Clinical pathways are the currency of health tech

As healthcare becomes more entrenched in the digital revolution, the need for an approved set of protocols for care delivery — clinical pathways — is becoming increasingly critical.

Clinical pathways, as defined by the Children’s Hospital of Philadelphia, are the “standardization of care that translates guidelines and/or evidence into localized infrastructure and processes.” These processes have significant financial implications, as they can decrease payer (insurance) denials, allow providers to enroll in performance-based reimbursement, or help resource-constrained provider systems better allocate financial resources.

These financial benefits, coupled with current macroeconomic forces — the struggle for profitability in hospital systems, the rise of team-based care (via non-physicians), the challenges of utilization management at scale for insurance companies, and strict legislation around patient communication and healthcare systems interoperability requirements — have paved the way for pathways to become the de facto operating system for healthcare.

Pathways provide a currency for patients, providers, payers, and technology companies to prove a return on investment (ROI), both clinically and financially. Ultimately, this has created a unique opportunity for emerging and legacy healthcare companies to build around pathways, leveraging new datasets, delivering novel reimbursement models, preparing for and complying with new transparency and interoperability legislation, and utilizing advanced AI to provide personalized understanding and delivery of information.

Data is the driver of pathway’s success in clinical and operational settings

In the past, national guidelines dictated decisions made all the way down to the local level. Now, local, and even personalized, evidence-based pathways are driving decisions, thanks to the ability to access, create, and analyze new datasets.

Access to data, standardization of data rights, and the utilization of HIPAA-compliant collaboration tools, such as Datavant, will continue to improve compliance and create a more democratized, fine-tuned system of analysis for personalized pathways. The Centers for Medicare & Medicaid Services (CMS), under advisement from several healthcare companies, is now implementing an approved format for hospital charges as single machine readable files (MRFs), which will be leveraged to standardize all charge information.

With the ability to access, create and analyze new datasets, personalized, evidence-based pathways are driving more healthcare decisions.

This will allow both large (national) and small (local) providers to access previously unavailable data that can then be used to enhance care coordination and delivery, promote quality improvement, advance research, and increase ROI.

We spoke with Eric Leroux and Dan Imler, emergency department MDs and the co-founders of clinical pathways startup Curbside, about the ability to utilize new data models in the transition from the national to the local level for clinical logic creation. They pointed out that while nationalized datasets have a role to play in insight generation, “the clinical and financial responsibility of point-of-care decision making is still inherently local . . . decisions must be governed there to have any real impact,” especially in value-based care constructs.

As more companies like Curbside and AvoMD work to bridge the gap between art (no guidelines) and science (NCCN guidelines) when creating pathways, we expect more investment in startups that focus on the intersection of digital health and fintech as evidenced-based pathways and localized reimbursement engines become more necessary.

Reimbursement models

Pathways have evolved as a “cornerstone of future reimbursement methodologies and quality efforts,” as described by Dr. Robin Zon in ASCO Connection. They can help providers avoid “time-consuming prior authorization and appeals with payers,” and capture “stage and molecular data for a more refined risk adjustment.” Today, clinical pathways can be used for reimbursement via a number of models, from value-based care (VBC) to legacy fee-for-service (FFS).

For example, CMS uses clinical pathways to create a benchmark for cost and quality in the Medicare Shared Savings Program (MSSP). Providers who can provide care at a lower cost than the benchmark and who can meet certain quality standards are eligible for shared savings.

Given the connection between pathways and reimbursement, the infrastructure to assist with creating, enhancing, and delivering pathways is becoming a focal point for stakeholders, and thus a ripe opportunity for startups to build and sell into established healthcare enterprises.

A recent Flatiron Health blog post noted that “technology is going to need to be more embedded in the workflows of clinicians, and it’s going to need to provide even more insight into optimizing treatment pathways to improve outcomes while achieving operational goals for health systems.” Pathways are the unique resource that bridges clinical evidence, which is constantly being updated, with point-of-care decision making via an agreed-upon reimbursement model that meets the needs of all stakeholders.

