Integrating Combined First Trimester Screening for Preeclampsia into Routine Ultrasound Examination.

Introduction The Fetal Medicine Foundation (FMF) London has developed a first trimester screening algorithm for preeclampsia (PE), based on maternal characteristics and past risk factors, mean arterial pressure (MAP), uterine artery pulsatility index (UtA-PI), and placental growth factor (PlGF). The aim of this study was to determine the feasibility of integrating PE screening into routine practice. Material and Methods All pregnancies with a fetal crown-rump length of 45 - 84 mm presenting to our ultrasound department between January 2014 and September 2020 were included in this analysis. Screening for PE was offered to singleton pregnancies only. The number of screening tests performed in the eligible population was assessed and the reasons for missed screenings identified with the help of the electronic clinical database. SPSS Statistics 25 and GraphPad version 8.0 for Windows were used for statistical analysis. Results 6535 pregnancies were included, 4510 (69.0%) of which were screened for PE. The percentage of patients being offered PE screening increased over the years from 63.1 to 96.7% (r s  = 0.96; p = 0.003), while the rate of screenings performed in eligible patients remained stable at a median [range] of 86.2% [78.0 - 91.8%] (p = ns). 2025 (31.0%) pregnancies were not screened for PE, 1306 (64.5%) because they were not eligible for screening. 145 (2.2%) women explicitly declined PE screening; their background risk was lower than that of women who accepted screening. Conclusion Our study shows that integration of PE screening into the routine first trimester ultrasound scan is feasible and widely accepted by pregnant women and health care providers. The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commecial purposes, or adapted, remixed, transformed or built upon. ( https://creativecommons.org/licenses/by-nc-nd/4.0/ ).

Introduction

Preeclampsia (PE) is a multisystem disorder that affects 2 – 7% of all pregnancies and is associated with short- and long-term risks for both mother and child 1  –  4 . Delivery is still the only treatment for PE available today, however prevention is possible in high-risk pregnancies if low dose aspirin (LDA) treatment is started before 16 weeks of gestation 5 . Screening by maternal characteristics and a previous history of risk factors as proposed by many obstetric societies detects around 30% of all preterm PEs (< 37 weeks) 6  –  12 . The Fetal Medicine Foundation (FMF) London has developed a first trimester screening algorithm combining background risk factors with placental growth factor (PlGF), mean arterial pressure (MAP) and uterine artery pulsatility index (UtA-PI). This screening identifies 75% of preterm PE cases 6 ,  7 ,  13 . The ASPRE trial (Aspirin for Evidence-Based Preeclampsia Prevention) confirmed the effectiveness of LDA in patients identified by this screening to be at risk and demonstrated a reduction of preterm PE by 62% in women who screened positive 14 .

Offering first trimester screening for aneuploidies is widespread and, in many countries, it is part of routine pregnancy care 15 ,  16 . In addition to screening for fetal aneuploidies, first trimester ultrasound also allows the detection of various structural fetal anomalies, and therefore first trimester screening has evolved to become the most important ultrasound exam in pregnancy 16 .

The introduction of first trimester PE screening into routine pregnancy care, however, is still an ongoing process, hampered by issues such as practicability and costs. While several validation studies and, more recently, implementation studies have been published which confirm the performance of the combined PE screening algorithm, little information is yet available about the performance of first trimester PE screening in a general clinical setting 10 ,  12 ,  17 . In particular, there are no studies regarding its feasibility and womenʼs acceptance of introducing this PE screening algorithm into the first trimester ultrasound scan. A subanalysis of the ASPRE trial investigated the reasons for accepting or declining PE screening. In this randomized controlled study, 6.8% of women declined participation for various reasons such as insufficient information and discouragement by significant others (partner, family, health professional) 18 . Reasons to participate were personal benefit, small risk, sufficient information, trust in professionals, and encouragement by others 18 .

First trimester screening for fetal anomalies and aneuploidies has been offered to pregnant women for nearly two decades at our department. In 2014, we introduced combined first trimester screening for PE along with the routine first trimester ultrasound. The screening test mostly performs according to expectations in our cohort as we could demonstrate in a recently accepted publication 19 . The aim of this study was to investigate the feasibility of introducing PE screening into routine first trimester ultrasound in clinical practice.

