Malaria; How Safe Are You?
Exploratory Data Analysis of the Malaria Report
If you’re asked to mention 10 life-threatening diseases, you’ll talk about Covid19, HIV/AIDS, cancer, etc., right? All true, but what if I tell you malaria is one of the top 5 causes of deaths in Africa? Yes, you read that right! I bet that didn’t cross your mind.
According to the WHO Malaria Report, in 2018, there were an estimated 228 million malaria cases with 405,000 deaths worldwide. Noteworthy is the fact that Africa has the lion share of the malaria cases(93%) and malaria deaths(94%); thankfully, the disease is preventable and curable.
In this article, I’ll summarize how I carried out an exploratory data analysis(EDA) of the Malaria Dataset that contains a report of cases and deaths from different countries from 2000 to 2017. This EDA answers some insightful questions on the geographical prevalence and progression of these events.
Data Understanding and Cleaning
As mentioned in my previous article, every dataset has to undergo cleaning for it to be suitable for analysis.
The Malaria Dataset has 1,944 rows(108 countries * 18 yrs) and 5 columns. Some of the rows have empty values for the ‘No. of cases’ and ‘No. of death’ columns, probably due to poor documentation.
Because values for each country differ significantly, it’ll be impractical to use the mode/mean imputation technique(read about it here). Hence, the rows with missing values were dropped, since we don’t have sufficient information on that row.
Data Analysis and Visualization
After cleaning the data, I used Power BI to analyze and gain insights into the relationship between variables. Here are some of the questions I was able to answer.
- What is the trend of cases and deaths over the years?
From the graphs above, we see that the number of reported cases increase annually, even more rapidly from the year 2011. The number of reported deaths fluctuated with the highest number reported in 2010.
2. Are some regions better off than others?
Africa region reported the highest number of cases and deaths (88% and 93%, respectively). Europe is doing quite well; some countries in the region have not reported any death in years while more are getting close to eradicating the diseases.
3. What are the most and least affected countries?
The Democratic Republic of Congo tops the list of malaria danger-zones. Sadly, the top 10 countries with the most deaths are all African countries. About eight countries have not recorded any death in 18 years! That’s impressive and enviable!
Conclusion
WHO Africa region needs all the help it can get, in form of funding, research expertise and materials. Better data collection protocols should be put in place so that the whole picture can be seen; this will help afflicted countries get back on track in the global response to the disease.
Awareness should be created on the effect of antibiotics abuse, especially how it has resulted in the development of parasites resistant to once-effective antimalarial drugs.
Researchers also need to pay attention to some parasites starting to develop the capacity to evade detection by the most commonly-used rapid diagnostic tests, making it hard to diagnose cases accurately.
If you want to see the full gist in Jupyter Notebook, here’s the Github link.
Thanks for reading.
