. The area under the curve of model 2, model 3 was 0.7651 (95% CI: 0.7023-0.8279)) and 0.7911 (95% CI: 0.7312-0.8511), respectively, for predicting PE. In addition, PPV and NPV for model 2 were 16.8% and 96.9%, respectively. PPV and NPV for model 3 were 18.55 and 96.9%, respectively.   Conclusion: The prediction models of preeclampsia vary depending upon healthcare resource. Complete model is clinically superior to primary care model but it is not statistically significant.  Prognostic models should be easy to use, informative and low cost with great potential to improve maternal and neonatal health in Low Middle Income Country settings. 

Ophthalmic artery Doppler in the prediction of preeclampsia

The ophthalmic artery is an easily accessible vessel for Doppler assessment that provides information on the less accessible intracranial circulation. In women with preeclampsia, compared with normotensive pregnant women, there is a decrease in impedance to flow and an increase in velocities in the flow velocity waveforms from the ophthalmic arteries. This study aimed to report the methodology for ophthalmic artery Doppler and summarize findings from the clinical implementation of such assessment in the prediction of preeclampsia. The Embase and MEDLINE were searched from inception to November 2020 to identify studies reporting on the use of ophthalmic artery Doppler in the prediction of preeclampsia. Of note, 2 small studies in high-risk pregnancies, one at 11 to 14 weeks' gestation and another at 20 to 28 weeks' gestation, reported differences between women who developed preeclampsia, compared with unaffected pregnancies, in ophthalmic artery Doppler and suggested that this is a useful biomarker for screening for preeclampsia. Another small study in high-risk pregnancies at 18 to 23 weeks' gestation reported that there was no marked difference in ophthalmic artery Doppler indices between the preeclampsia and unaffected groups. In addition, 2 recent, large observational studies in unselected pregnancies at 19 to 23 and 35 to 37 weeks' gestation, respectively, reported that, first, it is necessary to record waveforms from both eyes to get reproducible results; second, the waveform from the ophthalmic arteries is characterized by 2 systolic peaks and the ratio of the second to the first peak systolic velocity was increased in women who developed preeclampsia; third, in the study at 19 to 23 weeks' gestation, the peak systolic velocity ratio was superior to the uterine artery pulsatility index, mean arterial pressure, serum placental growth factor, and soluble fms-like tyrosine kinase-1 as individual biomarkers in the prediction of both preterm and term preeclampsia and the peak systolic velocity ratio improved the prediction of preeclampsia provided by all the other biomarkers; and fourth, in the study at 35 to 37 weeks' gestation, the peak systolic velocity ratio improved the prediction of subsequent development of preeclampsia provided by maternal factors alone and combinations of maternal factors with mean arterial pressure, uterine artery pulsatility index, placental growth factor, and serum placental growth factor. The ophthalmic artery Doppler provides a useful biomarker for the prediction of preeclampsia.

Ophthalmic artery Doppler in combination with other biomarkers in prediction of pre-eclampsia at 35-37 weeks' gestation

OBJECTIVE

To examine the potential value of maternal ophthalmic artery Doppler at 35-37 weeks' gestation in combination with the established biomarkers of pre-eclampsia (PE), including mean arterial pressure (MAP), uterine artery pulsatility index (UtA-PI), serum placental growth factor (PlGF) and serum soluble fms-like tyrosine kinase-1 (sFlt-1), in the prediction of subsequent development of PE.

METHODS

This was a prospective observational study in women attending for a routine hospital visit at 35 + 0 to 36 + 6 weeks' gestation. This visit included recording of maternal demographic characteristics and medical history, ultrasound examination for fetal anatomy and growth, assessment of flow velocity waveforms from the maternal ophthalmic arteries, and measurement of MAP, UtA-PI, serum PlGF and serum sFlt-1. The competing-risks model was used to estimate the individual patient-specific risks of delivery with PE at any time and at < 3 weeks after assessment by a combination of maternal demographic characteristics and medical history with biomarkers. The area under the receiver-operating-characteristics curve and detection rate (DR) of delivery with PE, at a 10% false-positive rate (FPR), in screening by combinations of maternal factors with ophthalmic artery second to first peak of systolic velocity ratio (PSV ratio), MAP, UtA-PI, serum PlGF and serum sFlt-1 were determined. The modeled performance of screening for PE was also estimated.