Patients and Materials

Screening methods

We performed an observational cohort study with a prospective analysis of retrospective data. All women who attended the Center for Ultrasound and Prenatal Diagnosis of the Department of Obstetrics and Feto-maternal Medicine at University Hospital of Bern for their 11 – 14 weeks scan between January 2014 and September 2020 and agreed to the further use of their data were included in this study. All women with a singleton or multiple pregnancy and a fetal crown-rump length of 45 – 84 mm are routinely offered early morphological screening and first trimester combined screening for aneuploidies.

First trimester screening for aneuploidies is performed by combining maternal age with fetal nuchal translucency (NT) and the biochemical parameters “pregnancy-associated plasma protein A” (PAPP-A) and “beta human chorionic gonadotropin” (β-HCG) 20 ,  21 . All women are informed about the possibility of having additional non-invasive prenatal testing (NIPT) or an invasive procedure. As of July 2015, NIPT in the form of sequential screening for all women in whom the combined first trimester aneuploidy screening test yields a risk for trisomies ≥ 1 : 1000 is covered by health insurance in Switzerland 22 . Invasive procedures are proposed to women who have a high risk at combined aneuploidy screening (individual decision, covered by insurance if the risk in the first trimester combined screening is > 1 : 380), a NT > 3.5 mm or if fetal anomalies are detected during the scan 22 . In addition, for families at risk of an inherited genetic condition, invasive testing is offered after genetic counseling.

First trimester combined screening for preeclampsia (PE) has been offered to singleton pregnancies since January 2014 during the same visit. Screening for PE is performed using the algorithm developed by the Fetal Medicine Foundation (FMF) London which combines maternal characteristics and a previous history of risk factors with placental growth factor (PlGF), mean arterial pressure (MAP) and mean uterine artery pulsatility index (UtA PI). PAPP-A is included in the screening algorithm if values are available from combined screening for aneuploidies. However, if only PE screening is requested, PAPP-A is not included. Women considered at risk are prescribed LDA. Both the risks for trisomies and for PE are calculated using the algorithm provided by Viewpoint version 5.6.25.284.

Patient selection

In this analysis, we first identified the patients eligible for screening. While we introduced PE screening for patients attending our department for antenatal care from January 2014, we only offered it to patients referred from external gynecologists after publication of the ASPRE trial. Women referred before July 2017 as well as all multiple pregnancies were therefore considered not eligible for PE screening in this study. We included them in the study to assess the overall screening rate for PE. However, when analyzing the rates of acceptance, we only included the eligible group.

Combined screening for aneuploidies was not repeated if it was previously performed by a private provider and the woman was referred to our center for second opinion, and it was not offered if an invasive procedure was indicated or NIPT was performed prior to the scan. All other pregnancies were considered eligible for first trimester combined aneuploidy screening in this trial. Screening tests that were incomplete due to one or several missing parameters were included in this analysis and considered performed if the result was communicated to the patient.

To assess the acceptance of screening, we analyzed the percentage of screening tests performed in the eligible population. In pregnancies not screened for aneuploidies by combined screening in the first trimester and not screened for PE or for both, we analyzed the reasons for screenings not being performed by studying the screening records of these patients in our electronic clinical database.

Data analysis

Data collection and analysis was performed using our encrypted medical database, and information was saved using Viewpoint ® software. Statistical analyses were performed with GraphPad ® version 8.0 for Windows (GraphPad Software, San Diego, CA, USA) and the SPSS ® statistical software package. Continuous variables were analyzed with Kruskal-Wallis or Mann-Whitney U-test, and proportions were evaluated using Fisherʼs exact test or χ 2 test where appropriate. Correlations were searched for using the Spearman rank correlation test. Statistical significance was considered to be achieved when p was below 0.05.

The study was approved by the Ethics Committee of the Canton of Bern.

Results

During the study period, 6535 women with 6906 fetuses attended our clinic for first trimester ultrasound screening. Patient characteristics are depicted in Table 1 .

