Evaluating the Implementation of the Rapid Prenatal Exome Sequencing Service in England

Worldwide, there have been large investments in genomic medicine services in the past 10 years [1]. In England, multiple reports have highlighted the importance of mainstreaming genomics to deliver the long-term goals of the NHS with an intention to position the UK as a global leader in genomic healthcare [2‒4]. The 100,000 Genomes Project [5] forged a path to launch a national NHS Genomic Medicine Service in 2018. The NHS Genomic Medicine Service delivers all genomic testing in England through seven regional NHS Genomic Laboratory Hubs (GLHs) and NHS Genomic Medicine Service Alliances with a National Genomic Test Directory which sets out which genomic tests are available [6]. Tests include rapid genome sequencing for acutely unwell babies and children, genome sequencing for children and adults with selected rare conditions and cancers, and prenatal exome sequencing (pES) analysed using a panel of genes associated with fetal structural anomalies.

Fetal structural anomalies occur in around 2% of UK pregnancies. They are phenotypically and aetiologically variable, ranging from minor isolated anomalies to complex multisystem anomalies. Many studies have now shown that pES can improve diagnostic rates when fetal structural anomalies are detected by prenatal imaging, but karyotype and microarray have been uninformative [7‒10]. Diagnostic rates are higher if genetic review considers a genetic aetiology likely [7, 10]. Accordingly, several professional bodies now provide guidance for the use of pES [11‒13]. Results from pES are valued by clinicians and parents as they can guide clinical management of the pregnancy and birth, inform decisions about whether to continue or end the pregnancy, and allow accurate counselling about recurrence risk [14‒16].

pES was implemented nationally in the English NHS Genomic Medicine Service in October 2020 [17]. The NHS Genomic Service, launched in October 2018 following a major reconfiguration of genetic services in England, is delivered through a network of seven GLHs and partner Genomic Medicine Service Alliances. Each GLH is responsible for coordinating testing services for a particular geographical region in England with some GLHs also delivering testing for certain indications on a national or sub-national basis [18]. pES is listed as R21 in the National Test Directory [19], and the laboratory work is performed at two GLHs (North Thames and Central and South) each receiving referrals from approximately half the country (Fig. 1). There are 17 clinical genomics services across England, and within each GLH, there are two or three clinical genomics services who work with the fetal medicine (FM) and obstetric specialists in their local regions to deliver the pES. The FM and obstetric specialists may work within large highly specialised FM units at tertiary hospitals in major cities or see patients at smaller District General Hospitals in rural and regional areas. Across a 12-month period to September 30, 2023, local services made 760 referrals for pES to the testing GLHs, with 442 proceeding to pES testing (North Thames and Central and South GLHs, personal communication).

In recent years, guidance has outlined the importance of fully understanding how complex interventions (which include health services) are implemented [20, 21]. Evaluating implementation contributes to understanding of whether service outcomes are attributable to the implemented service or other factors [22]. Medical Research Council Guidance states that process evaluations should evaluate context and contextual factors, implementation (including process of implementation, training, resources, what is delivered and how the service is delivered; including fidelity, adaptation, dose, and reach), and mechanisms of impact [20]. One challenge of evaluating implementation is that it is necessary to understand the intended service or intervention components to evaluate implementation, but these are not always fully operationalised [23]. Previous research has demonstrated that implementation of innovations and/or health and social care services [24‒30], including genomic services [31, 32], can vary substantially between regions when implemented nationally [32], with research showing that multiple factors can influence the implementation of genomic services both internationally [31, 33‒41] and in the UK [42]. A recent review of the preparedness of the UK NHS to implement genomic medicine proposed that implementation will be influenced by four overarching factors: (i) system level factors (including use of multidisciplinary teams (MDTs), coordination and integration, digital systems, lack of guidelines, access, and health service cost), (ii) training and workforce needs (including clinician preparedness and confidence, education, and specialists), (iii) professional attitudes (including clinician attitudes and views on compatibility with their practice and values), and (iv) patient and public involvement [42]. Other key factors relevant to implementing genomic services include leadership, genomic literacy, and resistance to change following implementation [36, 37].

