Introduction

Approximately 2%–3% of all newborns are affected by a congenital anomaly, and this represents a significant cause of fetal, neonatal, childhood, and adult morbidity and mortality [1,2]. In most developed countries, prenatal ultrasound for fetal anatomical assessment is offered to all pregnant women around 20 weeks gestation as recommended by WHO [3]. The routine 20-week ultrasound scan aims to detect congenital anomalies [4,5], and recent data demonstrate that such screening programs detect about two-thirds of babies with a congenital anomaly before birth [6].

While second-trimester screening for anomalies at around 20 weeks represents the current gold standard practice, the majority of fetal organogenesis is completed by 10 weeks gestational age [7,8]. In addition, technological improvements in ultrasound image quality now allow earlier visualization of fetal anatomy [9,10]. This means that a significant proportion of major anomalies are detectable earlier, at 11–14 weeks of gestation, in both low- and high-risk populations [11–14].

On this basis, international guidelines have begun advocating for first-trimester anomaly screening as an addition to routine clinical practice [15,16]. Nevertheless, no current recommendations for such early anatomical screening exist in most countries, including the USA and England. This is because it is not known whether such a policy would provide benefit within the context of existing screening programmes.

In England, all women are offered a scan at 11–14 weeks for confirmation of fetal viability, measurement of fetal crown-rump length (CRL) for gestational age estimation, identification of chorionicity in multiple pregnancies, and measurement of fetal nuchal translucency (NT) thickness for women who accept the offer of a “Combined Screening Test” for Down’s, Edwards and Patau’s Syndromes [17]. Crucially, there are no national recommendations for fetal anatomical assessment at this gestation, but studies suggest that screening for fetal anomalies is taking place in many centers. There is a lack of supporting evidence on (i) the impact of such ad-hoc screening at a population-level on the timing of fetal anomaly diagnosis and (ii) on how the screening should be optimally performed.

The objectives of this study were to evaluate whether first-trimester examination of fetal anatomy has an impact on the earlier diagnosis of major fetal anomalies. This was examined by establishing whether differences in first-trimester examination protocols are associated with differences in the proportion of major fetal anomalies identified early in pregnancy.

Methods

This study was conducted in two parts. The first involved undertaking a nationwide survey of current first-trimester ultrasound practice in all NHS maternity units in England in order to establish current first-trimester examination protocols. These findings were then linked with data from the National Congenital Anomaly and Rare Disease Registration Service (NCARDRS) database, to determine the impact of different first-trimester screening policies on the timing of diagnosis for a priori selected major fetal congenital anomalies. This study is reported as per the Reporting of Studies Conducted using Observational Routinely-Collected Data (RECORD) guideline (S1 RECORD Checklist).

Results

Of the 131 NHS hospital trusts participating in the NHS England antenatal screening programme, responses were received from 110 NHS trusts (response rate 84%). Completed questionnaires were submitted by more than one ultrasound unit within the same NHS trust in 8 cases, meaning 118 responses were received. In five cases, the first-trimester protocols in the two units submitting responses from the same trust were identical; in three cases the individual protocols within the same trust were different, and for the purposes of this analysis, the most detailed of the two protocols was considered. Thus, analysis from 110 NHS hospital trusts was undertaken, and based on first-trimester examination protocols, these were classified into Group A (n = 27), Group B (n = 22), Group C (n = 45), and Group D (n = 16). The responses obtained reflected a variety of healthcare settings, including district general hospitals (n = 74, 67%), university-affiliated and academic centers (n = 35, 32%), and tertiary care units (n = 12, 11%), and represented trusts from all nine designated health regions in England (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Response rates of trusts by regions as defined by Public Health England (PHE) at time of study.