Table 1  Characteristics of the study population. Comparisons are made between pregnancies which had combined screening for aneuploidies and the total study population and between pregnancies which had PE screening and the total study population. They are depicted in absolute numbers (N) and interquartile ranges [IQR] or as percentages (%). Additionally, the significance between pregnancies which had combined screening for aneuploidies and the total study population as well as the significance between pregnancies which had PE screening and the total study population are depicted in the third and fifth columns.

	Total (N = 6535)	Aneuploidy screening (N = 5401)	p †	PE screening (N = 4510)	p ‡
Figures are given as median and [interquartile ranges], figures in parentheses are percentages; IVF: in vitro fertilization.* Relevant medical history includes diabetes mellitus, hypertension, systemic lupus erythematosus, antiphospholipid syndrome. p < 0.05 is considered statistically significant; p † : Significance between pregnancies which had an aneuploidy screening and the total study population; p ‡ : significance between pregnancies which had an PE screening and the total study population.
Maternal age at term (median [IQR])	33 [29 – 37]	33 [29 – 36]	n. s.	33 [29 – 36]	p = 0.005
Maternal BMI (median [IQR])	23.2 [20.9 – 26.5]	23.2 [20.9 – 26.5]	n. s.	23.3 [20.9 – 26.6]	n. s.
Gravidity (median [IQR])	2 [1 – 3]	2 [1 – 3]	n. s.	2 [1 – 3]	n. s.
Parity
Nulliparous Parous, no previous PE Parous, previous PE	3205/6388 (50.2)3015/6388 (47.2)168/6388 (2.6)	2738/5312 (51.5)2429/5312 (45.7)145/5312 (2.7)	n. s.n. s.n. s.	2281/4450 (51.3)2013/4450 (45.2)156/4450 (3.5)	n. s.p = 0.046p = 0.010
Ethnicity
Caucasian Other	4994/6007 (83.1)1013/6007 (16.9)	4393/5337 (82.3)944/5337 (17.7)	n. s.n. s.	3655/4506 (81.1)851/4506 (18.9)	p = 0.007p = 0.007
Smoking	480/6282 (7.6)	419/5369 (7.8)	n. s.	340/4510 (7.5)	n. s.
Medical conditions*	196/5352 (3.6)	183/4759 (3.8)	n. s.	175/4491 (3.9))	n. s.
Conception by IVF	371/6370 (5.8)	323/5385 (6.0)	n. s.	243/4510 (5.4)	n. s.
Family history of PE	68/5282 (1.3)	63/4725 (1.3)	n. s.	65/4490 (1.4)	n. s.
Multiple pregnancy	331/6535 (5.1)	274/5401 (5.1)	n. s.	0/4510 (0.0)	p < 0.0001

PE screening

5225 (80.0%) of all included pregnancies were considered eligible for PE screening. Screening for PE was performed in 4510 (69.0%) pregnancies, with a median [range] of 585 [500 – 814] screenings performed per year. The percentage of patients eligible for PE screening increased over the years from 63.1 to 96.7% (r s  = 0.96; p = 0.003), as did the absolute number of tests performed per year from 2014 to 2019 ([500 – 814], r s  = 0.93; p = 0.022). The rate of PE screening performed in eligible patients remained stable at a median [range] of 86.2% [78.0 – 91.8%] (p = n. s.) ( Fig. 1 a ). The different reasons why no PE screening was performed in the remaining 2025 (31.0%) pregnancies are given in Table 2 . As regards the 145 (2.2%) patients who declined screening for PE, no significant change in these figures was noted over the years (15 – 31 per year, p = n. s.). 110 (75.9%) of them also declined combined screening for aneuploidies while 35 (24.1%) agreed to aneuploidy screening only.

Fig. 1

 Uptake of PE screening and combined screening for aneuploidies over the years. The numbers in the bars refer to absolute numbers and they are also displayed as percentages for each year. The rates of all women eligible for PE or aneuploidy screening who had a screening are included in red in both figures.

Table 2  Differential analysis of the 2025 patients who did not have a PE screening. They are depicted in absolute numbers (N) and as percentages (%). Additionally, the percentage of the total study population of 6535 pregnancies is depicted in the second column.

