Return to the Main Page.

Return to the Schedule.

Return to the list of potential projects.

Association of HIV with Causes of Mortality in Baragwanath

Overview

Dating back to classical work by John Graunt in the 1600’s, analysis of death records has a long history of providing insight into important epidemiological patterns, including the impact of major infectious diseases.

This group will focus on analysis of death data from a large urban hospital in South Africa.

Things to consider

• This group is recommended for:
  • Participants who would like to gain experience with data analysis
  • Participants who have an interest in understanding disease burden in urban settings
  • Participants who are interested in understanding the interactions between infectious diseases (e.g., HIV and TB)
  • Modelers who would like an opportunity to collaborate with clinical researchers
• This group will have the opportunity to engage in any of the following:
  • Statistical analysis of a large dataset to evaluate the association between HIV infection status and TB mortality
  • Discussions with clinical researchers about the dataset and its analysis
  • Development of process and observation models that may yield insights into biases inherent in the dataset and improve interpretation of the findings
• Potential group members are encouraged to review the following sessions before Week 2:
  • Public Health, Epidemiology, and Models (Monday AM)
  • Introduction to Thinking About Data II (Track A, Tuesday PM)
  • Study Design and Analysis in Epidemiology (Track A, Thursday AM)
  • Participatory coding for Variability, Sampling Distributions, and Simulation (Thursday PM)
• In addition the following sessions during Week 2 will be essential for this group:
  • Working with databases: management and manipulation in R (Tuesday PM)
  • Model assessment (Wednesday PM)

Background

South Africa has the greatest number of people living with HIV in the world, but detailed data on causes of death in those infected with HIV are still lacking. From 2006 to 2009, Dr. Andrew Black collected detailed data on the causes of death for everyone who died at Chris Hani Baragwanath Hospital, a large urban hospital that mainly serves the township of Soweto. This dataset is unique in Africa and will allow us to make important inferences about the relationship between HIV and other diseases in South Africa. During the period under investigation, more than 14,000 deaths were recorded. A previous publication describes the overall death rates, by age, gender, and HIV status. This project will extend this work to investigate the relationship between all causes of death and HIV status by age and gender. This group will engage with Dr. Black (via Skype), who will provide essential medical insights into the analysis and the interpretation of the data.

Data

• Data are available for individuals who died in the Medical Wards of Chris Hani Baragwanath Hospital between 2006 and 2009. The information available includes:
  • Age in years
  • Sex
  • Cause of death (ICD-9 code)
  • Date of hospital admission
  • Date of death
  • HIV status (positive, negative, or unknown)
  • For HIV positive individuals, whether they were on antiretroviral therapy at the time of death
• The data to be used for this project are owned by Dr. Andrew Black, who has generously made them available for use at MMED. Use of these data for any purpose beyond the analyses conducted as part of the MMED 2016 group project is strictly prohibited, unless prior written authorization is provided by Dr. Black.

Resources

References

• Black, A, J Kriel, M Mitchley, and BG Williams (2015) The burden of HIV in a Public Hospital in Johannesburg, South Africa. arXiv:1512.04781 [q-bio.OT] (download)

Tutorials

• Tutorial 4: Visualizing Infectious Disease Data in R
• Tutorial 5: Data cleaning and management in R
• Lab 3: Study Design in Epidemiology