By MARIE COPOULOS

I’ve had the great fortune of spending much of my career at the intersection of health care innovation and the underlying data that drives new models.

For those of us who’ve worked in this space for a long time, there’s a certain pattern recognition that comes with this work that is often immediate and obvious – both in terms of really cool developments but also gotchas. “Ah, you’re stumbling here. Everyone does that.”

The challenge, I’ve found, is that these ‘gotchas’ that can be so visible to the folx who’ve worked in health tech for the past few decades can be counterintuitive in the business and even met with resistance. Why?

I’m going to focus here on pattern recognition, with the goal of highlighting common stumbling blocks and, critically, ways you can interrupt them if you see them.

Pattern #1: Lacking a Clear-Eyed View of Market Data Gaps
Key Question: Do you understand how the market you’re in informs your ability to measure your work and use data to drive insight?

For those of you building models that change the status quo – this is for you. By nature these innovations break from existing care and financial models with the goal to improve them. We need this in health care. However, it’s common to overlook the fact that breaking with the status quo also breaks with the ways that we capture and serve up health data.

To this end, don’t assume you will be able to measure and show success, and that the data you need must be out there. The true differentiator is for both to align. Design with intention.

If you’re at the stage of thinking about a productized solution to a health care problem, then it is also the right time to look at the market with a lens toward data availability. In your problem space, what’s the data set you’re likely to lean on? Is it sufficient?

If the answer is that the data is not available or notoriously problematic in your market space for the problem you’re solving, this merits a pause. Can you find a way to survive in this reality? Can you create the data set you need? Can you adjust what you’re doing in some way to align with what is available? Is qualitative feedback ok?

Pattern #2: Accumulating Non-Technical Roadblocks Key Question: Do you have a good handle on the non-technical challenges impacting your data business?

A decade ago I would have approached this question differently. Technical challenges were often paramount as we tried to figure out how to solve the basics. Today, however, it’s often the opposite, in that business challenges are more likely to slow down technical progress than the other way around.

What do I mean by that? Most frequently I see organizations stumble on things like data acquisition, partnerships, and the right strategic vendor choices and these stumbles manifest in increasing technical debt that grind teams and reduce productivity.

In new models and approaches, in particular, there are often many players involved, eager to try something new. Because you’re doing something new, by design you won’t know all the stumbling blocks. What matters is not that you know what they are, but that you have good governance that allows you to work through these issues together.

It’s not that the problems are insurmountable, but the question of who is going to spend limited resources, in what order, on these very hard problems. Who owns that work and that risk? Who makes decisions? Think about this early.

Pattern #3: Lack of Focus
Key Question: Do you know what pieces of information provide disproportionate value?

There are many kinds of healthcare data (claims, EMR, ADT, pharmacy, labs, etc). Those different sources shine light on the same patient events–a single visit results in little bits of a story that are captured in many electronic systems.

Often talking about different data types feels wildly obscure. But, if there’s one concept to center on as a business leader, it’s getting to the bottom of this question: what are the pieces of information that are disproportionately valuable to run your business?

Some of the most value-add businesses in health care, in my view, have figured out how to narrow in on a piece of information (readmissions, medication fills) that is scalable and hyper focus on consistently improving on the patient and clinician experience, and outcome.

This reflects the reality of our industry today. Because health care data is messy and inconsistent, it takes a lot of work to get into usable forms. Absent that work, this information can be confusing, contradictory, and too frequently – noise.

Until we hit the point where this is not quite so hard, make sure you know what kind of business you are and where you want to invest your resources. From an infrastructure and product perspective, a business powered by a narrow insight looks different from a business powered by a holistic, normalized view of a patient. Which are you?

Pattern #4: Short-Term Wins that Don’t Build

Key Question: Do you feel comfortable with the tension between short-term wins and long-term wins and do you have an open conversation with your team on this topic?

This manifests in a couple ways. One is short-term wins that don’t build, and the other is a focus on long-term goals exclusively with unrealistic timelines. These problems are certainly not counterintuitive, but they are hard to interrupt and one of the reasons we see so much churn, burnout, and disappointment in major launches. My advice: aggressively look for ways to build in small, additive steps. 

An example: It’s really common in a new model to build a partnership to access information to provide a broader view of a patient population. Depending on the problem, you might find local, regional or national entities to support you in finding the right information.

A short-term win might have you build a partnership with a provider of that information with the goal of a quick win. However, these are the kinds of decisions that often weigh on technical teams in the long-term as they manage many partnerships and many interfaces, and in fact the cumulative effect can be devastating to productivity and innovation. It’s not just the weight of managing one-off work, but the sense of loss of having to rebuild again and again.

Building in an additive way takes a little extra time at the start, but reduces waste over time.  Consider how any small project will serve future efforts (i.e. is this a partnership that scales, contractually and technically?). These small wins build momentum and collective capacity.

Pattern #5: Siloed Technical Teams
Key Question: Do you have a good sense of what motivates your team to solve hard problems for you? 

Choosing to work in health care data means choosing to work in one of the most challenging technical segments – because of the weight of regulations, messy data, and old infrastructure. In my experience, a common motivator is mission. In the teams I’ve built, there is a palpable drive to help patients and improve systems for the better.

If I leave you with one point, it would be not to overlook this connection to mission and sense of belonging to the team that is helping improve patient lives. Yes, fair compensation and good benefits and work-life balance all matter. But, don’t forget ‘why’ these talented technical team members are sitting at your table, frequently doing work that is technically below their capabilities.

Ask them what makes them feel informed and connected to the whole. It will make it collectively easier to solve the messy, hard problems together.

Marie Copoulos, MS, is a public health professional and long-time health executive working at the intersection of analytics, population health, and climate. (She previously published on THCB under the name Marie Dunn).