Reason	N (%)	% of total population
Figures are in numbers and percentages. LDA: low dose aspirin; OBGYN: obstetrician and gynecologist
Not eligible (not offered)	1306 (64.5)	19.9
External referral before 06/2017 Multiple pregnancy	977 (48.3)329 (16.3)	14.95.0
Eligible	719 (35.5)	11.0
2nd opinion/pathologies Language difficulties LDA started before screening Refused by referring OBGYN Refused by patient No explanation	293 (14.5)11 (0.5)27 (1.3)28 (1.4)145 (7.2)215 (10.6)	4.50.20.40.42.23.3
Total	2025 (100)	31.0

Combined screening for aneuploidies

6070 (92.9%) pregnancies were eligible for combined screening for aneuploidies. Screening was performed in 5401 (89.0%) of them, which corresponds to a median [range] of 743 [684 – 914] tests per year. The rate of women accepting aneuploidy screening when offered remained stable during the study period at a median [range] of 90.9% [82.8 – 90.9%] (p = n. s.) ( Fig. 1 b ). 575 (9.5%) women eligible for screening declined first trimester aneuploidy screening; 460 (80%) of them accepted measurement of nuchal translucency without calculation of the combined risk for trisomies. 243 (42.3%) women who declined aneuploidy screening accepted PE screening.

Subanalysis of women declining PE screening

Characteristics of the women who declined PE screening compared to those who accepted screening are given in Table 3 . Women who declined PE screening had a lower risk profile with regard to their BMI, their obstetric history, mode of conception, and ethnicity. Other risk factors linked to previous medical history did not differ between the women who refused and those who accepted PE screening.

Table 3  Characteristics of the populations who declined vs. accepted screening for PE. They are depicted in absolute numbers (N) and as percentages (%) or interquartile ranges [IQR] where applicable. In the third column, the significance between patients who accepted and those who declined PE screening with regard to specific maternal characteristics are depicted.

	PE screening refused (N = 145)	PE screening accepted (N = 4510)	p
Figures are given as medians [IQR]; the figures in parentheses are percentages; IQR: interquartile ranges, TOP: termination of pregnancy, PE: preeclampsia, IVF: in vitro fertilization, n. s.: not significant.* Relevant medical history includes diabetes mellitus, hypertension, systemic lupus erythematosus, antiphospholipid syndrome.p < 0.05 is considered significant.
Maternal age at term (median [IQR])	33 [30 – 36]	33 [29 – 36]	n. s.
Maternal BMI at 12 weeks (median [IQR])	22.2 [20.0 – 24.3]	23.3 [20.9 – 26.6]	p = 0.0009
Nulliparous	52/139 (37.4)	2281/4450 (51.3)	p = 0.0014
previous miscarriage/TOP	10/139 (7.2)	564/4450 (12.7)	n. s.
Parous with previous PE	0/139 (0.0)	156/4450 (3.5)	p = 0.015
Ethnicity
Caucasian Black/South Asian	91/102 (89.2)8/102 (7.8)	3655/4506 (81.1)557/4506 (12.4)	p = 0.039n. s.
Smoking	5/123 (0.4)	340/4510 (7.5)	n. s.
Medical conditions*	0/83 (0.0)	175/4491 (3.9)	n. s.
Conception by IVF	2/132 (1.5)	243/4510 (5.4)	p = 0.047
Family history of PE	0/83 (0.0)	65/4490 (1.4)	n. s.

Discussion

The results of our study demonstrate that introducing PE screening into routine practice as part of first trimester ultrasound screening is feasible and very much accepted by pregnant women and doctors. The test was performed in 86.2% of all women considered eligible for PE screening, a figure that is comparable to the 90.9% of all women eligible for aneuploidy screening who accepted such screening. While nearly 10% of all women declined screening for aneuploidies because of the lack of consequences, only 2.2% of all women explicitly declined screening for PE ( Table 2 ). Acceptance amongst healthcare providers is also high; few referring doctors explicitly refuse PE screening for their patient ( Table 2 ).