To fully understand and evaluate delivery of the NHS pES service, it is important to consider how it has been implemented across England and the factors that might affect this. Researchers have suggested that additional training for health professionals may be required, with close working of MDTs and specialist input. Furthermore, it may be affected by practical issues, gaps in guidelines, staff and patient attitudes, and ethical challenges [15, 43, 44]. Of note, pES is the first prenatal genomic test to be implemented within the NHS Genomic Medicine Service. As a comparator, when prenatal microarray became available in the NHS from 2012, professional bodies developed guidance [45], but there was no overarching national service delivery strategy. National surveys considering microarray reporting indicated variation in testing experience and practice across England [46].

There are many implementation science theories that may help understand implementation of interventions or services (including process models, theories exploring factors influencing implementation, and theories evaluating implementation) [47]. As the implementation of pES in England can be considered as a large system transformation (transformation affecting multiple organisations and staff to improve services, patient care, and population-level outcomes) [48], we chose to draw on the major system change (MSC) framework [49], which outlines five domains that underpin the implementation of “major system change.” These five domains are (i) decision to change, (ii) decision on which model to implement, (iii) implementation approaches, (iv) implementation outcomes, and (v) intervention outcomes (e.g., clinical outcomes, parent experience, staff experience, cost, and cost-consequences) [49]. The MSC framework has previously been used to explore widescale implementation [30, 49, 50]. As the pES service is newly developed and had not yet been independently evaluated prior to this study, it was necessary to critically explore differences between how the service was intended to be delivered, how the service was delivered in practice, and what may influence local decision-making regarding models of care that are implemented. This evidence will support national and local services to further develop and optimise service implementation and models of care offered in future. In this study, we used the first four domains of the MSC framework to guide a structured exploration of the links between implementation approaches and outcomes, in order to develop meaningful lessons for improvement of the pES service.

In this study, we aimed to explore:

• 1.

  What drove the implementation of rapid pES in England? (decision to change).

• 2.

  How did the national pathway develop and what does the intended implementation look like? (decision on which model to implement).

• 3.

  How was this service implemented locally (implementation approach) and how did individual services vary to intended implementation and each other? (implementation outcomes).

• 4.

  What factors (barriers/facilitators) influenced the approach to implementation? (implementation approach).

• 5.

  How were implementation outcomes influenced by implementation approaches?

This study is part of the wider Optimising Exome Prenatal Sequencing Services (EXPRESS) study [51] that is evaluating multiple facets of pES implementation, including clinical outcomes, cost-effectiveness, and staff [52] and parent [53] experiences.

Our multi-site mixed-methods study has considered the implementation of the national pES service across England using key documents, interviews, and a cross-sectional survey. The study has 17 study sites: the 17 clinical genomics services, their associated FM and obstetric services, and their regional GLH. The MSC framework [49] has informed all aspects of the study. Figure 2 demonstrates which data were used to explore four implementation domains of the MSC framework (decision to change, decision on model implementation, implementation approaches, and implementation outcomes).

Co-authors include researchers, academics, clinicians, and clinical scientists, based within academic institutions, the NHS, or patient support organisations. Research expertise of the team include health services research (all), the evaluation of national healthcare services and implementation science (H.W. and N.J.F.), and genomics research (M.H., M.P., M.D., L.S.C., R.M., S.A., and H.D). Two co-authors were involved in developing and delivering the national rollout of the pES service in England (LSC – North Thames GLH, SA – Central and South GLH). The researchers involved in data collection and initial analysis and interpretation (H.W., M.D., M.P., and M.H.) are independent of the pES service.

We interviewed professionals involved in designing the national pES service (termed “national service developers”) and interviewed and surveyed professionals involved in delivering the service (clinical geneticists, genetic counsellors, clinical scientists, FM consultants, registrars, and midwives) at each of the 17 sites (termed “local staff”).

The topic guides used for interviews (online suppl. file 1; for all online suppl. material, see https://doi.org/10.1159/000543104), and the survey (online suppl. file 2) were informed by a previous study on pES offered in a research setting in England [15] and knowledge of the service from relevant documentation and discussions with key stakeholders. The topic guides were mapped against the four implementation domains of the MSC framework.