https://doi.org/10.1371/journal.pmed.1004709.t001

Despite the absence of a national recommendation, 83 of the 110 responding trusts (75%) routinely perform some form of first-trimester anatomy assessment, and this takes place across all types of units (100% of tertiary care centers, 80% of university and academic centers, and 69% of district general hospitals). Amongst the 83 centers performing anatomical assessment, there were variations in what anatomy was routinely assessed: fetal head (99% of reported protocols), limbs (95%), cord insertion (91%), stomach (72%), bladder (65%), heart (17%), spine (16%), thorax (14%) neck (12%), face (7%), and kidneys (6%, S3 Appendix). Of the 110 trusts responding, 94% primarily use trans-abdominal ultrasound for screening with the offer of a trans-vaginal scan when required for improved visualization, while the remainder offer exclusively trans-abdominal imaging. The majority of units (97%) allocate 20–30 min for their first-trimester ultrasound scan appointment, regardless of whether a formal assessment of fetal anatomy is conducted and how such an examination is performed.

Using the NCARDRS database, we identified that 1,030,224 pregnancies had pregnancy care between April 2017 to March 2019 (inclusive) in the 110 NHS trusts included in the study.

Overall, 6,808 fetuses were identified as having a congenital anomaly of interest, of which 913 (13.4%) were excluded from the analysis because either the gestational age of diagnosis was unknown, a fetal loss took place prior to screening, or the pregnancy was not booked in time for first-trimester screening (Group A: n = 206/1720 (12.0%), Group B: n = 155/1247 (12.4%), Group C: n = 374/2604 (14.4%), Group D: n = 178/1237 (14.4%)). As a result, 5,895 fetuses were identified as being affected by one of the 14 prespecified anomalies antenatally and were therefore included in the analysis. The prevalence of each of the anomalies was consistent with that expected based on available data published data from population-wide anomaly databases, NCARDRS [6] and EUROCAT [22].

Of the 5,895 fetal anomalies identified, 1,929 (32.7%, 95% Confidence Interval (95% CI) 31.5–33.9) were suspected or diagnosed prior to 16 weeks gestational age. Early detection rates for the 14 anomalies (composite) was lowest in Group A (27.7%, 95% CI 25.4–30.0), with a stepwise increase in detection in Group B (31.2%, 95% CI 28.5–34.1), Group C (33.2%, 95% CI 31.3–35.2) and Group D (40.4%, 95% CI 37.3–43.4, Chi-squared for trend P < 0.0001, Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Early detection of 14 fetal congenital anomalies based on the type of first-trimester anatomical screening policy.

https://doi.org/10.1371/journal.pmed.1004709.t002

Secondary analysis broadly identified three clusters: (i) anomalies where the majority of affected fetuses are currently identified in the first trimester regardless of screening policy, (ii) anomalies where detection was higher in units routinely using a first-trimester anatomical examination protocol, and (iii) anomalies which are poorly detected in the first trimester, regardless of policy (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Prevalence and detection rates of prespecified fetal anomalies <16 weeks gestational age according to first-trimester anatomy screening policy.

https://doi.org/10.1371/journal.pmed.1004709.t003

Discussion

In this study, we have demonstrated that screening for fetal anomalies in the first trimester impacts the gestational age at which anomalies are diagnosed within a population-based screening program. We present a comprehensive analysis of first-trimester ultrasound practices and how they relate to the timing of antenatal detection of fetal congenital anomalies. By examining population-based national data from 110 NHS hospitals in England, and linking this to the national congenital anomalies database, we have established that earlier anatomical screening results in higher detection rates of fetal anomalies in early pregnancy (prior to 16 weeks gestation) at the population-level.

In addition, we have identified significant inequity in healthcare provision: three-quarters of hospitals routinely offer women first-trimester assessment of fetal anatomy despite no national guidance to that effect. It appears that this absence of national recommendations has led to individual hospitals establishing screening policies independently, so that there is considerable variation in practice. While it is clear that units have been able to establish this practice within existing resources, the differences in the standard of antenatal screening have resulted in significant regional and local inequities in maternal care, with differential fetal anomaly detection rates prior to 16 weeks gestation.

