neoWONDER: Neonatal Whole Population Data linkage approach to improving long-term health and wellbeing of preterm and sick babies

BACKGROUND AND PURPOSE

Over the last 14 years in the UK, over 100,000 babies were born very premature (before 32 weeks) or with a condition requiring surgery in the first few weeks of life. With advances in neonatal care, more babies are surviving, but there is still a limited understanding of the long-term impact that many neonatal care and surgical interventions have. Understanding long-term impact requires following up children as they grow up. This is complex and costly. As a result, there is little information from the last decade to inform how these children are doing as a group. Without this information, it is difficult to know whether and how things can be done differently in neonatal units to improve longer term outcomes.

The aim of this application is to link existing data for a cohort of very preterm and surgical babies in neonatal units born 2007-2020 in England with data available in the National Neonatal Research Database (NNRD), held at the Neonatal Data Analysis Unit (NDAU) at Imperial College London (https://www.imperial.ac.uk/neonatal-data-analysis-unit/neonatal-data-analysis-unit/). The NDAU forms part of the Neonatal Medicine Research Group led by the Chelsea and Westminster Campus of Imperial College London. The NNRD gives near-complete population coverage of all very preterm and unwell newborn babies admitted to NHS neonatal units. The cohort supplied for linkage is therefore near-population. A letter was sent out to all neonatal units asking whether the data they each contributed to the NNRD could be included in the study. All units were given the opportunity to opt-out. Only data provided by units that have not opted out will be included in the study and supplied to NHS England for linkage.

This study could provide important information to better inform health and educational service planning. Better service planning could support children in reaching their full potential. It may help healthcare professionals to have evidence-based discussions with families about what the future may hold for their children. The study hopes to provide population level data on the long-term outcomes of this cohort and looks to examine how interventions and exposures in the neonatal period affect these outcomes. This is with the view of identifying those that may be modified and improved through changes in clinical practice and policy.

This project is part of a programme of research, 'Neonatal Whole Population Data linkage to improve lifelong health and wellbeing of preterm babies' (neoWONDER). This programme expects to link National Neonatal Research Database (NNRD) data to other population-level physical, mental health and education data and has been awarded funding by the National Institute for Health Research (NIHR) - https://fundingawards.nihr.ac.uk/award/NIHR300617.

This application is unique in that no study has previously linked together the rich clinical neonatal data held on the NNRD, the Paediatric Intensive Care Audit Network (PICANet) dataset, South London and Maudsley NHS Foundation Trust (SLaM) Clinical Record Interactive Search (CRIS) (approximately 6000 children in SLaM CRIS dataset) and Hospital Episode Statistics (HES) databases to later educational outcomes. ECHILD only links education data and HES data. By linking together with the NNRD, Imperial College London can explore the impact of neonatal care and interventions on later life health and educational outcomes. This may inform improvements in neonatal care.

SLaM CRIS holds rich free text data not available in other mental health datasets because they are derived from Child and Adolescent Mental Health Service (CAMHS) datasets for South London and other regions. These will be explored using Natural Language Processing and is a key part of the training programme for the NIHR Advanced Fellowship which funds this work.

DATA SUMMARY

The NDAU at Imperial College London require the following data from NHS England:

1) Demographics data: including forename, surname and up to date postcodes for all NHS patients. Demographics data is obtained under a separate Agreement (ref: DARS-NIC-609893-N5P5L) to allow linkage to the National Pupil Database (NPD) at the Department for Education.

2) Civil Registration (deaths) – secondary care cut data: linked to Hospital Episode Statistics (HES) data and will allow the NDAU at Imperial College London to capture the deaths of people who have been treated in English hospitals, irrespective of whether they died in hospital or not.

3) HES data: including data on admissions to NHS hospitals, critical care, outpatient appointments and attendances at emergency departments across England.

4) Mental Health Services Dataset (MHSDS): contains individual level data for all children accessing mental health services across the community, outpatient and inpatient settings in England.

The Civil Registration (deaths) - secondary care cut, HES and MHSDS data are required under this Agreement, DARS-NIC-283774-B9Z6K. The Demographics data is required under a separate Agreement as the Data Access Request Service at NHS England does not support the inclusion of more than one data recipient under a single Agreement. The Demographics data will flow to the Department for Education for linkage to the NPD whereas the Civil Registration (deaths) - secondary care cut, HES and MHSDS will flow to Imperial College London.