National service developers and local staff were contacted by researchers and asked if they would like to take part in an interview. Participants provided informed consent. Interviews were conducted by one of four researchers (H.M.D., R.M., M.P., and M.H.; between November 2020 and December 2022). Interviews (range 23–80 min, median 44 min) took place by telephone, face-to-face, or using video call. Interviews were audio-recorded, transcribed, and anonymised.

The survey was circulated via email to staff members involved in the delivery of the pES service (FM midwives and doctors, clinical geneticists, and genetic counsellors). Individuals within each local region were nominated by the 17 clinical leads for the local pES services to ensure appropriate individuals were contacted. The survey was open for 6 weeks (March–May 2022). Relevant documents relating to the implementation of the NHS Genomic Medicine Service as a whole and the pES service specifically were identified for analysis.

Interview data analysis followed the principles of thematic analysis [54] and included both inductive and deductive approaches [55, 56]. Rapid assessment procedure sheets (RAP sheets) [57] framed around the four implementation domains of the MSC framework [49] (Fig. 2) were used to conduct the analysis. Summaries of interview findings were added to RAP sheets for each of the 17 sites by researchers (M.D., H.W., M.H., M.P., and H.M.D.).

Researchers initially inductively coded the RAP sheets to develop a summary of themes and sub-themes relating to the topics of drivers to change (M.H.), national pathway development (M.H.), local implementation of care pathways (M.D.), and factors influencing implementation (H.W.). The researchers then used findings to develop a coding framework that built on the MSC framework [49] (deductive component) and findings from coding the RAP sheets (inductive component). The coding framework was independently piloted on two transcripts, and researchers discussed and agreed any discrepancies in coding (M.D., H.W., M.P., and M.H.). The coding framework was then applied to all transcripts by one of the four researchers.

To analyse survey findings, descriptive statistics were used to explore care pathway components (e.g., referral process, consent process, testing, return of results) and the involvement of some staff groups such as genetic counsellors and genomics associates across sites (M.P.).

Data from documents, interviews, and surveys were triangulated by inputting findings into individual spreadsheets for each of the 17 study sites that included referral processes, consent process, testing, return of results. Researchers (M.D., H.W., M.H., M.P., and H.M.D.) used findings to develop models of local pES implementation at each of the 17 sites. Models were discussed with the wider research team and then shared with a clinician from each of the 17 study sites to ensure that resulting models accurately represented each sites’ care pathway. The resulting models were used to compare factors influencing implementation between sites and explore links between implementation approaches and implementation outcomes.

Data were collected from all 17 study sites. Interviews were conducted with national service developers (n = 8) and local staff (including clinical geneticists, genetic counsellors, clinical scientists, FM consultants, and midwives) (n = 55). Surveys were received from local staff (n = 159 complete surveys, 57% response rate) (see Table 1, online suppl. file 3). Results are structured around the four implementation domains of the MSC framework. Illustrative quotes are provided in online supplementary file 4.

The pES service was implemented as part of a newly established national Genomic Medicine Service in England. Interview participants welcomed this introduction and highlighted that a national service would allow greater equity of access to genomic tests, more efficient ways of working across regional geographical areas and a means to embed genomics in mainstream medicine. The use of the National Genomic Test Directory and reconfiguration of genomic laboratory services was central. Reconfiguration often involved either combining existing laboratories or rationalising the service so different laboratories in the region are responsible for different tests. Participants described benefits of having a centralised service such as “using machinery and equipment more efficiently” (Professional 42 – Clinical Scientist – National Service Developer) and felt that it made sense to have expertise focused in specialist centres. Challenges from the reconfiguration related to changes to testing pathways and clinical boundaries and the loss of existing relationships between clinical genetics teams and their onsite laboratory. Experiences varied between GLHs and the reconfiguration had involved “a lot of change and that’s still really being embedded” (Professional 29 – Clinical Scientist – National Service Developer).

