Applying Human Centered Design to IV Infusions: Interview with Dr. Beth Kolko, CEO of Shift Labs

With the aspiration to address healthcare gaps around globe by applying human centered design to medical technologies, Shift Labs, a 2015 Y Combinator company, is first tackling the infusion market with DripAssist. Formerly known as Drip Clip, as reported previously by Medgadget, DripAssist automatically calculates an IV infusion drip rate, total volume of the infusion, and can set an alarm to alert a nurse when the drip rate changes or stops. DripAssist has been in the veterinary market since 2014 and has been available for human use thanks to an FDA clearance beginning earlier this year. We had a chance to sit down with Dr. Beth Kolko, CEO of Shift Labs and Professor at University of Washington, to learn more about DripAssist and the incredible strides Shift Labs has taken over the past couple years.

Mike Batista, Medgadget: Let’s start with a little bit of your background and what led you to the creation of Shift Labs…

Dr. Beth Kolko: I’ve been a professor for over 20 years. I actually started my career as a rhetoric/literature professor. I studied the early days of the internet – back before it had pictures! In those days, the internet was all about reading and writing, so it made sense for someone with my background to focus on research internet technologies. Eventually, I got to the point where I was doing work that examined how technology design affected whether and how different groups of people used it. At some point, I realized I could either keep writing about the limitations of technology and its design, or I could help to change how it’s built. With some luck, I moved over to engineering, so I could help make better stuff. Because engineers build things!

Fast forward another 10 years and a decade of work in the developing world which eventually focused almost exclusively on healthcare related projects. My rhetoric background was informing the design work of those projects. To my mind, anything, including an ultrasound machine, could ultimately be seen as an act of communication between the user and the machine. We built all kinds of prototypes, digging into the soul of technological artifacts and shaping them so that they spoke to their users in clear, intentional ways.

I worked with amazing students and colleagues on a vast array of projects in different countries. But again, I was running up against the limitations of “impact.” We were very successful in academic terms with publications and grants, but I wanted to really push innovative technology into the market – into emerging markets. It turns out that commercializing healthcare products for emerging markets is tricky. Business models usually hinge on larger ticket items or consumables — things that don’t serve the needs of users in emerging markets. So I stepped back and thought – what if you started from scratch, with a human-centered design perspective from the ground up – and built a business that would have a laser focus on creating products that people wanted and that served their needs — rather than assuming current ways of doing business were the only ways of building a business.

 

Medgadget: What sparked Shift Labs’ interest in IV administration leading to the creation of DripAssist?

Dr. Kolko: We started the company with a vision, but not a core technology to build out. Also, we had no cash. But we were committed to this idea of building affordable medical devices that clinicians really wanted. So we leveraged our relationships and past experience in healthcare facilities. I sat down with several doctors and nurses who had worked around the world, and we basically did an ethnography via interview. We talked in depth about their challenges in providing care, what was hard, where were the gaps in patient safety, etc. Then I compared notes across all of those clinicians, and I looked for commonalities – where was everyone struggling. And it turned out everyone was frustrated with the current landscape of IV infusion. There just seemed to be this giant gap in available technologies that would be lightweight and work seamlessly across multiple environments.

There were actually about five or six areas that were pain points for all the people I talked with, not just infusion. Once we identified those, we sat down with a bunch of hardware hackers, and asked “which of these could we solve cheaply.” There were a handful of technologies where we saw a path to an inexpensive solution. IV administration was one of them, and it was the one that seemed to have the largest market potential, so that’s where we focused.

 

Medgadget: Now that you’ve created DripAssist in response to that need, how does the device work?

Dr. Kolko: The device is quite simple on the surface. It’s a handheld monitor that slips onto the drip chamber of an IV set. The infusion is set by the nurse using the roller clamp that comes with the drip set, as with any gravity drip. Once the IV begins, however, the nurse no longer needs to count drops and calculate a flow rate. In seconds, DripAssist reads the drops and shows flow rate and volume, displaying the information on a large, easy to scan screen. DripAssist also includes an optional alarm that lets you know if the drip rate changes or stops. This makes DripAssist the only way to continuously monitor a gravity infusion.