The NDAU at Imperial College London also require NHS England to be the independent third party to conduct the linkages between the NNRD, PICANet and the SLaM CRIS system. PICANet and SLaM CRIS have provided confirmation to NHS England that they are content for their data to be linked with NHS England data for the purposes of this study.

PICANet contains details of the treatment of all critically ill children in paediatric intensive care units (PICU) across the UK.

SLaM CRIS contains Child and Adolescent Mental Health Service (CAMHS) datasets for South London. SLaM CRIS holds rich free text data not available in other mental health datasets because they are derived from Child and Adolescent Mental Health Service (CAMHS) datasets for South London and other regions. These will be explored using Natural Language Processing and is a key part of the training programme for the NIHR Advanced Fellowship which funds this work.

Personal identifiers are necessary to conduct the linkage between the NNRD and other health and education databases:

i) NHS number, date of birth, gender, postcode and unique ID will be used to link the NNRD to HES, Civil Registration (deaths) – secondary care cut, MHSDS, PICANet and SLaM CRIS data (under DARS-NIC-283774-B9Z6K).

ii) Forename, surname, date of birth, postcode, gender and unique ID are required to link the NNRD to the NPD (under DARS-NIC-609893-N5P5L) because the NPD does not hold NHS numbers (holds Pupil Matching Reference numbers instead) and the NNRD does not contain forename, surname or recent postcodes (only the postcode at the time of discharge from neonatal care). Therefore, the NHS number from the NNRD needs to be linked to the Demographics data held by NHS England prior to linkage to the NPD. The linkage to the NPD will be undertaken by the Department for Education. The Department for Education have provided confirmation that they are content to support the neoWONDER project and data linkage with the NPD - evidence of this has been provided to NHS England.

Linkage of the NNRD to other datasets, specifically HES, has previously been undertaken. A 2019 study demonstrated that NNRD to HES linkage was feasible (https://www.ncbi.nlm.nih.gov/books/NBK546526/). It also highlighted that the quality of neonatal data in HES was variable, including for crucial parameters such as gestational age (20% missing) and birth weight (1.5% biologically implausible), and that “data quality and completeness of recording were better in the NNRD than in HES for most key variables”. Linkage to the NNRD is necessary to ensure completeness and quality of the data, as well as to include data items not covered by HES.

JUSTIFICATION FOR REQUIRING NHS ENGLAND DATA:

Broadly, the outcomes of interest (relating to this application) in this study can be divided into health resource utilisation and clinical outcomes.

Health resource utilisation is an outcome of interest on a societal level (for resource planning), as well as at the individual family level. The NDAU at Imperial College London require data from emergency departments, mental health services, critical care, inpatient admission and outpatients from NHS England in order to build a full picture of the ongoing interaction the cohort has with health resources as they age. There is no information from the last decade to inform how these children are doing as a group. Linking existing health data to education data in England offers a practical solution to this problem. The data required from NHS England may provide important information to better inform health and educational service planning, which could better support preterm children to reach their full potential. It may also help healthcare professionals to have evidence-based discussions with families about what the future may hold for their children.

Clinical diagnoses are also an important outcome of interest. Clinical outcomes of interest have so far been informed by the core neonatal outcome set. The Data could help facilitate the evaluation of mental health and behavioural conditions, chronic health conditions, special educational needs etc. over time. Imperial College London are particularly interested in civil registration (deaths) – secondary care cut data as mortality is an important outcome of the study.

The Data will also allow Imperial College London to adjust for confounding factors. For example, the Data will include fields which will provide information relating to an individual's Index of Multiple Deprivation, geographical location (by region) and ethnicity.

COHORT

For England, a cohort from the NNRD will be linked to other health and education databases. The cohort from the NNRD comprises approximately 120,000 individuals. Whilst linkage between health datasets will include up to those born in 2020, the linkage to the NPD and SLaM CRIS will be limited to children born up to the end of 2016 as the youngest cohort will be at least school-age by 2020.