The implementation of pES was driven by a combination of factors that included: research evidence demonstrating the diagnostic yield and clinical utility of pES, the ability to return results within 2–3 weeks, professionals’ enthusiasm for both the benefits of the information from pES for clinical care and empowering parent decision-making, the knowledge, expertise, and infrastructure arising from previous pES research undertaken in the UK (such as the PAGE [8] and pES for skeletal dysplasia studies [10]), and key people driving it forward. Many participants saw the implementation of the pES service as “natural progression” from offering prenatal microarray in clinical practice as “it was time to move on to that next level of testing” (Professional 54 – Clinical Geneticist). Other views supporting implementation included the potential cost benefits (e.g., cost savings from recognising genetic conditions earlier). Opportunities from offering pES as part of a national service were also highlighted such as greater equity of access, more efficient use of existing infrastructure, and the creation of opportunities for national discussion and learning.

Perceived challenges for implementation included delivering a uniform service with tight turnaround times across two testing laboratories, monitoring referral patterns, streamlining pathways and communication, reaching agreement on eligibility criteria (some views that conservative eligibility criteria would allow the laboratories to manage numbers and establish a cost-effective diagnostic yield, and others that broader eligibility would enable more testing, albeit with potential lower diagnostic rates), education, and engagement of FM doctors and midwives (particularly in smaller local units) was challenging.

The national pathway and guidelines for the pES service were informed by research evidence, national and international guidance from professional bodies, and clinical and laboratory expertise. Following the launch of the NHS Genomic Medicine service 2018, an NHS England led service development working group that included FM and genetics clinicians and clinical scientists oversaw the development and the piloting of the service. Clinical teams across England could refer parents for pES testing from October 2020 (Fig. 3).

The launch of the service was supported by a national guidance document, the test request, and consent forms and a monthly national meeting led by NHS England where the two testing GLHs could provide updates and local teams could ask questions and discuss the service. The guidance sets out the key elements for providing the national service, including eligibility criteria; testing pathways and referral processes; roles of home and testing GLHs; approach to analysis; reporting pathways; approach to reporting incidental findings; turnaround times; need for clinical geneticist involvement for case selection; test request; parental counselling; MDT attendance; GLH workloads; NHS England audits; and considerations for consent (e.g., test limitations) [17] (see online suppl. file 5 for a summary [17]). The guidance is less specific on expectations regarding roles and responsibilities and workload and staffing models at the individual local clinical services and home GLHs. Specification of staff roles for some aspects of the pathway are less clear, with guidance often referring to “referring clinicians” or “requesting clinicians” for analysis and reporting. For consent, the guidance states that this can be undertaken by healthcare professionals who have undergone relevant training, and for reporting, the guidance states that discussions may include other MDT members. Overall, guidance stipulates that it is local teams’ responsibilities to arrange workloads and ensure appropriate capacity. The national pathway and guidelines for the pES service continued to evolve over time (e.g., adding eligibility criteria, introducing an additional pathway for non-urgent pES cases when the pregnancy ended before pES testing commenced) (Fig. 3).

Nationally, two initiatives were developed to support local implementation. These included (i) a national educational meeting, which evolved from initial service development meetings and offered a learning opportunity for clinical teams to reflect and share learning, and (ii) a national clinical oversight group that was set up in March 2021 to provide independent decision-making on eligibility for complex or “borderline” cases. The group includes at least one clinical geneticist from each GLH.

To understand local implementation, we mapped the local pES pathway for each of the 17 study sites. Some aspects of the pES pathway were consistent across all 17 sites (involvement of staff from both FM/obstetric and genetics teams, the use of MDT meetings, and the pathway processes of referral, consent, testing, and return of results) (Fig. 4).

However, there was local variation in implementation between sites, independent of the site’s GLH. Seven different models of service delivery were developed (Table 2). These models of care may not necessarily all be indicative of models of “good practice.” The models were based on key overarching factors which underpin service delivery of pES: (a) who leads the service (genetics or FM or both), defined as who initiates the process of requesting pES and takes consent, and (b) the core staff involved in the overall running of the pES service at the site.

