Mode of delivery for childbirth in the UK Clinical Practice Research Datalink (CPRD) and Hospital Episodes Statistics (HES), 1998-2020: investigating the usability, completeness and temporal trends in mode of delivery data

Date of Approval: 
2020-09-24 00:00:00
Lay Summary: 
- The CPRD pregnancy register is a large file containing information on almost one million pregnancies occurring between 1998 and 2020 from around the United Kingdom. - This register is a very helpful resource for people studying pregnancy. - The register contains anonymous, non-identifiable data on each pregnancy. We know if there was a miscarriage or an abortion. If the baby was born, we know if the baby was born alive or was still born. We also have an estimated start and end date of each pregnancy. - At the moment, we do not know what type of birth the mother had. In other words, we cannot see which mothers pushed their baby out, whether instruments like forceps were used to assist the birth, and which had a Caesarean section. This information would be helpful for people who want to research how the type of birth is associated with the health of the mother and her baby. - We aim to develop a way of defining type of birth data to the register. We will use data held in the mums GP and hospital records. We also plan to describe how type of birth methods have changed over time.
Technical Summary: 
The CPRD’s pregnancy register is a database of more than 5.8 million pregnancies, including 833,165 pregnancies ending in a delivery, between 1998 and 2020. This large-scale register is a valuable resource for researchers interested in pregnancy related epidemiological questions. The register contains information on the pregnancy, including its outcome (live birth, still birth or early pregnancy loss), the timing of the pregnancy (the estimated start and end date), and additional details such as whether it was a singleton or multiple pregnancy. The register currently does not provide information on whether the delivery was a vaginal delivery (VD), an instrumental VD, or emergency or elective Caesarean section (CS), which is a key variable for understanding the health effects of CS and instrumental VD, particularly in the context of increasing rates of CS. We therefore aim to enhance the pregnancy register by developing an algorithm to define mode of delivery, using hospital data and primary care data. We will analyse the completeness of mode of delivery data, use Cohen’s Kappa statistic to estimate the level of agreement between delivery data from different sources and calculate the rate of indications for, and complications of, deliveries to assess the data’s reliability. Finally, we will describe how delivery methods have changed over time.
Health Outcomes to be Measured: 
Primary outcome: An algorithm that will derive mode of delivery (Non-instrumental VD, Instrumental VD, Elective CS, Emergency CS). A further outcome is testing the validity of this derived variable. Secondary outcomes: Using the algorithm to derive, and testing the validity of, both a cruder (binary) and more refined 21 level categorical variable of mode of delivery, and the World Health Organisations Robson’s classification system (which was developed as a global standard for assessing, monitoring and comparing CS rates)1: - Binary mode: VD or CS - Delivery with finer categories: 0=VD vertex (top of the baby’s head comes first) 1=VD other cephalic, unassisted* 2=VD other cephalic, assisted* 3=VD other cephalic, assistance unspecified* 4=VD breech, unassisted* 5=VD breech, assisted* 6=VD breech, assistance unspecified* 7=VD presentation unspecified, unassisted* 8=VD presentation unspecified, assisted* 9=VD presentation and assistance unspecified* 10=VD instrumental, low forceps 11=VD instrumental, high/mid forceps 12=VD instrumental, unspecified forceps 13=VD instrumental, vacuum 14=Elective CS, upper uterine segment (UUS) 15=Elective CS, lower uterine segment (LUS) 16=Elective CS, unspecified 17=Emergency CS, UUS 18=Emergency CS, LUS 19=Emergency CS, unspecified 20=CS Unspecified 21=Other *Assisted refers to manual manipulation by a healthcare professional Additional outcome: Planned (as opposed to actual) mode of delivery.
Application Number: 
20_146
Collaborators: 

Harriet Forbes - Chief Investigator - London School of Hygiene & Tropical Medicine ( LSHTM )
Harriet Forbes - Corresponding Applicant - London School of Hygiene & Tropical Medicine ( LSHTM )
Caroline Minassian - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Deborah Anne Lawlor - Collaborator - University of Bristol
Jennifer Campbell - Collaborator - CPRD
Krishnan Bhaskaran - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Liam Smeeth - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Melanie Griffin - Collaborator - University of Bristol
Sadie Mullin - Collaborator - University of Bristol

Linkages: 
CPRD Mother-Baby Link;HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation;Pregnancy Register