Cohort 1 Born 2007-2020 in England: link to health data (HES, Civil Registration (deaths) - secondary care cut, PICANet, MHSDS). Cohort 1 will include preterm babies born less than 32 weeks and surgical babies (all gestations) with one of six surgical diagnoses:

1. Necrotising Enterocolitis – condition causing the tissue in the bowel to become inflamed.

2. Hirschsprung’s Disease – rare condition affecting the large intestine; causes problems with passing stool.

3. Gastroschisis – birth defect resulting in a hole in the abdominal wall; causes the baby’s intestines, and sometimes other organs, to be found outside of the body.

4. Oesophageal Atresia – rare birth defect causing the oesophagus to not form properly.

5. Congenital Diaphragmatic Hernia – defect in unborn babies where the diaphragm fails to close during prenatal development; creates an opening which allows contents of the abdomen to migrate into the chest.

6. Posterior Urethral Valves – condition only affecting male babies; causes a blockage in the posterior urethra (near the bladder) resulting in difficulties in passing urine.

Cohort 2 Born 2007-2016 in England: link to school age outcomes (NPD and SLaM CRIS). Cohort 2 will include preterm babies born less than 32 weeks gestation.

Cohort 3 Born 2012-2016 in England: link to school-age outcomes (NPD and SLaM CRIS). Cohort 3 will include surgical babies (all gestations) with one of six surgical diagnoses: necrotising enterocolitis, Hirschsprung’s disease, gastroschisis, oesophageal atresia, congenital diaphragmatic hernia and posterior urethral valves.

Cohort 1-3 will include those born in England only. Population: around 8,000 babies are born <32 weeks in England each year. Outcomes for the whole population will be described. Cohorts will then be formed for comparative studies, examining exposures during the neonatal period and long-term outcomes.

DATA MINIMISATION

The earliest data from the NNRD is 2007. The NDAU at Imperial College London require data for surgical and preterm babies born 2007-2020 (approximately 120,000 individuals) as the aim is to examine short and long-term outcomes over the life-course. The oldest cohort in the 2022 download will be 15 years old. The Data will by minimised to only patients who meet the inclusion criteria – criteria outlined in subsection titled ‘cohort’.

Data covering the whole of England is necessary as this is a whole population study. The NNRD covers England, Wales and Scotland. As part of the neoWONDER study, the NDAU at Imperial College London will also link data for very preterm babies to the Wales Secure Anonymised Information Linkage (SAIL) databank. This process is separate from this Agreement.

Patient identifiable data is required for linkage only (under DARS-NIC-609893-N5P5L). Once linked, data will be pseudonymised. In line with the HES analysis guide, only aggregate-level data with small numbers suppressed will be published. It is necessary that the data be linked at record-level to identify different exposures and interventions, and compare outcomes to identify associations.

ETHICAL / LEGAL CONSIDERATIONS

Due to the need to access personal identifiable data for the linkage, the NDAU at Imperial College London have gained Confidentiality Advisory Group (CAG) approval (Ref 21/CAG/0081) under Section 251 of the NHS Act 2006 to flow confidential information without consent.

The NDAU at Imperial College London have CAG approval to link existing data on health and educational databases for two reasons:

1. The alternative is obtaining long-term data through consent-based face-to-face cohort studies which are complex, intrusive and expensive - thus limiting the validity of findings to a whole population. Another major drawback of opt-in consent-based studies, is that seldom heard groups, including those whose English is not their first language, may not participate in such studies. To exemplify this latter point, the UK EPICure studies (https://www.ucl.ac.uk/womens-health/research/neonatology/epicure) followed up babies born before 26 weeks in 1995 and 2006. 92% were assessed at 2.5 years and 71% at 11 years in EPICure 1. Those lost to follow-up were more likely to have a non-white ethnic origin, unemployed parents and cognitive impairment. 55% were followed up at 3 years in EPICure 2. As survival improves and numbers rise, these studies are unfeasible and overburdensome for families. Recently, the US National Children’s Study and the UK Early Life Study were both abandoned due to slow recruitment, resulting in a waste of US $1.2 billion and £9 million.

Furthermore, specific to this study, contacting over 100,000 families who had preterm babies born in the last 15 years itself will require linking personal identifiers to obtain contact details. Importantly, contacting families or individuals with experience of preterm birth may cause unnecessary distress especially if the child has subsequently died or has complex needs.

2. Inclusion of data from the whole population through data linkage would maximise the utility of these data and result in the most meaningful and generalisable findings.

In addition to s.251 support via CAG, this study has also received Research Ethics approval (HRA approval: IRAS project ID 293603 REC Ref 21/EM/0130).

