Data — The Core Bottleneck to Digital Health (Part 4)
This is Part 4 of a 5 part series. Here’s Part 1, Part 2, Part 3, and Part 5.
This post was originally published on Medium.
This is Part 4 of a series. Here’s the Overview, Part 1, Part 2, Part 3, and Part 5.
How does healthcare’s data problems impact digital health?
As patient experience and outcome becomes increasingly crucial to healthcare, more and more healthtech startups are entering the foray. There’s two schools of thought for the survival of these startups. While Kevin O’Leary believes “if you’re trying to disrupt the healthcare system, you are going to need to partner with incumbents effectively in order to stand any chance”, Neal Khosla’s take is that companies need to become “existential threats” that “[scare] the shit out of the healthcare establishment.”
Essentially, to be truly disruptive, these startups have to demonstrate that they can create a viable business through solving a core problem in patient experience, through two options: partner with providers and/or payers and align their incentives & behaviors for the patient’s interest, or replace these incumbents to directly interact with the patient themselves.
In pursuing one of these two options, healthtech startups will inevitably encounter the complexities of navigating healthcare data. Here, too, exists two options: integrate with existing data sources, or build your own data infrastructure.
For healthtech companies taking the partnership route, integrating with legacy data infrastructure will be a bitter pill to swallow. When your clients are change-averse health systems or payers with paper thin-margins, the ability to provide immediate ROI is the line between life and death.
However, integrations are expensive and time-consuming. Conservatively, it will take an industry-leading integration team (involving full-time account managers, engineers, etc.) 1–2 months per new enterprise healthcare client, and that number can quickly expand based on variables such as type of EHR, implementation of change management processes, etc. (for larger, slower organizations, this could easily extend to +6 months). Take this process and multiply it by the amount of sales cycles a start-up could go through, and it quickly becomes a multi-million dollar scaling problem for the start-up over time.
Talk to employees within the start-up, and you’ll see this materialize into multiple, tangible pain points. Partnerships and account management will seek faster turnaround time and reduced costs on client go-lives, but implementations and integrations are slowed with every new data quirk. Product wants lowest scope efforts for highest impact, yet any build involving healthcare data requires high scope for an incremental amount of impact. Engineering dreads wrangling clunky legacy data and interacting with ancient code bases lacking detailed, friendly documentation. The organization as a whole feels a drain as these data problems start to shift focus away from delivering value to end users.
Alternatively, if a company seeks to circumvent integrating with cumbersome systems, it will have to own the experience itself and build its own data infrastructure. This is the approach for companies like Oscar, Flatiron, Ro, which have redesigned claims systems, EHRs, patient relationship management platforms, etc. to be able to ingest and utilize data for their specific use cases and deliver a modern healthcare experience.
Obviously, building modern versions of these systems is a laborious undertaking that requires significant time and capital — it’s the reason that only the most well-capitalized companies have found any degree of success. Yet, beyond this, maintaining the usability of the data itself persists as a massive business problem. Oscar, for instance, makes 2,200 data edits per day to its own provider directory to maintain accuracy and effectiveness.
Finally, as care continues to unbundle from the hospital, these new healthtech companies are developing new ways to access, pay for, and experience your care, creating new sources for healthcare data. However, it’s critical that these solutions create a synergistic ecosystem that elevate the patient experience, not exacerbate fragmentation. All participants need to be able to access and contribute to a patient’s longitudinal data history with minimal friction. This in turn will dramatically improve their value proposition, addressable market, and utility for the patient.
In healthcare’s current data landscape, this is simply not accomplishable. Aside from hindering healthtech companies of today, data interoperability and quality issues slow healthtech from bringing in the future of care.
Ultimately, all healthtech startups, regardless of their approach, are individually reinventing the wheel every time they interact with healthcare data, unable to maximize their utility for the patient. This is healthcare’s bottleneck.
In sum:
To be successful, digital health companies have to solve for an improved patient experience through two paths: partner with incumbent players and integrate existing data sources, or replace incumbents and build their own infrastructure.
Integrating existing sources is expensive, time-consuming, and difficult. Variables in the process mean that this has to be repeated for each new sales cycle, becoming a critical scaling issue decreasing business health and increasing employee backlogs.
Both of these paths cause each healthtech start-up to reinvent the wheel, creating a bottleneck.
Building new data infrastructure is impossible in and of itself, but managing that data’s usability significantly adds complexity.
As healthcare data increases in quantity and variety, the opportunity to create a healthcare ecosystem that works for every stakeholder is larger than ever. Healthcare’s data bottleneck actively impedes this from happening, driving fragmentation instead of synergy.
What’s happening right now that points to the alleviation of this bottleneck? Click here for Part 5.