A baby born before 37 weeks gestation is classified as a preterm birth and is the single most important factor in how likely a baby is to survive and what quality of life they will have. Preterm birth effects 8% of births in England and Wales and is estimated to cost the health service £3.4 billion each year.
Predicting who is at increased risk of preterm birth is important for planning supportive treatment to prevent it, or at least delay birth as much as possible so that care can be organised to be in hospitals that are equipped to give the best possible care for the mother, and for the baby after delivery.
There are several recognised risk factors for preterm birth, one of the most common being infections that happen during a pregnancy however, there is no evidence for women who have had significant infections before their pregnancy.
We propose, that like with an allergic reaction, when someone’s body excessively reacts to something it has previously come across before, a similar dangerous set of processes can occur for a pregnant woman who has had a recent significant infection before her pregnancy, making her have a higher risk of having a preterm birth.
We will estimate the risks of preterm birth associated with patterns of infections experienced by women in the five years prior to pregnancy, accounting for preterm birth risks that happen during pregnancy.
We postulate that a significant infective event prior to pregnancy can alter downstream signalling pathways that lead to priming and exacerbation of an inflammatory cascade, which in turn increases the rate of birth prior to 37 weeks gestation. Using the Clinical Practice Research Datalink (CPRD) general practice data with individual patient links to Hospital Episode Statistics (HES), including maternity HES information we will examine all pregnancies. We will extract outcome, delivery and birth details using mothers’ and children’s HES and primary care records to characterize timing and type of preterm delivery, e.g., spontaneous or other reasons/unexplained where possible. Using Read codes and ICD-10 codes in primary and secondary care data respectively, we will extract infections occurring within the 5 years before pregnancy, and characterize these as they present clinically (e.g., respiratory, urinary tract, vaginal, gastrointestinal, dermatological), the prescribing frequency of antibiotics or antivirals and the infective agent e.g., influenza, pneumonia, rubella, hepatitis, herpes simplex virus) wherever specified in the clinical record.
We will calculate incident rates of pre-pregnancy infections in the five years before pregnancy and use this to define exposure based on frequency and timing of infections in relation to onset of pregnancy and a cumulative load of infection. Gestational age at delivery and groups categorised by term of delivery (e.g. very/moderately preterm) will be compared between women with pre-pregnancy infections and those without, and will be adjusted as appropriate for sociodemographics (e.g., age, smoking, BMI, socioeconomic deprivation), pregnancy factors (e.g., parity, fetal and labour factors such as sex and mode of delivery) and clinical factors (e.g., pre-existing and gestationally presenting comorbidities). Groups will also be compared by other relevant exposure groupings according to the patterns of pre-pregnancy infections (e.g. frequency and timing). Cluster analysis will be used to account for women contributing more than one study pregnancy.
Gestational age of spontaneous (non-iatrogenic) birth; Term categorisation of spontaneous birth (extremely/very preterm/ moderately preterm/ term/ late term); Gestational age of any episode of threatened pre-term birth;
Gestational age of pre-term pre-labour rupture of membranes (PPROM).
Laila Tata - Chief Investigator - University of Nottingham
Mark Chester - Corresponding Applicant - University of Nottingham
Raheela Khan - Collaborator - University of Nottingham
CPRD Mother-Baby Link;HES Admitted Patient Care;Patient Level Index of Multiple Deprivation;Pregnancy Register