PATIENT AND PUBLIC INVOLVEMENT (PPI)

To ensure diversity and inclusivity of participant involvement, a website was developed and launched in September 2020 to raise awareness (https://www.neowonder.org.uk/). neoWONDER was advertised through charities such as BLISS (https://www.bliss.org.uk/), Twins Trust (https://twinstrust.org/), Smallest Things (https://www.thesmallestthings.org/), parent networks and the researchers’ social media pages (Twitter - @neoWONDER20, Instagram - @neowonderUK). To date 586 parents and adults born preterm have signed up to the website and receive regular newsletters, and the PPI group continues to grow.

The design of the study has been informed throughout by patient and public involvement, and there has been an exploration of the acceptability of using patient identifiable data in this study without consent. The first 6 months of the 5-year neoWONDER programme was dedicated to the workstream 'Parent and patient perspectives on linkage between existing data to evaluate long-term health and wellbeing of preterm babies'. This workstream received REC approval and commenced in October 2020 (reference 20/YH/0330 IRAS number 291612). Four parents and ex-patients helped co-develop the neoWONDER research proposal. Subsequently, a wider patient, parent, public involvement workstream was developed, comprising focus groups, interviews and a national survey, involving 543 parents and ex-patients. Survey data is not being supplied to NHS D for linkage. A write-up of the findings from the national survey is currently in draft format, with expectation for these results to be submitted for journal publication in Spring 2023.

The findings of the survey indicated that the majority of respondents were supportive of the concept of linking together existing routine data using identifiers without consent, as long as the final data available to researchers is pseudonymised.

In addition, as part of a larger NIHR-funded Medicine for Neonates research programme using routine real-world data for research, the acceptability of linking neonatal health and education records without consent was explored with parents on the neonatal unit. 1319 parents and families were surveyed. Over 80% and 85% were very or fairly confident, respectively, about data security and accuracy. Nearly two thirds agreed that opt-out should be the default position for data-sharing agreements. There was strong support for the use of identifiable data without consent to enable data linkage to occur.

Details about the study will be disseminated widely through charities, neonatal units and social media, including a co-designed video animation explaining the study and opt-out processes. There is a very engaged patient and parent group in workstream 1 advising and supporting this work, together with BLISS, a co-applicant of the neoWONDER programme. It will be possible for individuals to apply to opt-out even after data has flowed to NHS England as Imperial College London will provide NHS England with an updated cohort file with those participants removed prior to NHS England performing linkage with the required datasets.

CONTROLLERSHIP

Imperial College London are the sole data controller who will also process the data. This project has been registered with the Imperial College Data Protection Team. At no point will researchers have access to identifiable information. Researchers (all substantive employees of Imperial College London apart from one PhD student from Oxford University who holds an honorary contract with Imperial College London) will only have access to pseudonymised data; one linked dataset (containing health and education data), and one linked dataset (without educational data) held at the Office for National Statistics (ONS) Secure Research Service (SRS).

The Department for Education (under DARS-NIC-609893-N5P5L) and the ONS and Microsoft Limited (under DARS-NIC-283774-B9Z6K) will be data processors of NHS England data for the purposes of this study. These organisations will not have access to data from other organisations (or each other). The Department for Education are a data processor as they will be matching the linked NNRD-Demographics data supplied by NHS England to the NPD. Microsoft Limited are a data processor as the Department for Education uses Microsoft Azure cloud hosting for the storage and processing of data (including NHS England data for the purposes of this Agreement). ONS are a data processor as the data supplied under DARS-NIC-283774-B9Z6K will eventually be stored at the ONS SRS. Per Part 1, Section 3(4)(a) of the Data Protection Act 2018 the storage of data is considered a type of processing.

The wider projects include collaborators at the University of Oxford. The University of Oxford are collaborating and partially funding the work with surgical outcomes. The University of Oxford will not access or process NHS England data. The University of Oxford will also not determine the aims and objectives of the project. To this end, they are not listed as a data controller or a data processor. The PhD student is affiliated with Oxford University. The student is named on the CAG application. An honorary contract between the student and Imperial College London is in place so that the student can legally access and process NHS England data.

The National Institute for Health and Care Research (NIHR) are funding the study. NIHR have no means to access or process the data and do not determine how or why the data are processed.