DripAssist uses infrared to detect drops, and the electronics convert that information into a flow rate and volume display. The secret sauce is in the design and the software. Gravity infusion is incredibly erratic, and we worked for a very long time to make a device that would give nurses usable, actionable information – while maintaining 99% accuracy.

 

Medgadget: Given your background both teaching and researching human centered design, how did human centered design play a role in the development of DripAssist?

Dr. Kolko: Human centered design was at the core of all our decision-making around DripAssist. Even before we had a product, the work to determine what technology to build was informed by human centered design. Once we started building, we used human centered design to keep our focus on users and what they wanted. I won’t lie – it was a struggle at times. There were some core engineering decisions that were made much harder because of that human centered focus. Our choice of battery was driven by human centered design – and it made our mechanical engineering work a lot harder. Probably the biggest challenge was our determination to keep DripAssist as all one piece that didn’t need to be mounted on a pole. That was core to the users’ delight with our earlier prototypes, so we really wanted to stick with that design feature, but it posed lots of challenges, including getting the device to fit the many different dimensions of tubing sets. I argued with a lot of engineers over that one.

 

Medgadget: I see DripAssist is also marketed for veterinary use. Which came first and why did you decide to tackle two use cases at once?

Dr. Kolko: We always intended DripAssist for the human market. But we did an early crowdfunding campaign to raise initial cash, and that attracted the attention of a small veterinary distributor. The distributor reached out to us and said he thought there would be real interest in the veterinary market. At the time, he basically had a customer who wanted a drip counter and he didn’t want to build one for them, so he went online to try and find one. He found us.

At the time, we were still in the prototype stage, but he kept pinging us and eventually we had a product to show. So we went to a big veterinary exhibition with him to see what vets thought and whether we should go ahead and make a few hundred to sell to them. We had a great response, so we launched on the vet market to get early customer feedback while we pursued the regulatory process for human sales. Our experience in the vet market has been really great. That distributor I mentioned is an engineer, and he makes the very best ventilators in the business. So being associated with him as we entered the market was great. But he’s also one of the nicest people I’ve ever met, and he’s counseled us throughout the process – everything from engineering tweaks to packaging suggestions. Also, vet shows almost always have adorable animals on hand, so that’s a win.

 

Medgadget: Is there any difference in the technology between human and veterinary use cases?

Dr. Kolko: The core technology is the same. There are cosmetic and labeling differences, and right now they are manufactured at difference facilities since there’s quite a bit of overhead in maintaining the standards required for a medical device. This allows us to sell the veterinary product at a lower price.

 

Medgadget: You mentioned the crowdfunding campaign, I know it was mentioned here on Medgadget before, how did that end up going?

Dr. Kolko: We ran the Indiegogo campaign back in 2012, it was actually one of the first, if not the first, medical device to be crowdfunded. The campaign raised $12,000 and was the first money into the company. On that money alone we were able to go all the way from concept to working prototype that we were able to bring to India. Overall the campaign was a great way to not just raise a little cash but also get exposure and connect with clients, manufacturers, and distributors.

 

Medgadget: DripAssist seems like such an incredibly simple yet effective tool. Has anyone tried solving the challenges of IV infusion before?

Dr. Kolko: It is incredibly simple! Listen, we didn’t spend a decade in the lab building DripAssist. It’s not a sexy new nanoparticle or implantable or robotic interface – and let’s face it, those things are cool. There’s no reason someone else couldn’t have solved this problem earlier from a technological standpoint. What we did differently was take the problem space seriously, and then stay focused on a solution that put users’ needs first.

People have tried solving IV administration challenges, and there are fantastic technologies out there that solve big parts of the problem. Volumetric infusion pumps, syringe pumps, elastomeric pumps, etc. Those are all terrific technologies and they solve some major problems. What we did that was different was look holistically at the infusion space and identify the needs that weren’t being met. You could stand back and assume that eventually all healthcare facilities will have enough money that they can buy pumps for everything. Or, you could look at the millions and millions of infusions being done every day that don’t use pumps and say, “what can we make today that will improve the care of those patients.”

