“Let us not just create the best feeding management system, let us solve the baseline problem of helping NICU moms improve their pumping outcomes while away from their babies” – was the unanimous response from our development team in Spring 2016. Come Winter 2016, we embarked on an aggressive timeline to deliver not just a point solution for the multi-faceted human milk management problem statement, but to build an entire platform for the NICU and its several users viz. the nurses, the lactation consultants, the nurse management and last but most important, the mother.
Fast forward 5 months to April 2017, we had launched the first pilot at Penn Medicine and by August, 2017, the platform was announced for General Availability (GA).
It has been an absolutely magical 2017 – from introducing Keriton Kare to the world, to closing a $1M investment round, to winning SXSW Impact Pediatric pitch competition — a lot has happened during the year. Here’s what we learned, what we did right, what we could have done better and why we are excited about 2018.
WHAT WE LEARNED
Human milk management at a NICU is extremely complex. Period. The process is time-intensive and error-prone with little-to-no insights available for mom’s pumping patterns.
An average sized NICU does 10,000 feeds a month. An average feed life-cycle includes receiving breastmilk, combining milk bottles, fortifying milk and finally splitting a bottle to create the final feeds. Some steps in the process, like receiving breastmilk, need intensive data entries, while all steps in the process need 15-20 mental validations to ensure that the right milk is being prepared. This is why a 30-bed NICU spends ~13,000 hours/year on just managing breastmilk.
Would you be comfortable if you knew that the blood being used in your treatment was manually handled and eye-balled to make sure it matches your blood type?
Then why are the regulations so lax for breastmilk oversights, which is another bodily fluid and a carrier of HIV, Hepatitis, etc. Although the risk of transmission via breastmilk is low, 2-RN checks during the process of milk preparation not only takes away critical care time from the nurses, it is highly prone to errors like expired milk feedings, incorrect bottle feeding, incorrect bottles being combined, etc.
Add to the above list an increasingly difficult task for a NICU mom – pumping continuously while physically separated from her baby. Pumping itself is unnatural, the separation just makes it a lot harder. Lactation consultants from the NICU want to track a mom’s pumping supply, but due to lack of HIPAA-compliant solutions, they have to resort to manually tracking the mom’s supply — either via in-person conversations or by sifting through the NICU inventory in the refrigerator/freezer — neither of which allows for timely intervention with the mom.
WHAT WE DID RIGHT
There are several challenges, unlike any other, when executing on such a large platform for digital health.
First, the general product development perception in healthcare is that of a disconnected team developing a new product in-silo, solving a problem that the clinicians do not actually face. We took a different approach from Day 1 — we candidly accepted that NICU and breast milk management is not our domain expertise, our expertise lies in solving process orchestration problems with technology.
We took our technology expertise, merged it with HUP’s clinical domain acumen, which resulted in the the birth of the Keriton Kare platform as we know it today.
We visited the NICU at 6 in the morning, 12 in the afternoon, 6 in the evening and then again at 12 in the night to observe the process over and over again. We sat down with our users over design thinking sessions to learn and better evaluate the challenges and probable solutions. We also surveyed ~350 moms to understand the shortcomings on their end before nailing down the product wishlist and a product strategy to deliver on the wishlist.
Even post product deployment, we continued a very active conversation with our end-users and upgraded the platform with critical features that we would not have realized had we not embedded ourselves in the NICU. We really lived in the NICU, got it right and delighted our nurses and new parents along the way. One such significant functionality additional was that of a Donor Human Milk module.
Second, the typical fail fast silicon valley thought process does not bode well within the paradigms of healthcare innovation.
With patient’s lives at stakes, there is just no failing fast.
However, as an early-stage startup, you still need to stay lean and ship fast to ensure business growth. To this effect, we staffed a 1:1 software engineer to quality engineer ratio in the team. This was the single biggest reason we were able to ship an update almost every other week (to either Kare Mom or Kare Nurse app), while maintaining utmost stability on the platform.
Lastly, if you observe most digital health solutions, *cough* EMRs *cough*, most of them feel as if they were designed in the late 1900s. Robotic, banal and poorly designed. Enter data here, click there, enter data some more, click some more, click, click, click *ugh*. We set out to change that from day 1.
At the core of our design principles was empathy for our end-users – nurses, lactation consultants and mothers.
More so, we had to design our apps keeping in mind a variety of end-user personas, over a wide age-range – meaning extremely intuitive UI/UX and no fancy interactions. Our hearts were in the right place and our minds were on the right problem. The outcome? A system order of magnitude better than existing solutions, that not just dramatically reduces errors, expirations and effort, but also helps NICU moms improve pumping outcomes, ultimately ensuring babies get more milk. We are making a real, important difference, and this human impact, capable of driving economic impact, has always been the foundation of our company.
WHAT WE COULD HAVE DONE BETTER?
While we were fiercely executing on our product strategy, we grossly underestimated the task of training our end-users on the platform, considering how NICU staffing works with day-shift nurses, night-shift nurses, per-diem nurses, travel nurses, etc. Bedside training is not always the most efficient and pulling the nurses away from critical care is not always possible.
We learned our lessons from the first pilot and setup a formal training structure in conjunction with the nurse manager, from our second pilot onwards. Furthermore, we have also augmented our in-person training sessions with email updates, online video learning modules and self-learning packets to ensure continuous education and knowledge sharing.
WHAT TO EXPECT IN 2018?
From winning the Impact Pediatrics Pitch Competition at SXSW in March, to being heralded as the “Swiss Army Knife of Breastmilk Management”, to being celebrated as the Startup of the Year by Philly Geek Awards — 2017 has been a fantastic journey for the Keriton camp. Other than these external validations, it was the confidence of our end users and continued excitement whenever we sat down with them to discuss the product roadmap for the Keriton Kare platform that kept us going through the highs and lows of the year.
Continuing on our mission to improve NICU processes and outcomes, we are particularly excited for 2018, where we are focussed on expanding adoption of our platform, while providing a premium customer service experience.
With our platform now available for Android, iOS & WoWs (web browsers) and an active customer pipeline, 2018 already looks like a busy year.
As we wrap up the year, it would be remiss of me to not thank our unbelievably dedicated team, the confidence and support of our investors, the advice from our mentors and an army of individuals from Penn Medicine. Thank you all for your staunch belief in the mission and vision of our company.
Upwards and onwards, wishing you and your families a happy holiday season and prosperous 2018 — from everyone here at Keriton!