PE screening was not performed in 11% of all pregnancies, mostly either due to referral for a second opinion regarding a fetal problem or for no obvious reason ( Table 2 ). Our interpretation is that the involved physicians focused on the reasons for the referral and less on general screening options. However, pregnancies with increased nuchal translucency, increased risk of aneuploidies at screening and diagnosed genetic anomalies such as trisomy 13 are found to be at increased risk of preeclampsia if the pregnancy is continued 23 ,  24 ,  25 ,  26 . Another explanation could be that some doctors are less convinced that PE screening is useful. As demonstrated in the SPREE trial, even though 10.3% of all patients were screen-positive using the NICE method, only 4.5% were treated with LDA, despite the national guidelines 11 . Therefore, in future we should focus on offering PE screening to women with pregnancies showing abnormalities at aneuploidy screening or on the occasion of the first trimester anomaly scan. This also emphasizes the importance of patient counseling and the education of health professionals, as their attitude towards diagnostic tools may be reflected in their counseling and hence affect the decision taken by the patient.

The subanalysis of patients who declined PE screening in our cohort demonstrated a lower background risk than for women who accepted screening. Women who declined screening had a lower BMI, which correlates with a lower risk for PE 27 ,  28 . Maternal age is a risk factor for PE and trisomies; while advanced maternal age leads to an increased acceptance of aneuploidy screening, no such association was found for PE screening 29 ,  30 . This can be explained by the fact that the screening rate for PE decreased with higher parity, which is linked to maternal age. In addition, more nulliparous women accepted screening, possibly because nulliparity was perceived as a risk factor, but possibly also because nulliparous women generally accept more screening tests 29 ,  31 . Only two women who conceived with IVF declined screening and all women with a medical risk factor accepted screening. We consider this finding the result of patients being well-informed, as described by Möller et al. 32 .

Availability of a screening test is a crucial factor influencing the uptake of screening. Although there are inherent differences between aneuploidy screening and PE screening, mainly regarding the possibility of prevention and treatment of the condition screened for, it is worth taking a comparative look at the uptake of first trimester aneuploidy screening, which has been extensively studied in the past. While in the UK the uptake of aneuploidy screening increased to 75% once the test was introduced into routine care, it has remained much lower in the Netherlands where screening is only offered to women older than 36 years of age 15 ,  16 ,  31 . In Scandinavian countries like Denmark and Iceland, aneuploidy screening is part of routine care and the uptake is at least 90%, comparable to our setting and the uptake of aneuploidy screening in our cohort 33 . Trials investigating the acceptance of non-invasive prenatal testing (NIPT) with cell-free fetal DNA mostly confirmed that uptake was clearly related to reimbursement by insurance companies or the national health system 15 ,  34 . Other factors affecting the uptake of aneuploidy screening include counseling, the level of maternal education and trust in the healthcare provider as well as test-specific reasons such as maternal age, religious beliefs and the availability and acceptance of termination of pregnancy 29 ,  30 ,  32 ,  35  –  38 . To date, there are no studies published which have investigated the acceptance of first trimester combined screening for PE in the general population. One can only assume that some factors mentioned above also influence the uptake of PE screening.

Women who generally refuse any screening in pregnancy may not book an appointment for the first trimester ultrasound. However, the vast majority of pregnant women in Switzerland do, in fact, opt for first trimester ultrasound screening. Therefore, the acceptance rate in our cohort can in part be expanded to the general Swiss population. While the majority of women in our cohort who declined PE screening also declined aneuploidy screening and seemed therefore more critical of screening tests in general, many women who declined aneuploidy screening accepted PE screening. The reason for declining aneuploidy screening is mostly the fact that a positive screening test would have no consequence for the women as they would accept a child with chromosomal abnormalities. In addition, 80% of these women accepted the assessment of fetal nuchal translucency as a marker for other fetal anomalies apart from chromosomal abnormalities.

Conclusion

In conclusion, our study demonstrates that the integration of screening for PE using the FMF algorithm into first trimester ultrasound screening is feasible and accepted by pregnant women as well as healthcare providers. Counseling of all pregnant women on the issue of preeclampsia early in pregnancy is important, including the possibility of PE screening and prevention as well as symptoms and risks, as many women are not familiar with this disease.

Sources of Funding

None.