The goal of this project is to reduce maternal mortality and improve outcomes for pregnant women, new mothers, and infants in Nigeria. The maternal mortality ratio (MMR) in Nigeria was 545/100,000 live births in 2008 and increased to 814 deaths/100,000 live births in 2015 which is still unacceptably high.
The overarching aim of HelpMum is to help reduce MMR rates in Nigeria. The focus of this project is to process data that has been collected by HelpMum to gain a better understanding of the causes and outcomes of births taking place in rural Nigerian birthing clinics. There are two subgoals:
1) Process unstructured data collected by HelpMum into a structured format
2) Analyze the data to identify patterns in MMR across birthing clinics
HelpMum would also like to build a risk prediction model for mothers at risk of MMR. As of the outset of this project, it is not known whether or not the data already collected is sufficient for building risk models. To this end, the following subgoal will be accomplished:
3) Assess the feasibility for potentially building a patient-level risk prediction model for MMR and provide recommendations where appropriate
The data used for the project will be firsthand survey records collected by HelpMum from mostly rural maternity clinics across Nigeria. Each clinic responded to the following 6 survey questions:
How many pregnant women have you registered in the past one year?
How often do these pregnant women attend antenatal?
What are the challenges/danger signs they present or are faced with?
Did you refer them to the hospital?
If no, what are the solutions you proffered?
What was the end result of the situation?
In total, data was collected from 203 maternity homes owned by traditional birth attendants from across Nigeria. The data is stored in MSWord documents. Since the data contains healthcare information, a data sharing agreement will need to be signed that binds the volunteer team to strict confidentiality and privacy standards (note: the data does not personally identify any particular patients at maternity homes).
NPL will be used to transform the data into a structured dataset. A straightforward data analysis will be carried out based on analysis questions created by the volunteer team and the HelpMum staff.
Below are a sample of analysis questions:
How many pregnant women have been registered at clinics in the past one year?
How often do pregnant women attend antenatal?
What are the most common challenges and complications pregnant mothers face?
Which complications are most commonly referred to the hospital and what was the outcome?
A brief literature review will be conducted to see if findings about the causes of MMR in Nigeria are consistent with the projects findings.
The final deliverable of this project will be a report that contains the analysis findings, as well as recommendations for building an MMR risk prediction model. The code will also be handed over to the HelpMum team in a GitHub repository.