Table 5 Multi-professional views (N = 27): synthesis relating to addressing current challenges to using data to guide improvement efforts aimed at reducing risk of perinatal brain injury.

1. Standardised definition of potentially avoidable perinatal brain injury

• Lack of a standardised, consensus-built definition of potentially avoidable perinatal brain injury limits the ability to effectively monitor incidence and risk factors.

• Frequently suggested definitions focused on hypoxic-ischaemic encephalopathy, which was seen as a potentially avoidable condition in ideal care conditions (Table 6).

• Related or other frequently suggested definitions referred to causes such as absent, poor, late or suboptimal clinical management during perinatal, antenatal or intrapartum care (Table 6).

• Questions relating to avoidability should consider whether the outcome would have been different in a different healthcare unit, with different healthcare professionals, or using different clinical practices, while remaining cognisant of the challenges in determining avoidability.

2. Agreed set of data items relevant for monitoring incidence and risk factors

• An agreed set of data items relevant for monitoring incidence and risk factors of potentially avoidable perinatal brain injury, including postpartum follow-up, is needed to generate an integrated national data source.

• A large range of risk factors and diagnostic features for potential use in an integrated data source can be identified (Table 7).

• Follow-up of babies with potential brain injury is vital, since injury or developmental challenges may present some time after birth, even if there is no early indication of injury.

• Long-term follow-up may be challenging, but might be mitigated through use of existing community services (e.g. health visitors) and parents’ reports.

3. Addressing inconsistency and subjectivity across data sources and healthcare units

• Duplicative recording of data items across multiple data sources is common but should be avoided.

• To enable and optimise an integrated data source, the inconsistency and subjectivity involved in capturing many data items relevant to brain injury (e.g. therapeutic hypothermia criteria, Apgar scores, fetal growth restriction, intrapartum risk factors) should be addressed.

• Inconsistency may be associated to use of different types of input fields in the various electronic patient record platforms used across healthcare units, and varying data definitions of similar data items across data sources.

• Subjectivity arises partly because units and professionals differ in cardiotocography (CTG) classifications, have difficulties in defining a standardised parameter for risk factors that can evolve during labour, and differently interpretate results of some diagnostic procedures, such as brain magnetic resonance imaging (MRI) and continuous electroencephalography (EEG) monitoring.

• Resolving some of the subjectivity would likely require more evidence on the robustness of clinical indicators and risk factors associated to perinatal brain injury.

4. Systematic linkage and integration of data items across data sources and platforms

• A programme of work is needed to operationalise and standardise operational definitions that are currently not harmonised across data sources and associated electronic patient record (EPR) platforms.

• Ways to improve linkage of the data sources may include incentivising EPR platform suppliers to facilitate data linkage, “nudging” more communication between neonatal and maternity teams who have access to different data sources, and linking the baby’s NHS number to the mother’s health record.

• Future steps for integration could consider linking perinatal data to post-perinatal data that are usually captured in separate paediatric data sources.

5. Organisational change to improve consistent, reliable and feasible data collection

• Systems change, socio-technical change and culture change are needed to enable consistent and reliable data collection that is useful for improving care.

• Change should start with an assessment on needs and required resources for training, data entry, data management, data analysis, and quality assurance.

• Change may be enabled by prioritising recording of clinical data over administrative data, e.g. about costs.

• Making the data collection system more user-friendly, for example by focusing on default settings, mandatory inputs and electronic (non-human) interpretations and enabling some pragmatism, may support improvement.

• Solutions may be found in making relevant data items mandatory in perinatal data collections, a coding framework for text-based data, and use of innovative technology for standardised interpretation of data instead of relying on subjective interpretation.

• Data should mostly be recorded contemporaneously (in real time as part of routine care) rather than retrospectively, to reduce risks of poor data quality and professional burden – this could be achieved in various ways (Supplement 3).

6. Engaging, training and funding healthcare professionals involved in data capture

• Engaging with all relevant healthcare professionals is needed to reach shared understanding about the rationale and importance of the data.

• Effective communication is needed to mitigate the risk that professionals may think that data could be used “against them” e.g. in the event of controversy.

• “Professionalising” data capture could be supported by training so that it becomes part of clinicians’ skillsets.

• Funding and resources are needed for data management and should accommodate adequate time, training, and resources for digital midwives, nurses, neonatologists, paediatricians, informatics teams, data quality managers, and others dedicated to lead or support reliable data capturing.

• Funding allocation should be subject to the number and complexity of data items collected.

7. Co-designing systems with healthcare professionals and families to improve use of data

• Co-design with healthcare professionals and families is needed to ensure data presentation and use is relevant, acceptable and becomes part of everyday practice, not an “extra thing”.

• Comparing unit data to national data generates possible risks of blame dynamics, since national benchmarking may not sufficiently consider local populations and challenges unique to each unit.

• Sensible comparative methods need to be co-designed with healthcare professionals and families – for example co-design of feedback loops that combine individual case reviews with “big data” comparisons and priorities of families.

• Data exchange from local to national entities should utilise existing auditors to support local-to-national data exchange, and not cause further strain on healthcare services and professionals. Some national oversight of feedback loops might be helpful, but data feedback should primarily be a local enterprise.

• Where possible, data should be accessible to anyone, not just to those at senior levels.