Most sites adopted genetics led models; however, in the later interviews, it was noted that some of these sites were now shifting to more active FM involvement at the point of initiating the pES pathway. Model 3 was the most frequently reported model across the 17 sites, in which genetics lead the service and staff involved throughout include FM consultants, clinical geneticists, clinical scientists, and genetic counsellors.

Genetics teams (including clinical geneticists and genetic counsellors) most commonly took consent. Joint-working models (e.g., someone from either genetics or FM could take responsibility for consent) were also identified. In one model, only the FM team was involved in taking consent.

The most common staffing model involved, at minimum, a FM consultant, clinical geneticist, and clinical scientist. In other sites, the pathway also included genetic counsellors, FM midwives, or both. For example, in some sites, FM midwives were involved in taking consent or were central to supporting parents when having testing or were involved in sample collection and dispatch (Fig. 4).

Local implementation of pES was influenced by a number of factors, including “organisational trust factors,” “national and local pES service factors,” “local staff factors,” “parent factors,” and “societal factors” (see Table 3, online suppl. file 4).

Organisational Preparedness to Implement. Trusts with sufficient staffing across genetics teams, FM teams, and support teams, infrastructure for collaborative working (e.g., satellite teams or established MDT), and robust IT systems (e.g., linkage with electronic patient records) found the pES service easier to implement. As did sites with certain characteristics (trusts based in urban areas, tertiary settings).

Previous Experience. Previous experience of offering pES (and sometimes microarray) in a research setting or as part of a clinical service facilitated implementation. Sites with limited previous experience needed greater efforts to establish pathways and educate FM staff.

Availability of Data and Evidence. Being a research active trust and establishing local audits and evaluations of pES processes facilitated implementation. Many participants felt that research and national audit data were needed to support evidence acquisition regarding the effectiveness of the service and noted that more research is needed around prenatal genomics and fetal phenotypes.

Relationships, Communication, and Collaboration across FM, Genetics, and Laboratory Teams. Preexisting or newly established close working relationships between FM, laboratory, and genetics teams facilitated implementation. Strategies that supported collaboration included embedding genetics staff in FM teams, routine or ad hoc joint clinical appointments, co-location of genetics and FM teams at the same hospital, regular MDT meetings, and use of virtual MDT meetings (Box 1).

All sites reported the importance of working within an MDT. MDT meetings were a key forum to discuss eligibility of individual cases, the interpretation and implications of results, and gather expert views of fetal phenotyping of complex cases. MDT meetings were also used for updating staff generally on the pES service, sharing educational case studies, for knowledge exchange and for tracking of individual cases. Approaches to MDT meetings varied in terms of frequency, format, and staff involved (Box 2).

Clearly defined pathways and good communication were important in supporting peripheral units referring parents for pES. Some peripheral units worked directly with the local genetics team to discuss referrals, while others discussed complex cases with the tertiary FMU in their region or routinely referred parents with complex pregnancies to the tertiary FMU. Several participants highlighted that a key reason for onward referral was access to expert scanning and fetal phenotyping (Box 3). Barriers to successful implementation included gaps in communication and collaboration (e.g. casual email requests, peripheral units not engaged), a lack of strategies to support collaboration, and no clear point of contact in genetics teams.

National Guidelines and Direction. Sites differed in their views on the role of available national guidance and direction (e.g., through national meetings), with some sites reporting that national guidance was helpful for successful implementation, and other sites reporting that more detailed national guidelines were required to help streamline and clarify some processes (e.g., on aspects such as what staffing is required, roles and responsibilities, follow up). National initiatives to support and educate staff, including the clinical oversite group and educational meetings, were valued by many participants. It was noted that attendance at the national education meetings varied, often due to a lack of awareness or understanding about the purpose of the meeting.

Logistics. Dedicated teams with clear roles and responsibilities were important to support implementation. Barriers included staffing issues such as not having enough staff on the ground to deliver the service, issues with testing pathways such as difficulties getting parental samples and administrative issues such as paperwork requirements and problems with local couriers. Several sites had taken steps to streamline the service and reduce delays in referrals and sample transfer. Staff education was important to overcome initial uncertainties around the processes and pathways.