RESULTS

The study population of 2287 pregnancies contained 60 (2.6%) that developed PE, including 19 (0.8%) that delivered with PE at < 3 weeks after assessment. The PSV ratio improved the prediction of PE with delivery at any stage after assessment provided by maternal factors alone (from 25.4% to 50.6%), maternal factors and MAP (54.3% to 62.7%), maternal factors, MAP and PlGF (68.3% to 70.8%) and maternal factors, MAP, PlGF and sFlt-1 (75.7% to 76.7%), at a FPR of 10%. The PSV ratio also improved the prediction of PE with delivery at < 3 weeks after assessment provided by maternal factors alone (from 31.0% to 69.4%), maternal factors and MAP (74.1% to 83.4%), maternal factors, MAP and UtA-PI (77.1% to 85.0%) and maternal factors, MAP and PlGF (84.8% to 88.6%). The empirical results for DR at a 10% FPR were consistent with the modeled results. Screening by a combination of maternal factors with MAP and PSV ratio also detected 59.4% (95% CI, 58.6-82.5%) of cases of gestational hypertension with delivery at any stage after assessment, and 86.7% (95% CI, 82.4-100%) of those with delivery at < 3 weeks after assessment.

CONCLUSION

Ophthalmic artery Doppler could potentially improve the performance of screening for PE at 35-37 weeks, especially imminent PE with delivery within 3 weeks after assessment, but further studies are needed to validate this finding. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.

Ophthalmic artery Doppler in combination with other biomarkers in prediction of pre-eclampsia at 19-23 weeks' gestation

OBJECTIVE

To examine the potential value of maternal ophthalmic artery Doppler at 19-23 weeks' gestation on its own and in combination with the established biomarkers of pre-eclampsia (PE), including uterine artery (UtA) pulsatility index (PI), mean arterial pressure (MAP), serum placental growth factor (PlGF) and serum soluble fms-like tyrosine kinase-1 (sFlt-1), in the prediction of subsequent development of PE.

METHODS

This was a prospective observational study of women attending for a routine hospital visit at 19 + 1 to 23 + 3 weeks' gestation. This visit included recording of maternal demographic characteristics and medical history, ultrasound examination for fetal anatomy and growth, assessment of flow velocity waveforms from the maternal ophthalmic arteries, and measurement of MAP, UtA-PI, serum PlGF and serum sFlt-1. Waveforms were obtained from the ophthalmic arteries in sequence from the right eye, left eye and again from the right and then left eye. We recorded the average of the four measurements, two from each eye, for the following four indices: first peak of systolic velocity; second peak of systolic velocity; PI; and the ratio of the second to first peak of systolic velocity (PSV ratio). The measurements of the four indices were standardized to remove the effects of maternal characteristics and elements from the medical history. The competing-risks model was used to estimate the individual patient-specific risks of delivery with PE at < 37 and ≥ 37 weeks' gestation and to determine the area under the receiver-operating-characteristics curve (AUC) and detection rate (DR), at a 10% false-positive rate (FPR), in screening by a combination of maternal demographic characteristics and medical history with biomarkers. The modeled performance of screening for PE was also estimated.

