DATA STORIES | DIGITAL HEALTHCARE | KNIME ANALYTICS PLATFORM
Innovation in Healthcare: Solving the Opioid Dosage Problem
My Data Guest — An Interview with The Winners of the Digital Healthcare Hackathon
Co-interviewer: Prof. Dayanjan “Shanaka” Wijesinghe, VCU School of Pharmacy
In today’s episode, we hosted Alneeda and Brandon San, the winners of the Future Health Pioneers hackathon held virtually on 17–18 June 2023. The hackathon was co-organized with Prof. Dayanjan “Shanaka” Wijesinghe, Principal Investigator and Director of the Laboratory of Pharmacometabolomics & Companion Diagnostics at VCU School of Pharmacy.
Different challenges in digital healthcare were given to the participants, and the winning project by Alneeda and Brandon tackled the opioid epidemic in the United States. Using KNIME, the students created a data app that contributes to medical safety by calculating the Morphine Milligram Equivalents (MME) in alignment with CDC guidelines.
Our Guests
Alneeda San is a public health student at the University of Washington. Her focus is Global Health, where she investigates several research topics, such as infectious disease, health disparities and strengthening health systems. Currently, she works as a pharmacy technician in training at Rite Aid retail pharmacy setting.
Brandon San is a doctoral pharmacy student and Masters of Business Administration student at Oregon Health and Science University in Portland, Oregon. The focus of his program is mostly clinical pharmacy. He has experience working in hospital administration, clinical pharmacy and managed care. Currently, he’s working at a local health system as a clinical intern and at a regional health plan as a managed care intern.
Shanaka: How did you hear about the Future Health Pioneers hackathon, and was this your first hackathon?
Alneeda: Yeah, this was our first hackathon. I learned about it from Brandon. He sent me a text message while I was studying for my finals. I found it quite interesting and agreed to participate in it.
Brandon: I was scrolling on LinkedIn one day and I stumbled upon it. I was really enthusiastic about the prospect of a hackathon which had data, technology and healthcare: a combination that you don’t commonly see. I said to myself, ‘Why not give it a shot?’. Hackathons are usually more coding based and you try to find vulnerabilities in the software. With this challenge, it was quite different. Even if we hadn’t won, it would have been a valuable experience.
Shanaka: What was your initial impression of KNIME Analytics Platform and of its low-code approach?
Alneeda: I found KNIME Analytics Platform really easy to use. I wasn’t aware of visual programming-based tools before this hackathon. When I thought of programming, I thought of blocks of scripts, like Python. Building the visual workflow with KNIME was very intuitive, as I could monitor the output of the nodes at each step. And whenever the output was not what I wanted, I could debug the workflow in a matter of minutes. Debugging is a notoriously hard task for blocks of code but with KNIME, it’s a lot easier. Finally, with KNIME visual workflow, I found it easy to explain to others what’s happening in the workflow and how the data flows from A to B. There is this visual element for them to see that enables faster and clearer communication.
Shanaka: Coming from a healthcare background, how was your experience tackling a challenge from your area with a visual-programming based tool like KNIME?
Brandon: KNIME was an excellent tool to tackle this problem, especially for healthcare providers like us. Programming is not a common part of our curriculum. The approach that we took was particularly valuable to me, as it allowed us to write down our ideas on paper, sketch out our workflow and replicate it on the software. This is very transferable and easy to do, which is not always the case.
Shanaka: So far, you guys worked on a few versions and prototypes of your solution. What was the biggest insight or skill that you gained from the challenge?
Alneeda: The biggest insight that I gained by participating in this hackathon was mainly the confidence to transform an idea into an actual, real application. I’ve always thought that the development of data-driven applications was the exclusive realm of software engineers or computer scientists, not public health majors. After participating in the challenge, I realized that I can bring an idea and make it come to life.
Brandon: The most significant skill or insight that I acquired from this challenge was the capacity to develop healthcare solutions. I realized that lack of coding expertise is not an obstacle. I’ve never contemplated a career that melts healthcare and data science together but now I’m actively thinking about it and exploring the field. Ultimately, for healthcare providers, this hackathon provided the opportunity to dive into new avenues within the field and explore data science, technology and pharmacy combined.