Staff Knowledge and Experience. Staff having appropriate experience and/or specialist knowledge (e.g., FM consultants and midwives with genomics expertise) supported implementation. This included knowledge in clinical genetics for FM and obstetric staff and knowledge of FM for clinical genetics teams. Training sessions led by local genetics teams or NHS England supported understanding and awareness of the service.

Time and Capacity. All staff groups (genetics, FM/obstetrics, and laboratory) faced challenges in finding time and capacity to implement pES and attend education and training sessions. Time constraints limited capacity to participate in collaborative working such as joint clinics. Participants frequently reported that paperwork for the pES service took a long time.

Views Attitudes and Interest. Positive attitudes toward the service, self-confidence, and an interest in prenatal genomics facilitated implementation. Staff confidence increased with growing experience of delivering the pES service. Barriers to implementation included staff concerns about the service, such as worries about litigation, and a lack of confidence for implementing some aspects the service. In addition, some FM staff did not see this service as part of their role and were therefore reluctant to take on pES service-related responsibilities.

Parent Involvement in Service Development. Parents were consulted on the development of consent forms and written information but were not involved in the development of pES pathways or any other aspects of implementation.

Ability to Engage with pES Testing. Language barriers and a lack of parental understanding of genetics and pES influenced the implementation. Information for parents that is available in different languages (e.g., through interpreters or bilingual staff) and in a range of formats is needed. Transparent information on the eligibility criteria and clear expectation setting about whether parents will be eligible for pES was found to be important.

Access to Services. Participants described barriers parents face to access pES, including some people being generally reluctant to access any genetic services. Travel barriers were also noted that could be due to geography or affordability. Offering access to virtual or phone appointments was reported to facilitate implementation.

COVID-19. COVID-19 meant that workforces were stretched and that staff spent fewer dates on site but encouraged greater use of virtual ways of working (for MDTs, training, clinical appointments).

Geographical Location. Co-location of genetics and FM staff at the same or close hospitals facilitated implementation. Engaging geographically distant peripheral units and genetics/FMU collaborations were challenging at a distance.

Funding. The availability of funding benefited implementation when used for training or employing additional support staff to take on administrative roles. However, a lack of funding for clinical genetics teams to support pES delivery was challenging.

An analysis of the links between implementation models and factors influencing implementation indicated that some factors were identified in all seven models, some factors were identified in most models, and some factors were identified in only a small number of models. All services should consider the factors influencing all or most models of implementation to improve services in future. These include making sure collaborative strategies are in place, ensuring staff have the time and capacity to provide services, ensuring national guidance is clear and specific enough for services to implement, and developing strategies to improve knowledge, education, confidence, and interest of staff. In addition, support must be given to trusts that face particular challenges in implementation (e.g., those less prepared to implement or who faced particular logistical challenges). Furthermore, some challenges were present in only a small number of sites and models. These included organisations’ previous levels of experience of implementing similar services and being a research active trust.

The implementation of models involving four or more groups of staff (models 3, 4, 6, 7) were linked to some additional barriers (e.g., difficulties obtaining parental blood samples, parental access to services, funding), as well as some additional facilitators (e.g., clearer logistical procedures, specialist expertise, multidisciplinary knowledge).

The national pES service was rolled out in England as part of the newly established NHS Genomic Medicine Service because of a range of driving factors including evidence of benefits, having some infrastructure in place and support from stakeholders. Our evaluation of implementation indicated that while sites had similar pathway components (MDT working, referral, testing, analysis and return of results), models of implementation varied. Variation was at site rather than regional level, as sites within the same GLH had used different implementation approaches. Seven models of service delivery were identified that included differences in leadership, staffing, and MDT approaches. Many organisational, pES service, staff, parent and societal factors influenced variation. Some factors influenced implementation across all models of implementation. The implementation of models of service delivery which involved more staff groups tended to report more barriers and facilitators influencing implementation.

