Children's Intensive Care Outcomes and Health Inequities: A Mixed Methods Study of Social Determinants in Paediatric Intensive Care in England

Overview

This mixed methods study aims to understand how factors such as ethnicity, deprivation, language, and access to healthcare affect outcomes for children admitted to paediatric intensive care (PICU). The main questions it aims to answer are: * Why are some children more likely to be admitted to paediatric intensive care, or to have worse outcomes once they are there, depending on their ethnicity or background? * At what points in a child's healthcare journey could earlier support or intervention make a difference? The first part of the study will involve interviewing parents and carers of children admitted to intensive care and health care professionals working in or referring children to intensive care. The second part of the study will be an analysis of existing NHS data. Participants will have an interview with a researcher lasting 45 minutes to an hour.

Qualitative study: Aims To explore, using individual interviews, the experiences and perspectives of a purposively sampled diverse cohort of parents of children admitted to PICU and Health Care Professionals (HCPs) working in or referring patients to PICUs Objectives • To understand the experience of patient families and health professionals of PICU admission, including barriers (for example language, financial, access to specialist care, systemic healthcare biases) that can be faced leading up to admission, experience on the unit, examples of good practice and perception of factors that influence outcomes This work package will comprise a qualitative study in which semi-structured interviews will be conducted with families of children who have been admitted to PICU and HCPs who work in PICU or refer children to PICU. Setting Potential parent participants will be identified by Clinical teams (e.g. PICU liaison nurses or other local staff) at Great Ormond Street Hospital (GOSH), Birmingham Children's Hospital, St Mary's Hospital and Royal Manchester Children's Hospital. These four hospitals are in geographically, ethnically, and socioeconomically diverse areas. Sampling (patient families) Recruitment will occur over 12 months in four units. Purposive sampling will be used to recruit families who have a child aged 0-16 admitted to PICU. Feedback from our PPI activities has informed our recruitment strategy. Families will not be approached in the initial phase of their admission as they will be overwhelmed and will not be able to take in information. Staff from the direct care team who are not otherwise involved with the research study (e.g. PICU liaison nurses or other local staff at Great Ormond Street Hospital, Birmingham Children's Hospital, St Mary's Hospital, and Royal Manchester Children's Hospital) will screen patients using local admission history to identify potentially eligible parents/carers of children admitted to PICU. They will identify parents of any child who fits the sampling criteria and provide them with information about the study during their admission, but not in the first 48 hours of admission. The study information poster will also be displayed in the unit (including in family rooms and waiting areas) and handed to eligible families by staff. The poster will direct families to the study website, where they can access the patient information leaflet (co-produced with the parental advisory group) and consent to contact form. Participants who agree to be contacted about an interview will complete a consent-to-contact form. At this stage, participants will provide their contact details (telephone number and/or email address) and limited demographic information, including ethnicity and socioeconomic indicators, first language and limited details about their child (age, if they have any long term medical conditions and whether this is their first PICU admission) collected to support purposive sampling. Consent to contact forms will be completed via REDCap, with personal details stored securely on Imperial College London servers. Data from the consent to contact forms will be deleted at the end of the recruitment period, or within 12 months of collection, whichever is shorter for both those who do and do not go on to participate in the main study. Parents/carers will be contacted by the study team to provide more information, answer any questions, and arrange a suitable time and place for the interview if they wish to participate. Informed consent will be obtained electronically via REDCap prior to interview. At the start of the interview, the researcher will verbally confirm with the participant that they are the individual who completed e-consent and that they remain happy to take part and for the session to be audio recorded. The REDCap form will be completed by the participant using a unique link sent directly to them, providing a record that consent was given by the specific individual. No study procedure will take place without consent being in place. A sampling matrix will be used to ensure diversity in terms of Family socio-demographic characteristics (ethnicity, socioeconomic information, first language) Child characteristics (age of child, presence of comorbidities, multiple PICU admissions in the previous year) The sampling matrix is a form of stratified sampling that will allow us to select families that represent a broad range of experiences and backgrounds. Our PPI group was clear that bereaved parents were an important group to hear from. Prior research has shown that bereaved parents' experience of participation in research is generally positive. Our PPI group emphasised it was essential to have a very sensitive approach with all parents, ensuring that they did not feel "blamed" or the investigators didn't make them "relive the trauma". They suggested approaching bereaved families once they had time to process what had happened, recognising that not all families will be at the same point in their journey. Based on prior work by supervisors bereaved parents will be approached 6-12 weeks after discharge with approach made by the clinician following up the family, with the aim of including interviews from 3-5 bereaved parents. A sample of interviews facilitated by an interpreter with families