SLaM: Data Controller for CRIS.

Healthcare Quality Improvement Partnership (HQIP) / NHS England: Joint Data Controllers for patients treated in England and included in PICANet.

SLaM and HQIP (HQIP have delegated authority to review and sign off data access requests on behalf of NHS England) have provided signed confirmation that they are content to support the neoWONDER project and the data linkages involved, as outlined under neoWONDER’s CAG approval documentation.

SLaM and the Universities of Leicester and Leeds (Data Processors of the data provided by English NHS providers for PICANet) are cohort providers only and as such are not considered data controllers or data processors under this agreement. They will not access nor process NHS England data.

LEGAL BASIS FOR PROCESSING

The lawful basis for the data processing proposed for this study is that of Public Task, as set out under Article 6(1)(e) of the General Data Protection Regulation (GDPR).

The legal basis for processing special category data is under Article 9(2)(j) of the GDPR. This is processing necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes. Appropriate safeguards will be in place when processing data in accordance with Article 89(1) of the GDPR - the use of pseudonymisation to respect the principle of data minimisation.

Imperial College London is a public authority as described under Schedule 1 of the FOI Act 2000. Imperial College London’s Royal Charter confers power on the University “to provide the highest specialised instruction and the most advanced training, education, research and scholarship in science, technology and medicine”. Imperial College London can rely on this legal basis as the data are required to better understand the long-term impacts of neonatal care interventions. The outcomes of this study could provide important information to better inform health and educational service planning. To this end, the study meets the conditions set out under DPA 2018 Schedule 1 Part 1 (4).

By signing the Data Sharing Agreement, all organisations party to this agreement must comply with the Data Sharing Framework Contract, including requirements on the use (and purposes of that use) by "Personnel" (as defined within the Data Sharing Framework Contract i.e.: employees, agents and contractors of the Data Recipient who may have access to that data).

The transfer of personal identifiers for data linkage will be limited to:

i) between the NDAU at Imperial College London and NHS England

ii) between NHS England and the Department for Education

The transfer of personal identifiers between the NDAU and NHS England and NHS England and the Department of Education will utilise the secure electronic file transfer (SEFT) service provided by NHS England.

Personal identifiers are necessary to conduct the linkage between the NNRD and other health and education databases. NNRD/PICANet/SLaM CRIS identifiers to be sent to NHS England:

NHS Number,

Date of Birth,

Gender,

Postcode,

Unique ID.

The file NHS England sends back will have a unique pseudo ID, plus a flag for who is a PICANet baby or SLaM CRIS baby so Imperial College London can ask for the relevant clinical data from PICANet and SLaM CRIS researchers.

For the purpose of data linkage only, personal identifiers will be made available to a restricted number of staff in the NDAU at Imperial College London, NHS England and the Department for Education.

A “Split-file” process will be used to separate personal identifiers from the clinical dataset so that only identifiers (without clinical data) are shared with the independent third party for linkage. NHS England for England, and Digital Health and Care Wales for Wales will act as the third parties to conduct the linkage. The data flows are designed such that no organisation will hold data that they do not already hold, and researchers will only analyse pseudonymised data. No clinical data will be transferred from the NNRD/PICANet/SLaM CRIS to the third parties. Once third party linkage has been carried out, all personal information is removed and the linked records will only retain the pseudonymised unique ID. The pseudonymised data will then be securely transferred to the NDAU at Imperial College London to be linked back to clinical data from PICANet and SLaM CRIS using the unique ID (under DARS-NIC-283774-B9Z6K) and to the Department for Education for linkage to the NPD (under DARS-NIC-609893-N5P5L). No researchers will have access to these identifiers and individual children cannot be re-identified in the data set.

Split file process of the clinical data: NNRD, PICANet, SLaM CRIS

Step 1: Split NNRD, PICANet, SLaM CRIS into 2 files: File 1 (identifiers) and File 2 clinical data (no identifiers). Both hold a unique ID. File 1 is sent to NHS England, acting as a Trusted Third Party (TTP) to be linked to HES, ONS mortality, MHSDS data using the NHS number, date of birth, gender, postcode and unique ID. File 2 is retained in the NDAU (without identifiers).

