Kenya National Bureau of Statistics
Institute for Health Metrics and Evaluation (IHME)
The African Covid-19 Vulnerability Index (ACVI) is replicated from the Covid-19 Community Vulnerability Index (CCVI). The purpose of the index is to identify the geographic regions less resilient to the impacts of Covid-19. The index is built off the Centres for Disease Control and Prevention’s (CDC) Social Vulnerability Index (SVI). As defined by SVI, social vulnerability refers to the resilience of communities when confronted by external stresses on human health, stresses such as natural or human-caused disasters, or disease outbreaks.
The African Covid-19 Vulnerability Index examined the CCVI themes and indicators and adapted them to the African context. For example, the lack of water, sanitation and hygiene are major problems experienced in Africa and contributes to the risk of Covid-19, which is why these indicators were included in the index for African countries.
The methods and calculations used in the ACVI are taken from the CCVI methodology and SVI data documentation respectively.
The ACVI is calculated for various African countries covering four themes:
Identifies communities with health issues that make them more vulnerable to the effects of Covid-19 such as HIV, Tuberculous and diabetes.
Covers topics such as health worker density and number of hospital beds among others.
Identifies areas with a high density and high percentage of elderly people aged 65 and over who are more vulnerable to the effects of Covid-19.
Identifies areas with no access to water and sanitation.
The data used in the index has been sourced from various surveys, reports and estimates listed here [link to data sources page].
All raw data and calculations can be found here [link to openAFRICA with all ACVI data].
A detailed list of indicators used are represented in the table below:
|Epidemiological Factors||Cardiovascular conditions||South Africa||Cardiovascular Prevention (prevalence of non-raised blood pressure)||District Municipality||District Health Barometer, 2018/19|
|Epidemiological Factors||Respiratory conditions||South Africa||Respiratory infections under 5 year olds||District Municipality||Institute for Health Metrics and Evaluation (IHME), 2017|
|Epidemiological Factors||Diabetes||South Africa||Diabetes prevalence 15 years and older||District Municipality||South African Health Review, 2019|
|Epidemiological Factors||HIV/Aids||South Africa||HIV prevalence estimates 15-49 year old||District Municipality||Institute for Health Metrics and Evaluation (IHME), 2017|
|Epidemiological Factors||Tuberculous||South Africa||Incidence per 100,000 population||District Municipality||District Health Barometer, 2015/16|
|Epidemiological Factors||Tobacco smoking||South Africa||Percent smoking tobacco||District Municipality||District Health Barometer, 2018/19|
|Healthcare Systems||Number of hospital beds||South Africa||Public sector hospital bed density per 1000 uninsured population||District Municipality||South African Health Review, 2019|
|Healthcare Systems||Number of healthworkers||South Africa||Public sector ealth professionals per 100,000 uninsured population||District Municipality||South African Health Review, 2019|
|Healthcare Systems||Health insurance||South Africa||Medical scheme coverage||District Municipality||South African Health Review, 2019|
|Healthcare Systems||Access to medicine||South Africa||Proportion of public sector health facilities with essential medicines||District Municipality||District Health Barometer, 2018/19|
|Healthcare Systems||Access to pharmacies||South Africa||Public sector pharmacists per 100,000 uninsured population||District Municipality||South African Health Review, 2019|
|Demographics||Age||South Africa||Population by age group||Local Municipality||Community Survey, 2016|
|Demographics||Population density||South Africa||Population density||Local Municipality||Community Survey, 2016|
|WASH||Access to sanitation||South Africa||Main type of toilet facility used||Local Municipality||Community Survey, 2016|
|WASH||Access to water||South Africa||Main source of water for drinking||Local Municipality||Community Survey, 2016|
|Epidemiological Factors||Cardiovascular conditions||Kenya||Population reporting having heart problem||County||Kenya Integrated Household Budget Survey, 2016|
|Epidemiological Factors||Respiratory conditions||Kenya||Population reporting having respiratory infections||County||Kenya Integrated Household Budget Survey, 2016|
|Epidemiological Factors||Diabetes||Kenya||Population reporting having diabetes||County||Kenya Integrated Household Budget Survey, 2016|
|Epidemiological Factors||HIV/Aids||Kenya||HIV prevalence estimates 15-49 year old||Subcounty||Institute for Health Metrics and Evaluation (IHME), 2017|
