In healthcare, we focus a lot of energy on patient outcomes. This makes perfect sense since improving health – or curing people of whatever ails them – is the goal. But what if we took a similarly outcome-based approach to data? After all, despite all the talk about using data in healthcare, it seems like we could be doing more to ensure that there are positive outcomes from all we collect and analyze. Our guiding question should be: how can we find value in the data and then share our insights with our patients?
As you might guess, a few barriers stand in our way. For starters, patient privacy concerns and complexity drive access issues for healthcare employees. Only when we solve these challenges can we begin putting our data to work to improve the care we give and the experience we deliver to our patients.
My background is in biomedical science, forensic psychology, and clinical research. My passion lies in transforming raw data into meaningful insights that improve patient outcomes. I get a lot of satisfaction from taking raw data and making it understandable for as many people as possible.
When we talk about an “outcome,” it’s really many different elements coming together. Patient health always comes first, but patient experience is important too. Data can guide us in making adjustments that can radically improve the patient experience. If we could make patients happy during the time they spend in the hospital, we can have a different type of positive impact on their lives.
One great example of this is wait times. From point A to point B and point C, there are so many adjustments we can make to improve a patient’s experience. When I worked in a hospital setting, I always wondered why wait times were an hour when they should be closer to five minutes. How could you fix it? You could pull data from everywhere to determine the root cause of the delays. Is security taking too long? Are patient charts too slow to load? All these elements come into it. You need to minimize the noise to get to a better outcome.
Be sure to ask yourself frequently what you are tracking and why. Is it just health outcomes? Are people better health-wise at the end of their visit? Are we doing anything in one area that detracts from an outcome in another area? Every hospital sends out a five- or 10-minute patient survey so people can rate their experience. It can be a big ask, though, when you request some of a patient’s time for something that they think benefits you more than them. So that’s one of the improvements you can track. When you move from hospital to hospital, that gets disconnected. I’m pretty sure that there are still some smaller clinics that don’t even have electronic records. Some are still using paper.
Challenges in accessing healthcare data
Our goal should be to make the data in our organization as accessible as possible to the people who can turn it into insight. But because we operate in the healthcare setting, we quickly run into barriers:
- Informed consent and explanation to patients
- Lack of clear understanding of data usage
- Limited access to de-identified data
- Real-time and updated data requirements
These are all real concerns, but it is possible to cut through the many complicated layers that limit access. Let’s talk first about patients. They all sign IRB-approved consent forms. This is our first chance to clearly explain what information of theirs we are using and why. If their data is de-identified, it should be in a central place where you can query it. Even if you have patient outcome data on some forms or program on your computers, you still need access and know-how to get it out of the system. If you don’t have that access, you may rely on your IT team, which is another hurdle.
Sometimes data is all over the place and exists in a variety of formats. There’s no consistency. You may have databases that can’t talk to each other or some digital and some analog assets. The focus should be on a single outcome and all the variables around it. What are the steps to get from where we are now to where we can know enough to improve patient outcomes? The hard work begins with making process improvements, so all these elements come together in a single, designated data warehouse.
1. Unlock the power of operational data
Operational data can be a powerful tool for decision-making and improving outcomes. What do I mean when I say “operational data” in healthcare? I’m thinking of things like:
- Identifying inefficiencies in patient flow
- Addressing long wait times and improving services
- Utilizing patient feedback to drive improvements
One person loudly making the case for change can sometimes be a case of “squeaky wheel syndrome.” You might address the concerns of a single person and realize later that the problem they identified was far down the list or wasn’t a problem at all. You might even end up making a reactive change that later needs to be changed back. That isn’t culture change–it’s going in circles. Culture change happens when teams have equal access to data and can create well-informed change based on that data.
By sharing as much data as you can across the team, you get a better real-time understanding of what’s going on. Clearly define stakeholders to identify significant issues and use the data as your guide to solve them. If you can identify the biggest bottlenecks in your hospital, you can intelligently craft a solution. If a noticeable majority of your patients give negative feedback around the same issue, you can be sure that it’s one worth solving.
How do you carefully grant more people access to your data? It can feel very abstract to many, but you need to create a culture of proactivity. We need to get out of that abstract mentality of what we think we know to what we know from what we see. That tends to be how I operate in general. You should derive the outcome from the data. To confirm the outcome, you must look back and get through it. And along the way, you will know where, when, and what needs to be changed. Even little adjustments can do a lot. For example, you send a patient to floor one, then you send them to floor two, then you send them to floor one again to wait to go to three. Why not share space on the second level so they don’t have to come back down? Your small adjustment can make a big difference for people who aren’t comfortable walking the stairs.
2. Promote interest and collaboration
If your goal is to have a culture of data-driven decision-making, start by creating a climate of collaboration. Form groups and advocate for more general collaboration across the organization. Also, be sure to share success stories from around the business that promote the value of accessible data. Nothing lures people out of their silos like the promise of a better way.
Sometimes in the clinical setting, you might have one person speaking loudly while other, quieter participants keep some worthy ideas to themselves. For this reason, I recommend starting by sharing data in smaller groups. You can then grow the number of small groups, elevating the best ideas and growing the network from there.
When I started doing this work, we created dashboards – not too many at first – and the more we produced out of the dashboards, the more we communicated to colleagues to let them know what we’d studied and what we’d learned. We almost immediately started getting more requests from people saying, “Now that we know you guys can do this, how about we add this or that to it?” Their input only made our efforts stronger, either by refining something we already had covered or starting a new line altogether.
I can’t overstate how promoting interest and collaboration pays off over time. Like I said, the way we improve outcomes might start small, but that small can be pushed and pushed a little more until it becomes something bigger.
3. Improve data literacy across the organization
You can use your existing quality teams to get started on your data accessibility efforts. I used to be in a clinical setting before I moved on to quality operations. When I started, there were two of us in the clinical data management that handled data analytics. We went to potential interest divisions in department meetings, and people would say, “I wish we could see this instead of using a certain program that has to be updated all the time. If we could get help setting certain things up, we could get help linking the data.”
We helped them overcome the challenges of manual data updates. And we helped them build out their data visualization and communication capabilities. For example, no one wants to look at a table with 30 columns in it. So, we would show them, for example, how to navigate dashboards and their visualizations. We would show how to filter the disease you want to isolate, hide other ones, snapshot it, put it in your PowerPoint if needed, and be done. This method picked up momentum over time and spread to other divisions and departments.
Conclusion
When I did my undergraduate degree, data analytic-related classes and programs were extremely limited. But the younger generation now understands that data analytics is a great field. And they understand this is a degree you can get and take in several directions.
The good news in healthcare is that we now have enough people paying attention to data analytics. It’s not as scary as it used to be for people. It’s something anyone can learn if they’re interested. The more exposure people get to data, the more comfortable they are working with it, and the more we all benefit from it.
When we make healthcare data more accessible to the right people, good things happen. More well-informed decision-making. Better patient outcomes. A better, healthier world. What’s not to like?
Pang Chaoprang Herrera is Senior Clinical Data Analyst for Seagen, Inc.