Do ethnic minority groups have higher preterm birth rates? Evidence from CPRD data

Supervisors: Dr Sungwook Kim, Dr Catia Nicodemo

Background:

It has been well known that ethnic minority groups are more likely to have a higher risk of preterm birth. Although the NHS provides accessible health services, minorities sometimes face barriers such as language differences, cultural misunderstandings, or discrimination, which can affect timely access and health outcomes. This is a significant public health concern linked to ongoing health inequalities among ethnic minorities. However, current evidence is outdated and does not adequately reflect the new landscape after the pandemic, which has exacerbated health inequalities.

Given this, evidence exploring how minor ethnicity is associated with preterm birth rates in the post-pandemic era is needed to support ongoing maternity policy initiatives.

 Although Midwifery Continuity of Carer (MCoC), which is a key recommendation in the 2016 "Better Births" national maternity policy in England, targets ethnic minority groups in the most deprived areas, it struggles with limited funding and understaffing of midwives (House of Lords, 2024, NHS, 2021). Staffing shortages continue to disproportionately affect ethnic minority groups, where fewer women receive continuity of care (Dodsworth, 2023). Given this, it is needed for decision makers to make the best use and strategic allocation of existing resources; however, there is no clear evidence post-pandemic to guide prioritising specific ethnic minority groups or deciding which socioeconomic groups should be targeted. The evidence will also be useful for the successor programmes of ongoing initiatives, such as Core20PLUS5 (Dodsworth, 2023, NHS England, 2021) and the  Saving Babies Lives Care Bundle (SBLCB) v2 (NHS, 2019) as these initiatives aim to reduce inequality in maternity care by focusing on ethnic minority. This DPhil project aims to provide evidence for the given research question using health economic methods.

 Data:

Clinical Practice Research Datalink (CPRD) and Hospital Episode Statistics (HES) linked data will be used. The CPRD pregnancy register provides a variable for premature delivery, and gestational age is available through HES maternity tail data. This data is also appropriate to visualise the composite index considering a temporal trend. CPRD data is securely connected to patient postcodes through a privacy-preserving mechanism. This process employs the NHS Postcode Directory (NHSPD) to associate postcodes with Lower Layer Super Output Areas (LSOAs) across England. However, to maintain confidentiality, neither the original postcodes nor specific small area identifiers are disclosed to researchers. The updated IMD 2025 will be incorporated, which was recently released by the Indices of Deprivation team.

Brief description

This DPhil project consists of coherent and independent quantitative health economic studies.

Study 1: Economic costs of maternal health service uses for ethnic minority groups.

This study will indicate whether minority groups incur higher economic costs than ethnic majority groups. Resource use will be obtained from Hospital Episode Statistics (HES) data. Cost variables will be generated based on a standard approach, where the use of a service (resource use) is weighted by a value reflecting the opportunity cost, or ‘price’, of the service (unit cost). We have verified that all these variables are present in the dataset. Using this information, mean costs of maternal health service use for ethnic minority groups will be reported. Details of medications or devices issued by the GP, consultation type (e.g., telephone, home visit, surgery visit), and duration of inpatient stay will be captured and assigned costs.

Study 2: Impact of ethnic minority on preterm birth rates

This analysis will be conducted using a matching method based on socioeconomic variables including education, maternal age and marriage status to examine preterm birth rates among ethnic minority groups (Black, Asian, Mixed, or Other ethnic groups). Preterm birth rates will be estimated and compared for the years 2020 and 2025 to examine whether the impact of ethnic minority status on preterm birth rates was worsened by COVID-19. A subgroup analysis of ethnic minorities living in areas with different levels of deprivation, as measured by the Index of Multiple Deprivation (e.g., IMD quintiles 1 to 5), can be conducted to disentangle the impact of area-level deprivation.

Dependent variable: Percentage of preterm births per 100 total births. The analysis will start with a binary variable of preterm birth: Before 37 weeks vs. 37 weeks to term delivery. Further detailed analyses can be conducted with the following categorisation: Extremely preterm: Less than 28 weeks Very preterm: 28 to less than 32 weeks Moderate to late preterm: 32 to less than 37 weeks.

Study 3: Association between preterm birth and ethnic minority groups in deprived areas

In many cases, area level deprivation is not distinguished from ethnic minority groups for ongoing initiatives. For this purpose, we will use a logit regression to examine the association between preterm birth rates and ethnic minority groups. This will explore the association between 1) preterm birth rate and 2) ethnic minority and between 1) preterm birth rate and 3) the interaction term of ethnic minority*deprived area. Covariates are: (i) Clinical characteristics: child's sex, mother's age at delivery, multiplicity (twins, triplets, quadruplets) and parity (nulliparous vs. multiparous) based on the availability CPRD-HES linked data.

Preferred applicant background/skills

Health economics