Step 2: PICANet, SLaM CRIS, NDAU will send file 1 (identifiers, unique ID) for babies born in the study years to NHS England. NHS England will retain linkages for matches and discard data for babies who do not have matching records on NDAU, PICANet and SLaM CRIS, respectively. NHS England will flag to the NDAU at Imperial College London who is a PICANet baby or SLaM CRIS baby so Imperial College London can ask for the relevant clinical data from PICANet and SLaM CRIS researchers.

Step 3: Identifiers are then removed with unique pseudonymised ID retained, and transferred back to the NDAU.

Step 4: File 2 (clinical data, unique ID, no identifiers) is then linked back to File 1 by Imperial College London using the unique ID without identifiers.

The Office for National Statistics (ONS) Secure Research Service (SRS) is set up for ONS-accredited researchers to access data from the National Pupil Database (NPD). Data cannot be downloaded from the SRS, and ONS procedures ensure it operates within a legal framework without disclosure of sensitive information.

The ONS SRS is a certified data processor under the Digital Economy Act 2017 and is compliant with the Data Security and Protection Toolkit. Access will be restricted to a limited number of Imperial College research staff with ONS accreditation working on the project.

All researchers accessing the Data are substantive employees of Imperial College London, the data controller, or hold an honorary contract with Imperial College London permitting them to process NHS England data. All researchers have obtained the appropriate data protection and confidentiality training.

No attempt will be made by the research team at Imperial College London to re-identify individuals once the data is in a pseudonymised format.

There is a fully funded comprehensive NIHR dissemination plan in the fellowship awarded. One parent/ex-patient representative hopes to attend and present at a conference (British Association of Perinatal Conference). Parents and ex-patients are anticipated to be co-authors in workstream reports, academic publications and a parallel report for public and policy makers. Written reports summarising the research findings hope to also be produced.

These hope to be disseminated to the neoWONDER PPI group, other families through UK neonatal units and the charity BLISS using neoWONDER's established communication channels (newsletters, social media - Twitter - @neoWONDER20, Instagram - @neowonderUK, website - www.neowonder.org.uk and volunteers). Social media sites are run by parents and patients.

BLISS also hope to facilitate dissemination through their social media channels (eg. Instagram - @blisscharity), network animation videos and leaflets (‘What is data linkage and how can it benefit you, your baby and other families?). BLISS is the leading UK charity for babies born premature or sick. Their vision “is that every baby born premature or sick in the UK has the best chance of survival and quality of life.”. They have expressed explicit support for the neoWONDER study.

Imperial College London anticipate publication of findings in Spring 2023.

The following publications are planned:

For public, parents, charities:

1. Co-designed resources to help raise awareness of the benefits of data linkage (podcast /video/infographics)

2. Co-designed leaflets to disseminate lay summary findings

For academics and health professionals:

Clinical Senior Lecturer at Imperial College London plans to write and submit the following to peer-reviewed journals and conferences:

1. Data linkage to evaluate the long-term health and wellbeing of preterm babies: What patients and parents want to know; Arch Dis Child; Royal College of Paediatrics and Child Health Conference

2. Methodological linkage between the National Neonatal Research Database and other health, education, environmental databases; BMJ/Arch Dis Childhood/ PLOS/International Journal of Epidemiology; Data Linkage conference

3. Causal inference methodology to evaluate the impact of maternal breast milk on long-term health and educational outcomes: American Journal of Clinical Nutrition/ Pediatrics/Journal of Pediatrics/JAMA; Neonatal Society conference

4. Long-term outcomes of very preterm babies born before 32 weeks in England and Wales: 2007-2018; Lancet/ New England Journal of Medicine; Paediatric Academic Societies Conference

5. Childhood mental health outcomes following very preterm birth; Arch Dis Child; British Association of Perinatal Medicine Conference

6. Environmental influences of physical health of very preterm babies in England and Wales; Lancet/BMJ/JAMA journals; Neonatal society conference

7. Socioeconomic influences of health and educational outcomes of very preterm babies in England and Wales; Lancet/BMJ/JAMA journals; Joint European Neonatal Societies Conference

8. The New paradigm of neonatal data linkage. An academic commentary/reflection on following secondment at the SAIL databank Wales SAIL and SLaM King’s College London; BMC Research Methodology; Data linkage conference

For health and education policy makers

• A lay summary of the findings via academic paper, social media and conference presentation

Imperial College London hope to use the results to describe and quantify later life complications for babies born extremely premature. These results may help inform public health and educational policies with regard to resource allocation and investment needed to optimise support and improve longer term outcomes for preterm babies. If the study results show modifiable influencing factors on longer term outcomes, such as certain neonatal unit interventions and care pathways, then this new knowledge may be shared nationally and care standardised. Individual babies in the future may benefit from improved neonatal and more standardised care. If this linkage proves successful, the methodology developed including data linkage and statistical methods hopes to be further used to benefit other patient groups such as moderate preterm and term babies with specific conditions such as congenital cardiac diseases.

