How no-code tools are democratizing EMR customization

Why custom no-code EMR workflows are ideal for specialty providers & small practices

No-code EMR tools are enabling specialty healthcare organizations, like primary care, psychiatry and weight-loss practices, to build custom workflows for their unique needs.

One such no-code tool is Canvas Studio. Launched on May 21 in beta, Canvas Studio is powered by Anthropic's Claude Code. Using natural language, clinicians can customize no-code workflows in Canvas Studio that serve as plug-ins for Canvas' EMR, a certified EHR technology (CERT) platform. CERT platforms conform to CMS and ONC standards on storing data in a structured format, allowing providers to easily retrieve and transfer patient data.

For instance, a doctor treating weight-loss patients could create a custom EMR workflow to make a nutrition plan or incorporate data from wearable devices, such as an Apple Watch or Oura ring, said Adam Farren, CEO of Canvas Medical, an EMR startup. Clinicians can also create data-collection workflows to understand how treatments impact diet.

Custom workflows allow clinicians to tailor their care to patients' needs and provide access that's more comprehensive than simply "schedule an appointment and go to a clinic," Farren said.

Not only that but primary care practices can use custom EMR workflows to meet payer documentation requirements for annual wellness visits, according to Farren. The custom workflows allow providers to easily repeat the documentation process so they can submit claims to insurance.

The tool allows practices to customize patient-facing workflows as well. In Canvas Studio, providers can create plugins that identify which patients may be due for an appointment and then automate a reminder to schedule a visit.

"This is a task that often happens manually through a spreadsheet list and phone calls but in Canvas, [it] could be fully automated with software," Farren said.

Not only does Studio allow providers to create their own workflows, which can be proprietary or closed workflows for that specific running version of Canvas software, but Canvas also makes workflows available in an open-source format. Providers can use these open-source workflows from Studio as a starting point to create their own. In general, Canvas' AI agent builds 80% of a workflow from an existing plug-in library, and then the provider completes the last 20% on their own, Farren said.

"Our customers own the intellectual property to the plugins or workflows they build in Canvas, and it's up to them whether they want to open source them for other customers to use," Farren said.

Custom workflows allow small practices to move from idea to workflow faster rather than waiting for vendor roadmaps or expensive customization projects, according to Shegewi.

In addition, custom EMR workflows allow smaller healthcare providers to use staff more efficiently by automating tasks such as routing questionnaires and triggering follow-ups.

"You can quickly build an intake form or care pathway and tailor it to niche services, so I think there is incredible amount of power for smaller facilities leveraging no-code -- and I'll even say low-code -- workflows," said Shannon Germain Farraher, senior analyst for healthcare at Forrester Research and a former nurse.

Like Shegewi, Germain Farraher sees speed to deployment as a major advantage for smaller practices when using custom no-code solutions.

"Provider organizations, especially the smaller ones, can't afford any real disruption to workflows or impact on the negative side of their revenue," Germain Farraher said. "If something is going to impact patient throughput, it better be a really quick disruption."

Custom no-code EMR workflows could also improve the user experience for clinicians more than EMRs have done in the past, as they allow clinicians to create workflows that work best for them, Germain Farraher suggested.

"Other [EMRs] were designed for clinical workflows, but they weren't designed by clinicians," Germain Farraher said. "They're just designed mostly to capture data."

No-code EMR workflows incorporate natural language, iterative processes

To create custom EMR workflows, providers communicate with the Canvas interface using natural language in an AI-powered chat interface just as they would with a tool like ChatGPT. For instance, for a weight-loss patient on a GLP-1, a provider could enter the following into the chat interface: "Build a workflow for renewal of a patient's year-one prescription when they've already established care with my organization," Farren said.

AI agents then develop the workflow and add it to the testing environment in the provider's Canvas EMR.

Farren further explained that using the no-code Canvas Studio platform is an iterative process.

"What you're doing is essentially prompting the AI and then refining its output with additional prompts and instructions, and so that's the iteration loop," Farren said. "[Providers] iterate inside the Studio interface, and within an hour or two, they have a result."

Further, clinicians are able to build workflows within the organization's existing protocols rather than building them outside the system and then integrating them.

"They're actually doing the work inside the EMR, which already is encoded with all the organization's rules and protocols," Farren said.

Although providers create workflows in Canvas' EMR, the company supports interoperability with other EMR platforms through health information exchanges. The FHIR data standard makes this connectivity possible, according to Farren.

But governance is important when managing custom EHR workflows

With proper oversight, providers can avoid creating too many custom EMR workflows, which could introduce inconsistencies across healthcare practice, Germain Farraher explained.

"Inconsistent workflows across even a small facility can slow things down," she said.

So, governance protocols like tracking and maintenance are essential when using no-code tools, she added.

IDC's Shegewi agreed, noting that governance is necessary to avoid scenarios where healthcare practices create an abundance of poor workflows.

"No-code can make a workflow more customized, but that doesn't make it a sound workflow," Shegewi said. "If the underlying process is poorly designed, customization and automation will scale dysfunction. So, governance is, in my view, a major challenge."

He also noted that healthcare workflows affect many areas, including clinical decisions, patient safety, coding and billing, compliance and policy.

"If every team starts building in their own way and customizing without proper oversight or guidance, you could end up with a sprawl of inconsistent workflows, duplicate [programming] logic, and technical debt," Shegewi said.

Farren discussed how building governance into no-code solutions creates guardrails, which is what Canvas has done with its Studio tool.

"The [Canvas] agent understands where it can customize and deviate for a user, and where it has to follow standard protocols," he explained. "We've built the agent to be aware of those rules, and the organization, not the user, sets those rules. It's part of the agent's code."

In addition to building guardrails, clinicians can avoid potentially inconsistent custom workflows with "governed innovation," according to Shegewi.

"Let's empower knowledgeable and credible clinicians and operators to translate needs into customized solutions, wrapped in the right permissions, processes, testing, audits, controls and ownership," he said.

With proper governance and the flexibility to build workflows suitable for specialty practices, clinicians could speed up EMR workflow development, gaining the efficiency of a larger health system with a full IT department.

Brian T. Horowitz started covering health IT news in 2010 and the tech beat overall in 1996.