Clinical Data Manager to Clinical Data Scientist—the Skills You’ll Need to Master in the 21st Century
posted: April 9, 2020
Because of the changes in the way studies are designed, and the paradigm shift in what is being collected and how, the clinical data manager role is in the process of evolving into a clinical data scientist.
With that evolution comes a need for those who work with clinical data to develop new skills, and also put greater emphasis on skills that may have been less important up to now.
Data managers have always needed to be extremely detail-oriented and to have therapeutic area-specific knowledge. There are also a handful of other competencies that have not always been seen as critical to the role, but which will come to define the more desirable and successful data manager/data scientist.
Data managers need to be able to communicate effectively, with their immediate team members, with sponsors, with CROs and other vendors. They need to understand what the sponsor is trying to accomplish with the study so that they can design the most effective database. While the best data managers have always known this, there was a time when DMs could get away with not having good communication skills because they mostly interacted with the data, not the people. This is not really true anymore, and the ability to communicate clearly will only become more important as the world of clinical data continues to evolve.
So what does a great data manger need to understand about communication?
Communication is more than just the words you speak or write. It’s the tone of your voice, how you say something, and why you say it. What is the intention behind your message? When do you choose to say something – during an argument, or in a quieter moment?
What you don’t say is just as much a part of your communication, and is sometimes louder and clearer than what actually comes out of your mouth.
Your body language speaks too, with facial expressions, the gestures you use, and an open or closed posture.
Great data managers will want to develop their listening skills too, as they build their communication skills. Listening is as much a part of communicating as talking. When you practice active listening, you listen twice as much as you speak, and you tend to learn far more.
To improve your listening, be actively engaged and interested in the other person. Don’t interrupt. Don’t be thinking about what you want to say next. Reflect back what you think you heard in order to make sure you really understand. Watch their non-verbal cues. And ask questions. Open-ended questions (“What..., How...”) elicit thoughts and feelings and encourage elaboration. Closed-ended questions, starting with “Did...” “Do...” “Would...” get you specific information or yes/no answers.
Data managers already need strong project management skills, and this is only going to become more critical in the near future.
There are several important components to project management. Strong project management abilities include:
Leadership: You need to be able to identify and share the vision for your project, create a roadmap to keep everyone on track, empower and motivate your team, while removing barriers and smoothing interactions.
Communication: See above.
Planning Skills: Whether you use digital tools, collaborative tools, or some other method, you’ll need to be able to schedule the elements of the project to keep it on track.
Organization: This seems obvious, but when you are project managing there are so many different things to juggle you must have an organizing method that works for you. Sometimes this combines with the planning tools, above.
Risk management: Everything we do involves risk of some kind, so we need to plan for events and problems. If you’ve planned ahead you can react faster when problems arise, and often get small issues handled before they grow into big issues.
Problem solving: Like with risk, there are always going to be challenges and issues. You need to be able to think clearly and come up with solutions quickly.
Sense of humor: This may seem out of place, and you don’t need to be a comedian, but having a sense of humor, being able to laugh and to make your team laugh, helps keep things in perspective. A good laugh reduces stress as well, and, as a bonus, can improve your ability to problem solve!
Critical thinking is one of the most important skills data managers are going to need to develop. This informs the other skills we are talking about here, and separates average DMs from great DMs.
Critical thinking includes the following:
Curiosity: When you are curious you are never content with your current understanding of the world. You are driven to raise questions and pursue answers. The more you know about a topic, the more you realize there is to learn, and DMs need to keep learning.
Humility: This is related to curiosity and recognizing your own limits. If you are arrogant and “know” everything, you have no reason to be curious. Humility makes us more receptive to new information.
Skepticism: Demanding evidence for a claim is healthy and an important part of critical thinking, rather than just accepting what someone says at face value. We also need to apply this skeptical approach to our own thoughts and beliefs.
Logical reasoning: Logic is indispensable when we are talking about critical thinkers. Skepticism keeps us on the lookout for bad arguments, while logic helps us figure out why that argument is bad. Or good, for that matter.
Creativity: I define this as the ability to come up with new combinations of ideas. It’s not enough for us to just be skeptical and knock holes in every argument we come across. To really be a critical thinker we need to come up with solutions.
Empathy: Rounding out critical thinking is empathy; being able to see things from another person's perspective. This expands our horizons and gives us greater range in ideation and solution generation.
Subject Matter Expertise
How do we help our organizations or teams to move along? Data mangers need to do the work and gain knowledge, to understand the details of clinical trials in general, and the common challenges. To be seen as a subject mater expert, you need to:
- Have a documented history working in field
- Have actually done the work
- Understand industry best practices / tricks of the trade
- Understand user needs
The landscape is changing and it's really important to be aware, as a data scientist, that we are going to need new tools in our toolkit. Here are a handful to focus on:
R Programming: This is becoming more popular, about 43% of data scientists are using R now, and it’s being adopted in a lot of different organizations as a data cleaning tool. It is a robust, open-source language that is both affordable and effective.
Data Visualization: A picture is worth a thousand words. As data scientists we need to "see" the data and put it into visual tools like Spotfire and Tableau. Data visualization gives an organization the opportunity to work with the data directly, and more quickly grasp the insights that will help them make decisions.
SQL Database: Data mangers/scientists are going to need to be proficient in SQL. It's designed to help us access, communicate with, and work on data. It gives us insight, and when you use it to query the database you learn more and more. Knowing SQL will help us understand the database structure and look at the data from a different perspective.
Python: We are seeing the emergence of Python as an important tool. It's a great programming language for data scientists. Right now about 40% of data scientists are using Python as part of their programming language set. It's the versatility that we see with Python that allows you to do almost anything and pull in different sources of data, from database to Excel spreadsheets, which really speaks to our jobs as we deal with more and more sources of data. It can pull in the data and then create the different types of analyses that we need to see.
Evolve and Grow With the Industry
New study designs, additional data collection tools, the advancement of biomarkers and genomic data, larger sets of data, requests to analyze the data as it comes in, all of these are affecting clinical data management. As leaders in the field we have to start evolving into clinical data scientists now, ahead of the curve, if we want to survive.