|Epidemiological Factors||Tuberculosis||Kenya||Population reporting having TB||County||Kenya Integrated Household Budget Survey, 2016|
|Healthcare Systems||ICUs||Kenya||Covid-19 Treatment Centers||County||Covid19 Treatment Centers, 2020|
|Healthcare Systems||Ventilators||Kenya||Percent type of emergency breathing intervention||County||Kenya Harmonized Health Facility Assessment (KHFA), 2018/2019|
|Healthcare Systems||Number of hospital beds||Kenya||Number of hospital beds||Subcounty||Kenya Master Health Facility List, 2020|
|Healthcare Systems||Number of healthworkers||Kenya||Core health workforce per 10,000 population||County||Kenya Harmonized Health Facility Assessment (KHFA), 2018/2019|
|Healthcare Systems||Health insurance||Kenya||Percent with health insurance||County||Kenya Integrated Household Budget Survey, 2016|
|Healthcare Systems||Access to medicine||Kenya||Percent of health facilities with essential medicine||County||Kenya Harmonized Health Facility Assessment (KHFA), 2018/2019|
|Healthcare Systems||Access to pharmacies||Kenya||Number of pharmacies||Subcounty||Kenya Master Health Facility List, 2020|
|Healthcare Systems||Access to pharmacies||Kenya||Number of Community Health Units||Subcounty||Kenya Master Health Facility List, 2020|
|Demographics||Age||Kenya||Population by age group||Subcounty||Kenya Population and Housing Census, 2019|
|Demographics||Population density||Kenya||Population density||Subcounty||Kenya Population and Housing Census, 2019|
|WASH||Access to sanitation||Kenya||Type of toilet facilities||Subcounty||Kenya Population and Housing Census, 2019|
|WASH||Access to water||Kenya||Main source of water for drinking||Subcounty||Kenya Population and Housing Census, 2019|
A vulnerability score has been created for each of the four themes at each administrative level. The overall ACVI has been calculated using the vulnerability score calculated for each theme. Each indicator was ranked against all the geographies in that particular administrative level. The rankings for each indicator were summed up for each geography and the sum was ranked against all geographies. The resulting value is the vulnerability score used to identify the vulnerability rating.
The overall ACVI consists of adding up the vulnerability score for each theme and ranking the sum against all geographies. The resulting value was the vulnerability score used to identify the vulnerability rating. This process was done for each administrative level.
All themes were weighted equally as based on the CDC SV model. The vulnerability rating is classified according to the CCVI model:
Code for Africa’s data team sorted through each country’s government statistics and national surveys and reports to find the required datasets. As this was a relatively large data collection task involving various countries, Google Sheets was used to track the status of the process. This allowed the assignment of tasks to team members and changed the status of the tasks easily.
A lot of the datasets were in pdf, but for the purpose of this project the datasets needed to be in machine-readable format. We scraped the data tables from the pdf reports using tools such as Tableau and Cometdocs to generate .csv files.
Some datasets were downloaded in spreadsheet format from interactive data dashboards, such as the Statistics South Africa SuperWEB.
The datasets often needed additional cleaning as scraping tools aren’t always entirely accurate – and to remove data not required. The datasets were cleaned using Google Sheets.
Each dataset was then uploaded as .csv files to openAFRICA. From past experience, reports or datasets are sometimes taken down or misplaced on a website resulting in broken links. This problem is solved by uploading all datasets used on openAFRICA. openAFRICA built by Code for Africa is Africa’s largest open data portal that allows an user to upload, search and download datasets in various file formats.
The dataset needs to be arranged in standardised format in order to be visualised in HURUmap. The dataset is required in long data format and the appropriate geographic data, such as the administrative level and codes need to be added. For this formatting purpose, an open source online tool called Workbench was used. We created a workflow with the relevant steps to produce the data into the required format.
Each dataset is presented into a separate .csv file which is uploaded to the PostgreSQL database. Once the datasets are uploaded to the database, the data visualisations were easily created using the HURUmap visual plugin on the website’s dashboard. The HURUmap visual plugin has features to select the dataset, chart type and chart title which creates the visualisations seen on the website.