5 Must-Have Skills to Become a Healthcare Data Scientist

Healthcare organizations all over the world continue to collect huge volumes of healthcare data by the minute. Who is going to be the individual to sift through the noise and turn this raw data into actionable information?

The healthcare data scientist!

Companies such as Verily, Health Catalyst, Clarify to name a few, are companies focused on using data to improve patient outcomes. If this is of interest to you, you might want to apply for a role of a healthcare data scientist. In this blog post, I am going to detail 5 important skills to obtain when looking to become a healthcare data scientist.

1. Analytical skills
   
   Analytical skills are quite important when it comes to working as a healthcare data scientist. Oftentimes you will see on job postings, experience working with R or Python, experience with data visualization tools such as Tableau and Power BI and experience writing SQL queries. These technical tools are important in analyzing, cleaning, visualizing and finally presenting the data to an audience. Therefore, proficiency with these technical tools is key to analyzing the data.
   
   
2. Strong statistical foundation
   
   Working in this field, a strong background in statistics is important when it comes to developing machine learning algorithms. Although there are many open source graphical user interfaces that can be used to quickly input your raw data into your machine learning model of choice. It’s important to have the statistical background to understand if you should use a logistic regression or decision tree classifier on your data. It’s also important to understand how to measure significance and if the results you have obtained are statistically significant. Another consideration, is that with healthcare data, oftentimes missing data is quite prevalent whether from patients not answering questions in a medical survey provided or the clinical staff forgetting to enter in the data. As a result, in your analysis you will have to deal with missing data in a statistically sound manner.
   
   
   
   
   Credit
   
   
   
3. Experience working in a multi-disciplinary team
   
   Working as a medical or healthcare data scientist you might find yourself working with a diverse group of people, from clinical staff to IT. Therefore, it’s important that you can take in various opinions and criticism of your work and make adjustments to the algorithm you are building when you are presented with clinical expertise. Many of these algorithms are built to help aid the clinician and therefore, should take into account their experience when building these technical tools.
   
   
   
4. Strong communication skills
   
   When it comes to presenting your findings, if you are working in healthcare most likely you will have to present your technical findings to a non-technical audience. One of my favourite series on YouTube is by Wired called 5 Levels, the premise of the video is for an expert to explain a concept to 5 different types of people: a child, a teen, a college student, a grad student, and an expert. It’s one thing to understand the concepts you have learned it’s another thing entirely to explain and be able to communicate your ideas to a diverse audience. When using healthcare data you might need to explain that the results of your analysis can help clinicians improve care if they carry out x process, if you want someone to change the way they have done their practice for a long time, it is critical that you are able to explain your why in a clear and rational manner.
   
   
5. Experience with clinical data
   
   You might see on multiple job postings that many of these medical or healthcare data scientist positions are looking for individuals who have some experience in the healthcare field. This can be experience with common EHR platforms such as, Epic or Cerner. Experience with clinical coding systems (e.g., ICD9/10 diagnosis codes) or experience working with large clinical datasets. Healthcare data can be quite unique in comparison to other industry specific data such as financial or geographical data. Oftentimes, you might need to understand the clinical background to make sense of the numbers and understand if the trends that might be trending downwards or upwards is positive based on the clinical question being asked.

I hope this overview of a mix of both technical and soft skills is helpful to you on your journey to becoming a healthcare or medical data scientist. Comment down below, what would be your dream company to work for, as a medical/healthcare data scientist?