Aline: Now let’s talk a bit about your solution, the Opioid Dosage Calculator that you prototyped with KNIME to improve safety to opioid access. How did you choose this topic? Does it somehow connect with research projects you do in your programs?
Alneeda: We started out just brainstorming ideas. When Brandon brought this idea up, we both found it very applicable to everyday life. I work in a retail pharmacy setting so I see this barrier everyday: people squeezing in time between work shifts to come to the pharmacy, making sure their kids are being taken care of by someone, waiting in a long line, and eventually by the time they come to the front they realize they need the pre-authorization by their doctors. We saw the importance of this issue and how much potential it has to improve people’s lives, so we decided to tackle it.
Brandon: Opioid epidemic is prevalent in the news and it has a devastating impact on people’s lives, communities and economies. This raises the concern of pharmaceutical practices and the need for effective treatment. Another challenge derives from physicians or prescribers not having access to the proper and accurate information that each and every health plan has, which was the second part of the project we wanted to tackle.
Aline: The opioid crisis is an incredibly important topic these days, which drew our attention to your project. Can you explain how the opioid calculator works?
Alneeda: Essentially, in our calculator, providers input the strength, the frequency and which opioid they will be using. Each opioid has a different conversion factor, which is multiplied by the strength and the frequency. This creates the morphine miligram equivalent (MME) value. If the value is above 90, then the program tells you that it’s not approved per CDC guidelines. If it’s below or equal to 90, then the program tells you that it is approved and asks the provider to continue the preauthorization coverage verifier, in which they will be given information on the steps for the preauthorization after inputting the insurance plan.
Aline: After the hackathon, you kept improving your solution and modifying the data app. What modifications have you implemented so far?
Brandon: We developed our calculator based on the CDC prescriber guidelines. However, we went beyond by creating another MME calculator, as there are numerous available such tools online. With the enhanced version of the original calculator, our goal was to establish a more concrete linkage between clinical criteria, insurance requirements and prescriber guidelines to streamline the access to care. We improved our workflow and its visual appeal, making it simple for providers to determine if the medication meets the ideal MME requirement. This way, healthcare providers can have easy access to information that they need to make well-informed decisions.
Aline: Given the relevance of the topic you chose to work on, do you folks have plans to continue to develop this project?
Alneeda: Yes, we definitely do. We became very passionate about it and it’s something that we see high importance in. My personal goal is to see it implemented and running in the real world. Right now, I’m at a ceiling in terms of how far I can improve the project. So we decided to transition into learning more advanced aspects of data science and KNIME in order to further apply what we learned to the project.
Shanaka: How would you explain to other students your experience creating a data app for the first time?
Alneeda: It was hard at first. When you have an idea for a project without knowledge of the software, it’s going to be hard in the beginning, especially considering that the hackathon only lasted two days. Through it, you gain so much experience, you learn to adapt and shape the workflow while learning the tool itself. KNIME has a fairly fast learning curve and if you approach the problem with an open mind, it becomes a lot more seamless.
Brandon: There is an initial learning curve, but that’s expected when you venture into something new. KNIME turned out to be a really easy to use platform for data app development. I’m thrilled to have discovered it during the hackathon. KNIME presents a practical alternative to Excel or physical calculators, which are commonly used in clinical settings by pharmacists. I highly encourage students to explore low-code platforms like KNIME. I was pleasantly surprised how straightforward it was to use.
Shanaka: Who did you have in mind as the end users of the opioid calculator, and how did this impact the design of the data app?
Brandon: When Alneeda and I are started this project, we were thinking that the end users would be prescribers but it turned out that they could be also patients, healthcare providers and even insurance companies. Design-wise, we implemented a friendly user interface with clear and concise information. Perhaps, in the future we can integrate it with electronic health records. The most important thing is to incorporate regular updates because regulations and guidelines can change. It’s also important to include a feedback mechanism so we make sure that there is an ongoing refinement of the calculator. This calculator could be used by anyone, there is no restriction to it.