Findings extend the applicability of the MSC framework [49], by demonstrating its value in evaluating implementation of the national rollout of the pES service in England. The application of the MSC framework enabled us to conduct a thorough evaluation of many stages of implementation (including decision-making around implementation of service and service development) and the relationships between these stages; in line with MRC guidance [20, 21]. Our findings also support previous research which has outlined challenges regarding service/intervention components not fully being operationalised prior to implementation [23]. This makes it challenging to fully evaluate aspects such as fidelity. However, this study provides a worked example of how elements of service fidelity and adaptation (intended versus actual implementation at national and local levels) can be evaluated even in cases where fully operationalised guidance is not available. This, together with findings on factors influencing implementation at a local level, allowed for in-depth analysis on which models may be indicative of good practice; even when services are in their infancy and do not have fully operationalised “Gold Standard” service to compare against.

We found substantial variation in the way in which the pES service was delivered locally, both within and across GLHs. Some variation was expected, as previous research highlights a need for balance between adaptation and adherence to ensure successful implementation [39]. All sites illustrated similar pathway components, but local pES services were implemented using different models of leadership and staffing, and illustrated differences in MDT working, and genetics and FM roles. There is a risk that variation may lead to inconsistencies in service delivery, which may influence patient care. When triangulating findings with patient experience, we found some indications that parents had differences in experience which may be attributable to variation, such as differences in and experiences of follow-up care after pES testing [53]. In addition, differences in diagnostic yield between GLHs were identified in an analysis of pES outcomes, which may be partly explained by variation in local services [58]. This variation (exacerbated by continual development of national guidance – including to eligibility criteria) emphasises a need to establish a clear process for updating national guidance to avoid local variation.

Findings support previous research indicating the importance of multi-levelled factors when implementing services at scale, and the importance of systems, training and workforce, interest, and patient and public involvement [31, 33‒42, 59]. National strategies such as the national educational meeting and the introduction of the clinical oversight group were valuable. This study extends previous research by linking implementation approaches and implementation outcomes to show which factors influenced all/some/few models of implementation. In particular, findings emphasised the importance of communication and collaboration between national and local teams (including MDT working), staff knowledge and education, and interest. However, it is important to note that many of these factors may be influenced by financial limitations. Furthermore, findings indicate that ways in which implementation occurred at a national level may have influenced the amount of variation seen. For example, gaps in the guidance provided to local services (e.g., around staff roles and responsibilities), lack of tailoring of support for different services, e.g., for inexperienced services, and the setup of complex logistical processes.

This multi-site, mixed-methods, and theoretically informed [49] study allowed for a comprehensive exploration of implementation approaches and impacts on implementation outcomes. This facilitated an in-depth understanding of how pES services were implemented (from national service development through to local implementation) and the factors influencing implementation.

One limitation was that models were developed from available findings. However, there was variation in the numbers of interviews, surveys, and documents available from each site and some staff groups, such as midwives, were under-represented. In addition, the models only represent service delivery at a snapshot in time and services continue to evolve as evidenced by findings in later interviews that FM teams at some sites were taking on more tasks in pES service delivery. It is probable that services will have changed since interviews were conducted. Some pathways may not have been captured within our dataset and may not be represented (e.g., smaller pathways from peripheral units).

Finally, it is a limitation that it was not possible in this manuscript to link “implementation outcomes” with the final MSC framework domain, “intervention outcomes,” which includes clinical utility and cost-effectiveness [49]. This means that from this manuscript alone, it is not possible to quantitatively evidence whether any of the models of care may be more optimum than others, in terms of clinical effectiveness. Our study has, however, illustrated what works in practice, how these models work and why, and established the groundwork for the interpretation of outcome findings to be explored in other papers from the wider project that will focus on outcomes [58].

Implementation of the pES service in England is still in its infancy. Findings can therefore be used to inform ongoing development and further optimisation of the pES service within the NHS Genomic Medicine Service and the implementation of pES services in other settings. The national rollout of the service happened rapidly from a local service perspective. Therefore, while these strategies may echo standard implementation strategies, findings have indicated that these strategies were not necessarily used in the initial implementation and that guidance documents were not fully operationalised. To improve this nationally implemented service moving forward, the following recommendations are proposed:

• 1.

  National or regional leads to identify and support organisations who may not have previous experience as these teams may require additional support in terms of rollout.

• 2.

