Categories
All posts Alternative Careers Digital Nomad Physician

Learning Data Analytics for Physicians

I know it sounds like a boring topic but you already are a data analyst – let’s start with that. You collect data, process it, analyze it, share it, and take action based on your analysis. Learning data analytics for physicians won’t be too difficult.

I have no desire to enroll in a master’s degree or Ph.D. program just so that I can learn data science. But data is one of the most valuable resources which is accumulating at an exponential rate. I want to know what to do with it and how it can improve healthcare.

Data Science in Healthcare

I am a family medicine physician who has done some consulting in the machine learning space. I have wrangled a lot of data but have very superficial knowledge. I want to figure out what I don’t know and whether it’s relevant to becoming a better clinician.

The cranial itch I’ve been having regarding this data process is that in big tech all the data collection and analysis is done by a separate department so that those who are best at taking action can concentrate on taking action. As physicians, we’re stuck doing all the work ourselves.

The whole point of data collection (gathering an H&P) is that I can make the right clinical decision for my patient. Maybe they need a TKR or they need to have their statin changed.

Physicians as data analysts get stuck collecting all the data, processing the data, and then having to take action on it. That’s way too much work for 1 person. I want to figure out how we can divide these tasks up so that the physician can focus on what they are most effective at or what they like doing most.

Data Analysis Process for Physicians

There is a rather strict process that most data analysts follow when it comes to processing data. It’s referred to as:

ask – prepare – process – analyze – share – act

Let’s translate this to how physicians approach data. So, we

  • find out what problem the patient wants to solve
  • gather a history
  • perform a physical exam
  • process our findings based on current research
  • analyze the findings as it relates to the patient
  • share it with the EMR and the patient
  • take clinical action

Here is a cool little data analysis workflow I found in the course I am taking.

data analysis steps for analyzing data

The Data Analysis Course

The point of this course is to train someone with no data science experience to obtain an entry-level data analyst position. It can take 6 months but you can blow through it in 1-2 months.

The course is put on by Coursera and Google is the instructor. You can find the link on the main course website. There are 8 courses in total you must complete to obtain a certificate:

  1. Foundations: Data, Data, Everywhere
  2. Ask Questions to Make Data-Driven Decisions
  3. Prepare Data for Exploration
  4. Process Data from Dirty to Clean
  5. Analyze Data to Answer Questions
  6. Share Data Through the Art of Visualization
  7. Data Analysis with R Programming
  8. Google Data Analytic Capstone: Complete a Case Study

The first section describes the work of a data analyst as collecting data, analyzing the data, and making recommendations to stakeholders as to how to effect the desired changes.

This first section is a 5-week course and there are small quizzes you take after each section. It’s a mix of video and interactive pages and writing. There is a discussion forum as well which I think is a good way to find like-minded individuals.

Motivation to Take a Data Analytics Course

I’m a nerd and proud of it. I like learning and I enjoy the challenge of medicine. And at this point in my career, I don’t think I am challenged with dispensing Macrobid for a UTI.

Instead, I want to know which of my patients with a UTI must be seen and who must have antibiotics for their infection. And I want to know which can skip it.

How do I come by this knowledge? Where is this information? Apparently, it’s all stored in large databases. I have no idea how to access them or how to manipulate the data even if I saw it in a big old spreadsheet.

I am not quite nerdy enough to want to sit and do the actual data analytics myself but once I understand the data analytics process as a physician I can cooperate with a data analyst to create such databases to answer these kinds of clinical questions.

Data Analytics Skills for Physicians

If you’ve been reading this far you’re as big of a nerd as I am. Here comes the juicy stuff, because data analytics skills can help you find new career opportunities.

Maybe you’re tired of seeing patients or you want a more location-independent lifestyle. Understanding clinical data or data in healthcare makes you a competitive candidate for work within healthcare startups and established companies.

Tech companies are trying to change the way we do business, not just in healthcare. Some of it is driven by greed but some are trying to improve healthcare. These companies need physicians who understand all the data left behind by patients.

Clinical data analytics is where we make sense of clinical data. It could be a patient’s weight or a patient’s particular genome sequence. And data is always dirty, it needs to be cleaned by someone who is an industry expert.

After the data is cleaned it needs to be interpreted. This interpretation requires creativity and insight into clinical medicine.

Physicians as Data Analysts

You don’t have to sit there and create a SQL query. You don’t have to rearrange data using R or Python. As a physician, you won’t even have to sift through all of the data on Excel and create a nice dashboard on Tableau.

A junior data analyst can do all of that for you. But you become the domain expert – the person who can help the junior analyst make sense of the data.