who do not have English as a first language will be done, with the aim of recruiting 5 families who do not speak English as a first language. Participant information leaflets and consent forms will be translated. The study team discussed whether any families should not be included, for example looked after children, or families with ongoing complaints against the hospital. Parents agreed it was important to hear from a diverse range of families, but this would need to be approached sensitively. Children with ongoing safeguarding investigations will not be included, but if parents have parental responsibility children who have a history of safeguarding concerns or being looked after will be included. Sampling (HCPs) HCPs working in PICU or referring patients to PICU will be recruited by sending an email to an identified professional at each of the participating units for circulation to the wider Healthcare Professionals groups with information about the study. HCPs will be invited to express interest in participating in the study via an online consent to contact form in which details will be collected on sampling criteria (ethnicity, socioeconomic position and professional group) which will be used to ensure diversity in participants. Forms will be completed on REDCap, with personal details stored securely on Imperial College London servers. Data from the consent to contact forms will be deleted at the end of the recruitment period, or within 12 months of collection, whichever is shorter for both those who do and do not participate in the study. Participants will be contacted by the study team to provide more information, answer any questions, and arrange a suitable time and place for the interview if they wish to participate. At the start of the interview, the researcher will verbally confirm with the participant that they are the individual who completed e-consent and that they remain happy to take part and for the session to be audio recorded. The REDCap form will be completed by the participant using a unique link sent directly to them, providing a record that consent was given by the specific individual. No study procedure will take place without consent being in place. Interviews Individual semi-structured interviews lasting approximately 45-60 minutes will be undertaken. At the beginning of each interview, participants will be asked to give their consent for participation and for audio recording of the session and they will have the opportunity to ask any questions. At the start of the interview, participants will be asked to provide limited demographic information to support data analysis, including their ethnicity and socioeconomic information. Parents will also be asked for their child's age, presence of comorbidities, and number of previous PICU admissions; healthcare professionals will be asked about their professional background. The study team will ensure all participants understand that interviews are confidential, and they will not be identifiable in any published results or reports. The study team will keep a reflective diary following each interview, particularly noting anything that might not be captured in the audio-recording and the investigator's perceptions and experiences of the interview. For face-to-face interviews, an encrypted digital dictaphone will be used. For remote interviews, Zoom (via a secure institutional account) will be used. Audio files will be kept only until transcription and accuracy checking are complete, after which they will be permanently deleted. If a parent does not consent to audio recording, they will not be excluded from the study. With their agreement, the researcher will take written notes during the interview instead. The participant will be advised that this may limit the ability to capture their account verbatim, but every effort will be made to accurately reflect their views. This ensures that individuals are not excluded solely because they prefer not to be recorded. Setting Families - face to face at a place of the participants choosing (ie private room in hospital or at their home) unless they particularly request an online interview HCPs - virtually or in a private room at the hospital Our PPIE group was clear that parents needed to be supported during the interviews, as they could potentially be distressing. The study team will make it clear to participants that they can stop the interview at any time, take a break or reschedule. Recognising that participants can sometimes find it difficult to stop an interview, a stop signal will be agreed with parents before the interview begins. A distress protocol will be used. The study team will signpost parents to relevant support services, which may include hospital psychology, GP and support groups. Bereaved parents can be signposted to bereavement services at the hospital and organisations such as Cruse. In addition, there is a possibility that interviews may highlight instances of perceived poor practice. Participants will be informed in advance about the limits of confidentiality. If information is disclosed that raises safeguarding concerns or indicates a risk of harm to a child or other individual, local safeguarding and governance procedures will be followed. If a participant described something that reflected poor care this would not be shared unless harm was involved but it would be reported in study findings. Data collection: interviews Demographic information will be collected from participants. Interview topic guides will be developed for each participant group (parents and healthcare professionals), informed by existing literature on health inequities, expert input, and feedback from the parent advisory. Co-development work with the parent advisory group will be required to develop the parent/carer interview topic guide; however, this cannot begin until the PhD has commenced, and the resulting materials will then be submitted as a substantial amendment. Interviews with parents will include Their child's healthcare journey leading up to and during their PICU admission to understand their experiences and challenges or things/people that helped (timely access to GP, availability of appointments, ability to get repeat prescription) Knowledge barriers (knowing where to access health care, understanding symptoms to