That’s one of the really fun things about the space. If you look at infusion, either people are using pumps which cost thousands of dollars, or they are counting drops by hand. There’s this enormous, unaddressed middle space, and that’s where we come in.

It’s important to note that DripAssist isn’t just a more affordable solution. The whole ecosystem of the product is designed to make it thrive in emerging markets with minimal resources, training, and infrastructure. It doesn’t require a proprietary disposable. It doesn’t need any calibration or maintenance. It runs off one AA battery. It takes less than 5 minutes to learn, even for a lightly trained healthcare worker. It’s durable – you can drop it and use it in a hot, dusty, humid environment. That’s all part of our human centered design focus as well – to create a product that works not just for the user, but for the whole context of the user. Much to our surprise, we found that if you build things well then people everywhere want it. The success in developing markets has been exciting since that’s where we began but now with traction in developed countries, we’re seeing opportunities for healthcare savings which is an impact we did not initially expect we would get to, but here we are!

 

Medgadget: What healthcare savings can be achieved with DripAssist?

Dr. Kolko: Research shows that 4 out of 5 times, hand counting drops results in incorrectly infusing the intended dosage which is bad for patients and even worse for pediatric patients who have a higher sensitivity to infusions. DripAssist provides 99% accuracy. Hand counting also takes a lot of time. We did a small internal study that showed DripAssist could save two minutes per hour per infusion. And if you assume a nurse is managing fluid bags for six patients, which is an average load per nurse, those minutes add up. In fact, if you add up those two minute increments, the result is about 1,752 hours saved per year which, when taking into account the average nurse salary, is approximately $60,000 dollars saved per year. That’s real cost savings not to mention the fact that the time saved can be used on putting the focus where it should be, back onto the patient. The ability to impact the user in such a meaningful way is something we’re incredibly proud that one little DripAssist can achieve.

 

Medgadget: Can you tell us about one or two success stories where DripAssist is already in use?

Dr. Kolko: We’ve been working with 5 hospitals in Haiti for about six months now, focused on improving labor and delivery. These are hospitals and clinics where eclampsia, preeclampsia, and post-partum hemorrhage are being treated – these are some of the biggest sources of maternal mortality. In these facilities, the drugs are available, but there are no infusion technologies, so all medication is administered by counting drops. Since May, the nurses have been using DripAssist to safely administer magnesium sulfate and Pitocin (this work was partially supported by USAID). As soon as we introduced the devices, they started using them across the board – antibiotics, painkillers, etc. The nurse loved the fact that they didn’t have to count drops, and that they could reliably and accurately provide their patients the right dose of medication. We have these great quotes from the nurses talking about how much more comfortable they are when they are using DripAssist, how much easier it is to control an infusion, and how the alarm gives them more confidence. In one of the larger hospitals, the pediatric nurses eventually wanted to use DripAssist as well, so now they are sharing the devices across the wards.

My other favorite story is from a pediatric hospital here in the US. There’s a pediatric infusion coordinator we met at the Infusion Nurses Society conference. He saw DripAssist at the conference and purchased one. What got me was an email where he said, “I had a chance to use the Drip Assist today with a 16 month old oncology patient receiving platelets and I just loved it. Easy to setup and use. It was very accurate and easy to control the drip rate to what I wanted.” He goes on to say he’s “so happy I got to see you at the INS conference and learn about the DripAssist.” I’ve got a lot of cancer and cancer deaths among my friends and family, and knowing that our product is helping a pediatric cancer patient is simply incredible.

 

Medgadget: Finally, what else is Shift Labs working on today?

Dr. Kolko: We’ve done early work in a variety of areas, but I’ve got to tell you – the infusion space is fascinating, and it’s simply huge. We’re deep into some other technologies that address the unmet needs of medication management in emerging markets and here in the US, too. We think there is tremendous opportunity to simplify and improve infusion in hospitals, in homes, in outpatient infusion centers, and in countries around the world.

Link: Shift Labs…