This application is unique in that no study has previously linked together the rich clinical neonatal data held on the NNRD, the PICANet dataset, SLaM CRIS and HES databases to later educational outcomes. ECHILD only links education data and HES data. By linking together with the NNRD, Imperial College London can explore the impact of neonatal care and interventions on later life health and educational outcomes. This may inform improvements in neonatal care.

1) Impact on families

a) For parents, the results of the neoWONDER study may provide information on what to expect if their baby is born very premature. This may help with antenatal and postnatal counselling and decision-making; often in very difficult circumstances due to the current lack of long term data.

b) neoWONDER hopes to quantify and shine a spotlight on the additional needs and resources at population level; this may help influence policy and help the government direct health and social care resources to where it is most needed. Families who have personal and family costs that come with the additional caring demands may benefit from better support. By being aware of these additional needs earlier on, families may also be able to plan and manage their finances either privately, or from seeking funds from agencies including the charitable sector.

c) neoWONDER has established a parent and patient involvement group of nearly 600 that regularly participate in ongoing focus groups and surveys. In some studies, they are part of the research team. Feedback from parents has been overwhelmingly positive. They say that by being involved, they feel they are being heard and that their experiences can be used to potentially help their child and others in the future.

A potential output is quantifying the scale and extent of the lifelong challenges babies born preterm may face, and resources required from public services to support these children and families. Therefore a potential benefit to this focussed research is encouraging or supporting additional funding or research into the prevention of preterm births

2) Impact on neonatal clinical teams

The lack of long-term data for preterm babies born in the last decade has significantly limited the ability of UK clinical teams to counsel parents on outcomes and expectations. This means that the latest Framework, which includes gestational limits of viability for extreme preterm infants published by the British Association of Perinatal Medicine (BAPM) and other internationally published studies (www.bapm.org/resources/80-perinatal-management-of-extreme-preterm-birth-before-27-weeks-of-gestation-2019), is based on cohorts of babies born more than 14 years ago.

By having long-term data on cohorts of babies who receive contemporary neonatal care, neonatal clinical teams hope to be able to evaluate their outcomes over time, and together with parents, use these data to help facilitate their shared decision-making.

3) Impact on other health care providers

neoWONDER hopes to be able to quantify the service needs from other care providers following discharge from neonatal care including paediatric intensive care and other specialist services. These data may help forecast the additional needs and resources required for the growing number of preterm babies surviving to discharge from neonatal care.

4) Impact on Commissioning: Service and economic implications for the NHS

The neoWONDER programme hopes to map out the health care journeys for approximately 120,000 babies as they grow up. These data can help quantify the additional NHS resources needed to optimise long-term outcomes for preterm babies and help inform commissioning of neonatal service pathways of care. Improved early years health care may benefit children, reduce future burden on the NHS and benefit society in the long-term.

5) Impact on the NHS and society

This research has the potential to significantly reduce the costs associated with preterm birth. Being born very premature (less than 32 weeks) carries substantial risk of later life health problems, developmental delay, behavioural problems and educational difficulties. With an increasing number of survivors, and the limits of viability moving down to 22 weeks even in the NHS, a growing demand for health and educational resources can be expected. By bringing together data from health and educational datasets for preterm babies born in the last 14 years, Imperial College London hope to begin to evaluate and discuss the wider implications of prematurity at a societal level across health, education and employment in the future. This may help inform policy and national targets.

Access to the Data has enabled:

1) Report survival rates of babies born extremely preterm and determine influencing factors, which has informed and changed national guidance

2) Report educational and health outcomes of babies born preterm for babies born in the last 14 years in England and Wales; this helps with counselling and surveillance of outcomes

3) Access to data has helped analyses using novel causal inference techniques that is informing potential interventions in the neonatal period that can improve outcomes and family experiences