Beast Cancer

America often gets knocked for having some of the worst healthcare for a developed nation. Is that totally the case? The answer is muddier than it may look. We set out on our Hackalytics journey by looking for a research topic that we wanted to explore more about. In this case, we wanted to learn how America's private health care system could stack up against those around the world globally. We started with what we believe constituted a robust healthcare system. Starting with the first point of contention which was spending, we deduced that America spent an egregiously large amount of money compared to other nations, demonstrating that economic efficiency was low. Following up with more data sets, we were able to pull together and overall manipulate and merge data for four more factors which included number of hospitals doctors, average life expectancy, diagnosis screenings per capita, fully vaccinated populations, and infant mortality. For each factor that we considered, we believe that they offer a glimpse into every facet of an effective healthcare system. For instance, the number of diagnosis screenings per 1000 people are a fantastic metric for access to care because diagnosis screenings are often used for preventible diseases and are therefore a symptom of widely accessible healthcare. In addition, the average life expectancy and infant mortality rate are fantastic metrics for the actual outcomes of the each countries healthcare because a stronger healthcare system naturally would support and positively buffer those two statistics. Following both the access to healthcare and its robustness, we wanted to include a wildcard factor which were the density of doctors in order to explore the possible issues of brain drain between countries. Under our assumption, countries were a higher density of doctors meant that the areas overall prioritized their respective healthcare systems and thus gave an incentive for doctors to work in. Finally, with the aforementioned cost of healthcare, we can say with confidence that these five factors overall are as close as possible based on available data to map out the different aspects of healthcare.

Going into more about the data analytics. Not only did we graph fiscal spending with each of the five factors which represented the monetary efficiency of healthcare systems by country, we additionally made a multi variable linear regression model to help succinctly describe our data. Essentially, using the formula

y = c1x1 + c2x2 + c3x3 + c4x4 + c5x5,

We aimed to normalize the data values in order to represent all of the research we had collected. One of the most mathematically intensive parts was normalizing the data as we first had to clean the outliers and then use the normalize function provided by the pandas library. However, that still begs the question, how do you compare the number of doctors with life expectancy as they are two completely different values? Essentially, under each factor, we ranked each country and assigned a numerical number between 0 and 1 (the actual statistics and math can be found in the Github). Then, because we understand everyone has different preferences in a healthcare system, we deliberately made it so we can manually alter the weights of each factor for the viewer/judge. As a result, this would give a non-biased answer as it is perfectly reasonable to have a different value system.

Moving onto to our final deliverable which is the Wix Vevo website, we wanted to emphasize our value system which gives more control to the user. Thus, we aimed to implement interactive graph but ultimately fell short due to the new wix environment and the time constraint. In the mean time, we have a graph of an example of what we want where there is all the factors listed as out as well as the weights where the user can (hopefully, with more time) alter. Currently, this feature is still a proof of concept. Our website also includes our youtube video which will not be discussed in this write-up as it will have its own criteria. Continuing on, the bulk of the website demonstrates different graphs that we made using pandas and matplotlib from datasets including OECD and the WHO that generally map out the US's wasteful expenditure with the results of their healthcare system. Finally, as another proof of concept, we have embeded Google Maps utilizing Google Map's API model which we want as a way for users to learn more about the healthcare system through specific examples. The goal is to work with different hospitals in order to chart reality with different scenerios if the United States operated under a Universal Healthcare system. Overall, we really enjoyed working on our first ever Hackalytics Hackathon and we learned a lot from cleaning data to data visualization.