Sep 13 UPDATE: I have removed the Daily Test Positivity Rate chart. However, as it is a derived value – from published dashboard data – I will continue to track it in my Excel sheet. The Positivity Rate was calculated as follows (New Cases/Test Results)*100. However, this assumed that the Test Results being reported on a daily basis did not include repeat tests, i.e., persons who had been tested previously. Using the dashboard above, the positivity rate for daily tests as of September 11 2021 would be: (62/2100)*100 = 2.95% I will attempt to reach out to the Covid Communications Unit to find out if their test results include repeat tests, or consist only of persons who are getting tested for the first time.
The Barbados COVID-19 Situation Report, also known as the “Covid Dashboard,” is quite possibly one of the most important pieces of information currently being published. The above dashboard represents our Covid-19 situation as at September 11, 2021.
I am no stranger to Covid-19 data analysis. I have collected, compiled and analyzed data across the region on and off over the last several months (Click here to view previous COVID-19 related posts). However, I have never bothered to track dashboard data, especially as it changed over time. Given the recent surge in Delta-related cases in Barbados, I have decided to address this.
Similar to my analysis of murders in Barbados during 2020, my analysis consisted of five key steps:
- Defining the question – A) What do the dashboard numbers look like over a period of time and B) Can any other pieces of data be derived from what is presented on the dashboard? The time period under review was from July 3 to Sep 11 (2021).
- Collecting the data – The least sexy part of data analysis, but the most important. Relative to the murder analysis, the dashboard analysis was easier as there is only one ‘authoritative’ source of data (the Barbados GIS website). I reviewed GIS web pages containing dashboard images from Sep 11 to July 3. I then entered the data into columns in an Excel sheet as follows: Period, Test Results, New Cases, In Isolation, Deaths, Male, Female, Under 18, 18 and Over, Age Unknown, Total Test Results, Total Cases, Total Male, Total Female, Total Deaths, Total 1st Doses, Total 2nd Doses, Male, Female, % of Pop Full Vaxx and Source. I also added my own “check” columns. For example, values in the Male and Female columns should add up to the value in the New Cases column.
- Cleaning the data – Not much to clean here thankfully. However, there were some minor annoyances. For example, the Sep 4 dashboard New Cases value = ‘101’ with Male = ’39’ and Female = ’62’. My first check passed, i.e., the values added up to 101. However, when adding the age groups, Under 18 = ’26’ and 18 and Over = ’72’, the total was 98. That particular dashboard did not track ‘Age Unknown’. As I peered further back in time, I noticed that some dashboards weren’t tracking Age Groups and % Population Fully Vaccinated (but this obviously changed as the dashboard evolved and more data was added). In terms of inconsistencies, several dashboards (old and new) would break down the Total 2nd Doses Administered by Gender (i.e., Male and Female), while others did not.
- Analyzing the data – The exciting bit. The dashboard generally looks at a specific point in time, but I wanted to see what the data looked like over a period of time, to see the rise and fall of the numbers. Excel’s charts allowed me to see this. I also wanted to see what was not being published on the dashboard, but could be derived from it. This led to the creation of extra columns: Increase/Decrease of New Cases over Previous Day, Positivity Rates for Daily Test Results, Cumulative Positivity Rate, No of 1st Doses By Day, and No. of 2nd Doses By Day.
- Visualize and share the findings or results – This was done via my blog.
Figure 1 Description – This chart shows the Percentage of the Population with 2nd Vaccination (i.e., those who are fully vaccinated). This chart was created using values obtained from the dashboard (July 10 to September 11). On July 10 this stood at 26.8%, and on September 11 it was 35.6%. This represents an 8.8% increase over 64 days, or 0.13% per day.
Figure 2 Description – This chart shows the Total Population with 2nd Vaccination. It was created using values obtained rom the dashboard (July 3 to September 11). The total population value I used was 270,000 persons (270,865 to be exact). This population value was calculated based on the September 1 dashboard which stated that 94,803 persons had their 2nd dose, which represented 35% of the population.
Figure 3 Description – This chart shows the daily number of persons in isolation. It was created using values obtained from the dashboard (July 3 to September 11). A lull in isolations numbers, 16 persons between July 23 (187) and August 26 (203) was preceded by a slight increase of 83 over 12 days (between July 12 and July 23), and a steep increase of 516 over 16 days (August 26 to September 10).
Figure 4 Description – This chart shows the No. of Daily Tests Results. It was created using values obtained from the dashboard (July 3 to September 11).
Figure 5 Description. This chart shows Total Doses (1st and 2nd) Administered. It was created using values obtained from the July 3 to September 11 dashboards.
Figure 6 Description. This chart shows the No. of 1st and 2nd Doses By Day. It was created using values derived from the July 6 to September 11 dashboards. It is calculated by subtracting the number of Total Number of Doses (1st and 2nd) administered on a given day, from the numbers administered on the previous day.
Going forward, I will be updating my Excel sheet daily and posting weekly updates to this website every Friday. If you want to share the charts above, please remember to reference caribbeansignal.com as the source.