Current screening guidance recommends assessment of fetal viability, measurement of the fetal CRL, and measurement of fetal NT [17]. However, as ultrasound imaging has increased in resolution, anatomical pathologies are often instantly visible to the screening sonographer. The finding that a significant majority of units currently undertake a first-trimester anatomical screening scan, despite no such requirement, is striking. This suggests that the technological advances of ultrasound and sonographer training are outpacing policies and recommendations. The resulting variation in practice range from no formal assessment to extensive and detailed fetal examination.

Differences in screening policies revealed by our survey were also reflected in prescreening information given, how inconclusive scans are followed up, and what referral pathways exist. Hospitals more likely to evaluate fetal anatomy were also more likely to have formal policies for performing anatomical assessment. As an example, in region A, all responding trusts performed early anatomy screening and almost 90% had a formal protocol; while in region I, less than half of the responding units offered routine anatomical screening and a formal policy was present in less than one-third. These findings strongly suggest the emergence of “ad-hoc” screening, highlighting important inequities in antenatal options offered to women based on where they live. In our view, consideration of a national first-trimester anatomical screening standard would be expected to reduce some of this variation and promote equity.

Our analysis focused a priori on the detection of 14 major congenital anomalies; 12 of these are currently the focus of the second-trimester FASP. We found that approximately one third (32.7%) of these anomalies were detected prior to 16 weeks gestational age in England. This is an important finding, as most patients and healthcare providers consider that assessment for congenital fetal anomalies takes place in the second trimester. The fact that about one-third of these major anomalies are diagnosed at earlier gestations has important implications for informed prescreening consent, parental expectations of possible outcomes after first-trimester ultrasound, and provision of referral pathways. While the highest detection rates were seen in those centers performing detailed first-trimester ultrasound using formalized policies and protocols (40.4%, Group D), a significant proportion of fetal anomalies are diagnosed essentially incidentally, i.e., without formal anatomical assessment (27.7%, Group A).

There is evidence to suggest that the great majority of parents favor access to earlier screening, as it provides early reassurance or diagnosis [23–25]. For those parents with an affected pregnancy, an early diagnosis offers advantages such as additional time for genetic testing, multidisciplinary input, discussions around in-utero therapy options, and informed and balanced decision-making around pregnancy management. For those women considering a termination of pregnancy, having this performed at earlier gestations may be safer with a significantly reduced risk of maternal complications and length of hospital stay [26–30].

Previous reviews assessing first-trimester anatomical screening have suggested that conditions can be grouped into those nearly always detectable, those that are potentially detectable depending on maternal, fetal, sonographer, and equipment factors; and those that are rarely identifiable [11,31]. Our findings concur with this suggested model using real-world data: for example, over 80% of fetuses affected by acrania, exomphalos, and gastroschisis are identified prior to 16 weeks gestational age suggesting that the fetal skull and fetal cord insertion are examined even when anomaly screening is not a formal objective for the scan. Notably, these structures are readily evident in the standard mid-sagittal view required for measurement of the fetal CRL, meaning these conditions are often visualized incidentally, even when formal assessment is not intended. We identified a further group of eight conditions where a formal anatomical protocol was associated with improvements in detection rates, consistent with findings from reviews of smaller studies [14,31,32]. Finally, detection rates for some anomalies before 16 weeks gestation were poor regardless of the imaging protocol used, namely bilateral renal agenesis, congenital diaphragmatic hernia, and lethal skeletal dysplasias. Here, the small fetal size, evolution through pregnancy, and impact on amniotic fluid volumes resulted in later detection [7,8].