You can partly own and drive what will be done with this data once it’s cleaned. For example, I recently worked on an RPM (remote patient monitoring) project where a lot of data was coming in from patients. We needed to know how best to handle the clinical aspects of this. I can’t get into the details but imagine a patient’s heart rate is being monitored – what do you do when it’s high, low, irregularly irregular? You get the idea; it’s easy for you and me but not for a product lead or data scientist.

Data Analyst to Clinical Consultant

I’ve mentioned the term clinical consultant before. Anytime you are offering your expertise as a physician to a business you are a clinical consultant.

There are a lot of opportunities in this space. The best ones are found by you reaching out to various companies. Perhaps you find what they are doing interesting and think that you can offer some advice – that’s the perfect lead email or lead message on LinkedIn.

I have found a lot of my income-generating gigs on LinkedIn and Upwork. I am also offering up some gigs on Fiverr which hasn’t yet been lucrative.

Once you go down the Data Analyst pathway you learn new skills. And one thing leads to another and you learn about lean process improvement and efficiency consulting and so many other things. I didn’t even know such concepts existed.

The question to ask yourself is what you’re interested in and what you want to do on a day-to-day basis. Creating that kind of work and related income is just the next step. It requires some creativity and brainstorming with others who have achieved the same but it’s always doable.

As I write this I am ending my 1-year contract with a great company for which I consulted for. And I am starting consulting work with a new company whom I found myself on LinkedIn. I’ll keep you guys posted.

12 replies on “Learning Data Analytics for Physicians”

It’s on my list of books to get through on Goodreads. Seems on par with what the medical futurist publishes but still interesting.

Many medical groups will hire physicians who are both clinically active as well as work in the health informatics department. So it should not be a problem.

I’m a medical doctor and just like you found out how data can be useful for diseases, treatment, personalized health data and now I am thinking it could be an exciting path to follow. I just started a course, supported by my country’s goverment and let’s see how it goes. I think my background in medical field and having a data scientist skills can benefit alot.

I think there will be a lot of income potential in data in the future. I don’t see any healthcare system using data for the sake of patient health improvement and because data work is quite labor intensive it’s unlikely that any healthcare startup will be able to do much with it. However, it’s a great field to go into and will be quite lucrative it appears.
My personal passion is data modeling for prediction algorithms which could easily be built into any EMR. Naturally, it won’t. But I think it’s an interesting topic to follow.

Hi! Do you know how I found your blog? Through a conversational AI to which I was asking about how to work in data analysis as a physician. This is so, SO exciting. I’m just starting this path but I so do share your feelings. Thank you for the Coursera link, by the way, I’ll be taking it. I’m really looking forward to know more about your journey.

Hi Angela, thanks for your message.
Check out the articles I wrote about health informatics and healthcare consulting, in general.
I find that space to be quite fun and rewarding.
The data manipulation side can be a bit entry-level but you’ll learn a lot about wrangling data and obviously, as a physician, you’ll be much more suited to really categorize data or understand the natural anatomy and physiology of data – how they are interconnected with other data sets, etc.
This field is so immature still that you’re still well ahead of the curve. We are in a bit of a hype mode with AI, once that’s over there will be a dip in social media talking about it but that’s really when these things take off … so stick with it and ignore what others say but find your niche and your voice in this space.
Pardon the unsolicited advice, it’s what I would have liked to hear when I first started.

Hi, I am an early career Hospitalist and find data analytics and its application in healthcare very interesting! This post was very helpful.
Do you think the coursera courses are accurate or is it worth getting certifications via a university? Also how was your experience with upwork?

Congrats on the hospitalist career – that seems to be a good way to have some time to pursue other interesting projects on the side. I generally only advise pursuing a university degree if that’s the only thing blocking you from getting a particular job you’re afer. Otherwise, the course I described and some entry-level positions will provide you with all the information you need. I recommend connecting with other physicians in this space who are generally quite welcoming to get a sense of what they do.
I have some youtube videos on the topic as well I recommend watching which might give you some more insight.
https://www.youtube.com/@DigitalNomadPhysicians/search?query=data%20analytics

Hi Dr.Mo

Thanks for your writing. It was wonderful. I see that the best advantage of learning data is to switch or expand career.

I wonder if my goal is to be clinical physician and will try to involve in meta-analysis, clinical trials in the future. Will learning the data help me in the future? I feel like data and machine learning are the future and the world is moving so fast. I just feel like ordinary doctors will be left behind because it seems like those skillsets are not that important now. But in the next 2-5 years who knows right? 😆

I doubt that the average doctor will be behind because the workload taken off the shoulders of physicians will allow more space for pure clinical work. When it comes to doing research a lot of large medical groups are collecting their own data which will transform the way we do a lot of clinical research. Learning data will offer a lot more clinical opportunities in the future but also research study design and helping design better clinical algorithms.
Essentially learning how to manipulate data is the biggest skill someone can learn – which is a good amount of statistics and organizing and managing data through programs like R or Python or SQL.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.