  Close collaboration and communication within and across genetics, FM/obstetrics, and laboratory teams are required and can be facilitated through a range of strategies including joint appointments, regular MDT meetings, strategic use of virtual meetings, regular communication, clear roles and responsibilities, and embedding of genetic expertise within FM teams.

  • a.

    If existing collaborative relationships are not in place, there is a need to help bring different services together to develop these relationships to support implementation.

  • b.

    Funding is needed to support effective collaboration and communication and more staff.

• 3.

  Better use of virtual technology for parent expert genomic counselling and for the expert review of imaging to improve phenotyping.

• 4.

  Further education for all disciplines involved in pES to support implementation. FM professionals may need further education in genomics and counselling around pES and more expertise in prenatal clinical genomics is required.

• 5.

  Increased staffing across the various disciplines involved, including administrative support and time in consultant job plans to support service delivery; with funding to support this.

• 6.

  Streamlined logistics and processes.

• 7.

  Potential expansion to national guidance for the pES service. For example,

  • a.

    Outlining the importance of adapting services to fit local contexts but to also provide expectations and clarity around which parts of the pathway should not vary and which parts may be better for services to adapt locally.

  • b.

    Providing further clarity on staffing and service pathway elements (e.g., follow-up processes and expectations around this and roles and responsibilities throughout different parts of the pathway).

There has been substantial variation in the local implementation of the pES service in England. An analysis of the links between the different models of service delivery (implementation outcomes) and implementation approaches indicated that some factors influenced all or most models. Findings have informed key recommendations for implementing pES services that will support the development of optimal care pathways that facilitate equity of access for all parents to a high-class pES service regardless of where they live.

We are very grateful to all of the participants who took part in the interviews and survey. We thank Rema Ramakrishnan for critically reviewing this manuscript and the wider EXPRESS team for providing feedback throughout the study.

This study was classified as a Service Evaluation by the NHS Health Research Authority (HRA) and did not require research ethics committee approval in accordance with local/national guidelines. The service evaluation has been registered with the Research and Development office at Great Ormond Street Hospital for Children NHS Foundation Trust. Informed written consent or audio-recorded consent was obtained from participants to participate in the study.

No competing interests were disclosed.

This manuscript presents independent research funded by the National Institute for Health Research (NIHR) Health Services and Delivery Research programme (NIHR127829). R.M., L.S.C., M.P., and M.H. were also funded by the NIHR Biomedical Research Centre at Great Ormond Street Hospital. All research at Great Ormond Street Hospital for Children NHS Foundation Trust and UCL Great Ormond Street Institute of Child Health is made possible by the NIHR Great Ormond Street Hospital Biomedical Research Centre. N.J.F. is an NIHR Senior Investigator. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the UK Department of Health. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript.

L.S.C., M.H., and N.F. conceived the study and acquired the funding. All authors were responsible for study design. H.M.D., R.M., M.P., and M.H. collected interview data, and M.P. led the survey data collection. Interview findings were analysed by H.W., M.D., M.H., M.P., and H.M.D., and survey analysis was undertaken by M.P. M.D., H.W., and M.H. M.P. developed initial models in discussion with all authors. H.W., M.D., and M.H. drafted the manuscript with contribution from all authors. All authors commented on drafts of the manuscript and approved the final version. Author contributions according to the CRediT taxonomy are provided as follows: H.W.: methodology, data curation, formal analysis, validation, writing – original draft, and writing – review and editing. M.P., M.D. and H.M.-D.: methodology, investigation, data curation, formal analysis, validation, writing – original draft, and writing – review and editing. R.M.: methodology, investigation, validation, and writing – review and editing. S.A.: methodology, validation, and writing – review and editing. N.J.F. and L.S.C.: conceptualisation, funding acquisition, methodology, supervision, validation, and writing – review and editing. M.H.: methodology, funding acquisition, investigation, data curation, formal analysis, supervision, validation, writing – original draft, and writing – review and editing.

The data that support the findings of this study are not publicly available due to privacy or ethical restrictions. Where appropriate, data may be made available upon reasonable request to the corresponding author (H.W.).