RESULTS

The study population of 2853 pregnancies contained 76 (2.7%) that developed PE, including 18 (0.6%) that delivered with PE at < 37 weeks' gestation. The ophthalmic artery PSV ratio was significantly increased in PE pregnancies, and the PE effect depended on gestational age at delivery; the deviation from normal was greater for early than late PE. The second peak of systolic velocity was also increased in PE pregnancies, but the effect did not depend on gestational age at delivery. The other two ophthalmic artery indices of first peak of systolic velocity and PI were not significantly affected by PE. The PSV ratio improved the prediction of preterm PE provided by maternal factors alone (from 56.1% to 80.2%), maternal factors, MAP and UtA-PI (80.7% to 87.9%), maternal factors, MAP, UtA-PI and PlGF (85.5% to 90.3%) and maternal factors, MAP, UtA-PI, PlGF and sFlt-1 (84.9% to 89.8%), at a FPR of 10%. The PSV ratio also improved the prediction of term PE provided by maternal factors alone (from 33.8% to 46.0%), maternal factors, MAP and UtA-PI (46.6% to 54.2%), maternal factors, MAP, UtA-PI and PlGF (45.2% to 53.4%) and maternal factors, MAP, UtA-PI, PlGF and sFlt-1 (43.0% to 51.2%), at a FPR of 10%. The empirical results for DR at a 10% FPR were consistent with the modeled results. The second peak of systolic velocity did not improve the prediction of either preterm or term PE provided by maternal factors alone.

CONCLUSION

Ophthalmic artery PSV ratio at 19-23 weeks' gestation, both on its own and in combination with other biomarkers, is potentially useful for prediction of subsequent development of PE, especially preterm PE, but larger studies are needed to validate this finding. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.

Ophthalmic artery Doppler in prediction of pre-eclampsia at 35-37 weeks' gestation

OBJECTIVES

First, to examine the potential value of maternal ophthalmic artery Doppler at 35-37 weeks' gestation in the prediction of subsequent development of pre-eclampsia (PE), and, second, to examine the variability between repeat measurements in the same eye and variability in measurements between the two eyes.

METHODS

This was a prospective observational study in women attending for a routine hospital visit at 35 + 0 to 36 + 6 weeks' gestation. The visit included recording of maternal demographic characteristics and medical history and assessment of flow velocity waveforms from the maternal ophthalmic artery. Waveforms were obtained in sequence from the right eye, left eye and again from the right and then left eye. We recorded the average of the four measurements, two from each eye, for the following four indices: first peak of systolic velocity; second peak of systolic velocity; pulsatility index; and the ratio of the second to first peak of systolic velocity (PSV ratio). The measurements of the four indices were standardized to remove the effects of maternal characteristics and elements from the medical history. The competing-risks model was used to determine the detection rate (DR) of delivery with PE at any time and at < 3 weeks after assessment, at a 10% false-positive rate (FPR), in screening by maternal factors alone and a combination of maternal factors and the adjusted value of each of the four ophthalmic artery indices.

RESULTS

The study population of 2287 pregnancies contained 60 (2.6%) that developed PE, including 19 (0.8%) that delivered with PE at < 3 weeks after assessment. The DR, at 10% FPR, of delivery with PE at any time after assessment by maternal factors was 25.0% (95% CI, 14.7-37.9%), and this increased by 25 percentage points to 50.0% (95% CI, 36.8-63.2%) with the addition of the adjusted PSV ratio (P = 0.005); the respective values for delivery with PE at < 3 weeks after assessment were 31.6% (95% CI, 12.6-56.6%) and 57.9% (95% CI, 33.5-79.8%). The other ophthalmic artery indices did not improve the prediction provided by maternal factors alone. There was good correlation between the first and second measurements of PSV ratio from the same eye (right eye r = 0.823, left eye r = 0.840), but poorer correlation in the first and second measurements between the two eyes (first measurement r = 0.690, second measurement r = 0.682). In screening by maternal factors and PSV ratio for PE with delivery at any stage after assessment, the estimated DR, at 10% FPR, was 50.0% when the average of four measurements was used (two from each eye), 49.1% when the average of one measurement from each eye was used, 47.3% when the average of two measurements from the same eye was used, and 45.8% when only one measurement was used.

CONCLUSIONS

Ophthalmic artery PSV ratio at 35-37 weeks' gestation can predict subsequent delivery with PE, especially if this occurs within 3 weeks after assessment. In the assessment of ophthalmic artery Doppler, it is necessary to use the average of one measurement from each eye to minimize variability of measurements. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.