seek help for) Financial barriers (access to transport, housing issues, ability to provide adequate food, heating) Practical barriers (child-care for siblings, other family health issues) Communication issues (feeling listened to by staff, barriers with language) Experience in PICU (positive and negative examples, perceptions of care, treatment by staff) Experiences of discharge and support in the community Suggestions for how care could be improved - and which changes feel most important to families Interviews with Healthcare Professionals will include Experiences of providing or coordinating care for children before, during, and after PICU admission Perceptions of how ethnicity, deprivation, or social factors influence a child's health trajectory and outcomes Barriers to effective care delivery (e.g. time pressures, communication challenges, gaps in services) Examples of good practice and models of care that have worked well Challenges in working with families with complex social or linguistic needs Views on communication between services, and between families and professionals Experiences of discharge planning, community follow-up, and continuity of care Suggestions for improving care and reducing inequities, including priority areas for change Half of the HCP interviews will be conducted prior to the parent interviews as exploratory interviews to understand HCP perceptions of possible contributory factors behind increased incidence of PICU admissions and worse outcomes in PICU. Questions will be asked about potential issues around communication that might make it challenging for staff. Participant reimbursement: Recognising that participants will be recruited from groups that are often under-represented in research, parent participants will be reimbursed for reasonable out-of-pocket costs incurred in taking part, such as childcare or travel expenses, up to a maximum of £50 (calculated at £25 per hour for up to two hours). Quantitative study: Aims To explore the mechanisms by which ethnicity and deprivation contribute to disparities in PICU admissions and subsequent outcomes using routine data. Objectives To create a birth cohort linking four national datasets: Hospital Episode Statistics (HES) data and ONS, PICANet data, and the Maternity Services Dataset (MSD). To characterise the population of children who end up needing hospital admission vs those who end up in PICU, compared to the general population, focusing on the impact of ethnicity and deprivation. To use statistical techniques, including mediation analysis, to understand how factors captured in routine data, such as comorbidities and healthcare use patterns, mediate the known ethnic and socioeconomic deprivation disparities in PICU admissions and outcomes (mortality or unplanned readmission). Dataset creation: Applications will be made to NHS England to link HES (which collects data on 97% of all births in England) and Maternity services data (MSD) to make a birth cohort of all infants born in England between 2003-2025. The infants can then be followed-up via consecutive HES admitted patient care episodes which will give details on any hospital admissions and ED presentations after birth. This will give information on rates and reasons for admission. The research team will obtain approvals from the Health Research Authority (HRA), an NHS Research Ethics Committee (REC), and the Confidentiality Advisory Group (CAG) to use unconsented, routinely collected national data. Applications will be submitted to NHS England and PICANet/HQIP for access and linkage permissions. All datasets will be securely stored at Imperial College London. NHS England will identify the cohort using HES Admitted Patient Care (APC) and A\&E datasets from 2003 to 2025 and MSD from 2015 to 2025. NHS England will provide a pseudonymised dataset, including clinical and mortality data from HES and ONS, to Imperial. Separately, PICANet will provide patient identifiers for all children with at least one PICU admission between January 2003 and 2025 for linkage to HES APC. PICANet will assign the same pseudonymised study IDs already used for the standalone datasets. NHS England will then send the pseudonymised PICANet record IDs for successfully linked records to Imperial. Data flow diagram. The final dataset will be stored at Imperial College London's Big Data \& Analytical Unit (BDAU). The BDAU SE, is an ISO 27001:2022 certified research environment and compliant with the NHS Data Security and Protection Toolkit (EE133887- BDAU). Outcomes: Hospital admission PICU admission Intensity of care (PICU length of stay or duration of organ support) Mortality in PICU Unplanned readmission after discharge from PICU Data sources: HES is a national Hospital admissions database covering England which captures details of roughly 99% of NHS funded hospitalisations. Child socio-demographic information (age, sex, ethnicity, area where child lives for deprivation scoring) Comorbidity information (more completely captured than in PICANet alone) Information on Hospital admissions (APC) and Emergency Department presentations (A\&E) PICANet has been recording PICU admissions since 2003. The dataset includes Child socio-demographic information (age, sex, ethnicity, area where child lives for deprivation scoring) admission information (elective, date of admission, admission diagnosis, some comorbidity information) Information about care received in PICU (invasive ventilation, extra-corporeal therapies) Length of stay and mortality information Maternity services data Maternal socio-demographic information (age, language preference, ethnicity) Maternal health information (BMI, smoking status) Family information (IMD of area where family live, number of other children) Maternal psychosocial risk factors By linking multiple national datasets, this study will offer a comprehensive view of children's healthcare trajectories from birth through to adolescence. It represents an important first step in understanding how ethnicity and socioeconomic deprivation contribute to unequal outcomes in paediatric intensive care, with a particular focus on the role of chronic medical complexity. The findings will directly inform subsequent qualitative research and support the development of targeted interventions to address these disparities.