This study is the first to characterize first-trimester anomaly screening practice at population level, and to examine the impact of screening method on fetal anomaly diagnosis. The main strengths are the high rate of population-level coverage, with over 80% of NHS Hospital trusts in England, including over 1 million pregnancies and in every region. Although a limitation, the non-response rate was very low; it is possible that non-responders had a lower interest in first-trimester anatomical screening; or unable to support first-trimester services in their current form. Even assuming that none of the non-responders undertake any form of first-trimester screening for fetal anatomy, this would still result in 64% of trusts performing such screening, and the associations are unlikely to be changed. The use of a national anomaly registry with independent and anonymized data collection is another strength, as all NHS trusts in England are mandated to report, ensuring that data were as complete as possible. Limitations of the analysis include that we did not undertake hospital-level analysis, instead grouping into four clusters (A, B, C, D) based on the first-trimester anatomical examination protocol. This was a requirement of our review board, as the rarity of individual abnormalities could inadvertently lead to the identification of an individual in a small unit. This grouping has some limitations, as use of a screening policy alone is just one indicator of quality; other factors that may influence screening were not taken into account, such as the time given for examinations (similar between units); or skill of individual sonographers in each unit (not possible to measure at this large scale); or modernity of ultrasound equipment used for screening. In addition, protocols in each group were broadly similar rather than identical.

The data from the Nationwide Survey of Practice were collected in early 2019, before the COVID-19 pandemic, and therefore reflect screening practices at that time. Data analyzed from NCARDRS were from April 2017 (after the reporting became nationally mandated) to March 2019 (inclusive), and we assumed broad consistency in practice over this 2-year time period in all trusts; we believe this is reasonable given no new national guidance was released over this time.

These limitations would be more critical if the aim of this study was to determine with accuracy the sensitivity of the first-trimester scan in England. Rather, the objective was to explore whether reported changes in practice towards early anomaly screening reflected how and when fetal congenital anomalies are diagnosed. The results suggest that this is in fact the case: despite an absence of formal recommendations, a programme of first-trimester anatomical assessment is already underway in most centers in England. This ad hoc approach to screening may be lacking consistency and standardization between trusts, but is having a real-world impact on the diagnosis of fetal congenital anomalies in pregnant women.

Finally, an important limitation is that the NCARDRS database captures only those fetuses affected by congenital anomalies. This has supported an accurate assessment of first-trimester antenatal ultrasound screening sensitivity. However, without data on pregnancies impacted by a false positive result, assessment of specificity could not be undertaken. Based on the available literature and recent large-scale systematic reviews and meta-analyses, estimates suggest that the false positive rate for first-trimester anomaly screening is low, with estimates of 2.1% and 3.1% for major cardiac anomalies and noncardiac anomalies, respectively [31,32]. The consequences of a false positive screening result for parents will include referral to a fetal medicine specialist for ultrasound assessment (now as a diagnostic test) which may refute the original screening result or lead to additional serial ultrasound assessments, and possible offer of genetic testing, all of which may cause unnecessary anxiety and uncertainty for parents as well as additional healthcare cost. In the first trimester, there is additional complexity, as it can be difficult to identify what constitutes a false positive result as anomalies evolve. For example, it is widely accepted that the majority of bowel-only exomphalos in euploid pregnancies diagnosed in the first trimester will have resolved at later gestational ages without long-term associated complications. For these reasons, healthcare providers must be prepared to undertake appropriate counseling based on screening findings and parents well informed of the possible outcomes. A consistent nationwide approach supported by accessible patient information highlighting risks and benefits of screening, evidence-based guidelines, and educational teaching modules for healthcare providers would be expected to support informed patient consent and ensure provision of a high-quality and equitable screening service.

These nationwide data reveal significant variations in first-trimester ultrasound practices and demonstrate that more detailed screening protocols are associated with higher early detection rates of fetal anomalies. This finding highlights the need for standardized national guidelines to ensure equitable and effective antenatal care for all pregnant women. Standardizing first-trimester screening protocols within a formal extension of scope of the current screening programme could enhance early detection, improve parental counseling, and optimize pregnancy management, ultimately leading to better outcomes.

Supporting information

Acknowledgments

The views expressed are those of the authors and not necessarily those of the UK NHS, the NIHR, or the Department of Health.