Aline: Can you tell us about a specific feature, node or component in KNIME Analytics Platform that was particularly helpful during the hackathon?
Alneeda: The most valuable resource for me was the KNIME Community Hub. It’s a pillar of the KNIME open source community, which I really loved. I could easily download and inspect other people’s workflows and get inspired. Another great feature is the drag&drop functionality that allowed me to effortlessly import workflows, nodes, extensions or components from the KNIME Community Hub onto my KNIME Analytics Platform. With Python you would have to download packages, take care of dependencies, ensure environment compatibility across OSs, etc. With KNIME, this is very easy and seamless.
Aline: How did you split up the tasks for the hackathon, and how did you leverage your different strengths?
Brandon: Alneeda and I engaged in a collaborative brainstorming session to develop the calculator. I drew upon my clinical pharmacy background to ensure that the calculator aligns with the necessary guidelines. Additionally, it was important for us to design a user-friendly application for the end user. With her coding experience, Alneeda took the lead on the KNIME workflow creation and I took the lead on the beginning and ending aspects of the calculator. We fostered an open-minded environment that allowed us to listen to each other’s ideas.
Shanaka: Can you tell us about the most challenging aspect of the hackathon and how you overcame it? And how did you keep up your motivation?
Alneeda: Translating our ideas onto a concrete workflow was challenging at first. We didn’t have any familiarity with KNIME so it was hard for us to figure out where to start. We learned as we progressed. Whenever we weren’t sure how to proceed, we’d talk about the problem out loud. This allowed us to know what’s going on in our minds.
Shanaka: And the most rewarding aspect?
Brandon: We were so ecstatic when the workflow and the calculation worked for the first time, for it had been a long process before reaching that result.
Aline: Winning a hackathon is a big deal! And the main award we offered you was a Summer Internship at KNIME in 2024. What do you folks expect from this experience?
Alneeda: I’m really excited for the internship! I’m looking forward to gaining exposure to data science and a deeper understanding of KNIME Analytics Platform. Hopefully, I will be able to apply this understanding to my future project. I’m also eager to make meaningful contributions to the company through my work.
Brandon: I’m very ecstatic about this opportunity. I’m excited to get my feet wet and to see if this is something that I’d like to do in the future. I’m also excited for the hands-on learning, mentorship, project experience and learning all the amazing things that you guys are doing.
Aline: Was there a “wow” moment where you couldn’t believe that you could do such a thing with a low-code tool like KNIME?
Alneeda: My “wow” moment was the creation of a data app using the component interactive view. After wrapping everything in a component, I opened the view and was able to see all of my progress and interactive elements. I thought it was just great and super easy. This is something that I wouldn’t be able to do in Python.
Brandon: My “wow” moment was the whole hackathon in general. Coming in with not so much experience or knowledge of KNIME, being able to complete a calculator that sets clinical criteria and somehow winning the hackathon was unbelievable.
Shanaka: Do you use any resources (e.g., publications, books, online journals, social media, etc.) to keep yourself up-to-date about the newest trends in the digital healthcare space?
Alneeda: I try to read one non-fiction article every morning, usually health based. I think that the most important resource is having people around you that are in similar spaces as you are. When I discuss these topics with fellows, I learn the most.
Brandon: I’m a member of AMCP and they talk a lot about innovation, healthcare and digital health. I get emails from them everyday. What’s most insightful is going to their conferences, which they hold twice a year. The best learning for me happens when I talk to people within the field and ask them where the field is going.
Aline: Before we say goodbye, we have one last question. When are you graduating and what are your plans for the future?
Alneeda: I will be graduating next year and after that, I’d like to further my academic career in pharmacy school. While there, I will keep an open mind on what to focus on.
Brandon: I have two more years of pharmacy school and my postgraduate plans continue to evolve.Currently, I’m considering a path in the managed care or pharmacy industry but I’m also excited to research how data science roles align with my future goals.
Aline: Many thanks to Alneeda, Brandon and my co-host, Professor Shanaka, for the great conversation and for making us aware of how data analytics and KNIME can be used to help in the opioid epidemic.
Watch the original interview with the Winners of the Digital Healthcare Hackathon on YouTube.
