Why screen scraping isn’t the enemy—but the symptom of a bigger healthcare integration problem

To kick off our new Redox "Connections" thought leadership series, Redox Field CTO Tim Kessler reflects on why healthcare integration is still so hard—and what we can collectively do about it.

The integration workaround no one wants—but many choose to use

It’s May of 2025. The healthcare industry is now 16 years and billions of dollars of investment from the signing of the HITECH Act that drove widespread EHR adoption. As an industry, we’ve invested a lot of time and energy into “solving” interoperability. Despite federal rules aiming to eliminate information blocking, a lack of consensus in interpretation has led to confusion and sub-optimal compliance. While we’re continually improving, the promise of seamless interoperability is still largely unfulfilled.  

For background, when the HITECH Act and its successors were passed, the vision was clear: digitize healthcare and then connect it. Since Meaningful Use, digitizing healthcare at scale has been largely successful as EHR adoption in the US is near-universal, with over 95% of hospitals and 88% of office-based physicians leveraging EHRs. But the second part, accessing and connecting healthcare data, continues to lag. 

To their credit, EHRs have made meaningful strides toward interoperability, often navigating complex, evolving regulatory landscapes that can hinder consistent progress—in addition to dealing with a flood of eager third-party integration requests, customer security concerns, and long-standing infrastructure and technical debt. While gaps remain, these challenges reflect their reality of having to balance innovation, compliance, and provider demands in an increasingly fragmented ecosystem.  

Despite their continued effort, most technical options are still limited to a patchwork of opaque, inconsistent, and costly standards, especially when organizations try to manage the complexity on their own. The result is a system where digital health companies doing the “right thing” is time-consuming, costly, and riddled with uncertainty. 

To bridge the gap, the digital health industry has adopted methods used in many other industries when healthcare data isn’t easily accessible. One such method is Robotic Process Automation (commonly known as RPA or “screen scraping”) that navigates complex integration processes. As a result of the burdensome technical and operational environment, provider organizations and digital health companies alike have looked to RPA to close gaps that hold them back. RPA emerged not as a visionary solution, but as a necessary workaround. It provides an immediate way to extract and move data by mimicking human behavior—logging into web portals, downloading documents, and re-entering information into downstream systems. 

In recent weeks, we’ve seen pushback to RPA as a viable integration tool in healthcare with arguments centering around it being unstable and insecure. If you’re reading this in hopes of finding a passionate defense of RPA, then you should probably stop reading now. 

At best, RPA offers a quick and initially cheap way to integrate healthcare data, despite its limited capabilities. For example, for Providers, RPA can streamline administrative burdens by automating rule-based data entry tasks and reducing human error. At its worst, RPA is a brittle stopgap that is unreliable, difficult to maintain, and hard to audit. For example, for EHRs, it can compromise the experience they offer to clinicians as their user interface (UI) breaks and tech support tickets pile up blaming them, while also introducing security and compliance concerns regarding what data is accessed and how that data is used. 

Digital health companies often face tight timelines, limited budgets, and the need to access data from systems that haven’t fully operationalized the interoperability needs that innovation is driving. For many, RPA is the most practical option; it offers something no “approved” path can always guarantee: a quick integration pathway that offers tactical automation when one is under pressure to move fast.

No one wakes up hoping to build their integration stack around bots and screen scraping—but when the front door is locked, you are forced to find alternative options, especially when the future of your company depends on access to this data.

RPA isn’t the problem. It’s the signal.

Yes, RPA is fragile. It’s susceptible to breaking when a UI changes. It lacks context-awareness. It doesn’t understand the meaning of the data it’s moving. It can be difficult to monitor, hard to secure, and challenging to maintain over time. While it’s nowhere near an ideal solution, RPA can get the job done, helping lean teams work fast within tight budgets.

And so, digital health companies will keep choosing the workaround—not because they want to, but because they have to. It’s a signal that the system still isn’t working the way it should. If the only way to move data is to mimic a human clicking through a portal, we haven’t really solved interoperability—we’ve just replaced paper shuffling with digital duct tape.

So no, the answer isn’t banning RPA; doing so would be like shutting down the only road through town before the highway is built. If we want to move beyond screen scraping—and we should—then we need to make sure the “right way” of integrating is actually achievable. My recommendation for organizations that want to sophisticate and scale is to focus their efforts on building strong integration ecosystems.

That means:

• Open, well-documented APIs that are transparent and accessible.

• Consistent adherence to standards like FHIR, implemented in a usable, interoperable way.

• Clear onboarding paths that don’t require enterprise-scale deals or six-month timelines.

• Vendor cooperation when it comes to third-party integrations.

This isn’t just a plea to any player in the ecosystem—it’s a call to everyone in the industry. We all have a role to play in making data flow more easily, more safely, and more sustainably. Because when the better road is finally built, no one will want to keep using the detour.

Where Redox stands—and where we’re headed

If you are grappling with integration gaps, you might wonder: should we turn to RPA to close them? Our view is simple—relying on RPA might offer a short-term fix, but it’s not a strategy for scale. And while we don’t offer RPA ourselves, our platform can support those who choose RPA-based integrations with Redox as a routing mechanism to your solution of choice. 

At the end of the day we are here to meet customers where they are and strive to create a sustainable, secure, and standards-aligned interoperability platform—so you don’t have to. Your team can focus on delivering innovation, while our team stays ahead of the evolving integration landscape, navigating variability across EHRs, standards, and vendor policies. We’re here to help build something better. 

As to what will happen with RPA? Let's see! But meanwhile, let’s stop treating workarounds like solutions, and start demanding—and delivering—the connectivity this industry deserves...and more importantly, desperately needs.