We are already witnessing how tech-forward companies, such as Memora Health and Story Health, are able to push and pull from clinical pathways and leverage advanced technology, such as natural language processing (NLP), to generate negotiated pathways, provide evidence-based enhancements, and offer care delivery coordination. As such, pathway technology facilitates novel care and reimbursements models, for which large health organizations are taking a buy versus build approach, an opportunity for emerging startups and investors alike, in order to meet the requirements set by new government legislation around data transparency and integrated new solutions.


Legislation is continuing to drive the relationship between pathways and payers, and this will only become more codified and transparent. In February 2023, CMS issued a proposal designed to address the administrative burden of prior authorization. In a conversation we had with Brendan Keeler, head of product at Flexpa, he noted that this proposed rule “could more broadly be termed ‘Provider/Payer Collaboration.’”

With this new legislation, Keeler reiterated that sharing of claims data from payers to providers will work to “highlight where providers have gaps in their insights. . . . [T]he documentation rules that payers use to evaluate authorization will allow us to move the needle from phone calls and tribal knowledge to encoding of payer logic into provider pathways.”

While this process is not a requirement for employer plans, nor is it mandated for public exchange plans, which includes qualified health plan (QHP) issuers on the Federally Facilitated Exchanges (FFEs), it is likely a first step toward further integration and transparency between patients, payers, and providers. For the first time, technology companies, like Turquoise Health and Cascade Health, are increasingly engaging with patients (who want transparent pricing), payers (who need to provide real-time prior authorization and utilization management), and providers (who are now required to be part of the process).

Ultimately, data could become structured and available to create and assess patient pathways, creating an opportunity for startups to provide real-time information, learnings, procedures and communications, all of which will require stakeholders to utilize AI-based tools.

AI and NLP

Enhancements in AI and natural language processing (NLP) will play a critical role in elevating pathways’ impact because of the technology’s ability to distill and communicate large sets of information with substantially less manual effort. Memora Health is built on “NLP systems that can interpret meaning out of unstructured data, enabling automation across more use cases and reducing tasks that would otherwise fall on human staff . . . [while also] allowing patients much more flexibility in sharing information that could be relevant.”

We are already observing advancements in AI and NLP, including generative AI (GenAI), for the models that underpin a number of areas, including, but not limited to, coding and billing coordination (e.g., Nym), differential diagnosis (e.g., Glass Health), and documentation (e.g., DeepScribe). Many of the companies focused on these domains deploy both legacy AI systems, as well as new AI models, which provide superior real-time analysis and communication due to a deeper understanding of language and data than was previously possible.

We see a future where AI models facilitate real-time pathways based on real-world evidence and real-time payer negotiation. While we believe generative AI will have a profound impact on provider, administrator, and patient communication and education, we view AI more holistically as a powerful engagement engine across the entire healthcare ecosystem.

Value-based pathways

The recent advancements in data, reimbursement models, legislation, and AI are transitioning pathways from setting quality standards at a population level to becoming more personal, localized platforms driving financial ROI to VBC models.

With a heightened industry focus on risk-sharing and VBC models, value-based pathways will help drive clinical metrics, reduce unnecessary care, standardize decision making, and capture revenue, bonuses, and pay-for-performance (P4P) incentives. For example, as the industry moves toward a P4P standard, this will work to further blur the lines in the continuous favor of improved quality and efficiency.

In addition, value-based pathways could create an underlying fintech ecosystem where financial tools intersect, take on risk, and build new pathways, similar to what Fika portfolio company Accorded is doing for actuarial intelligence across healthcare. An opportunity exists for more fintech-enabled healthcare startups to build within this ecosystem and underwrite, securitize, and service patients at certain parts of the patient journey, allowing for the expansion to better care pathways.