Our guest on this episode is Siva Namasivayam. Siva is the co-founder and CEO of Cohere Health, an AI-powered platform that simplifies the time-consuming prior authorization process for health plans & providers. Cohere has raised $66M to date, and is backed by Polaris Partners, Longitude Capital, Deerfield Management, Flare Capital Partners, and Define Ventures.

Prior to Cohere, Siva was a founder and CEO of SCIO Health Analytics — a healthcare predictive analytics company for health plans, providers, life sciences, and pharmacy benefit managers. The company was acquired by EXL for $250M in 2018. Siva also served as the Sr. Vice President and Chief Revenue Officer at Mphasis, an HP company, the Senior Vice President of Payor Services at Dell, and a Senior Software Engineer at Intel.

In this episode, Siva and I discussed:

  • His path to entrepreneurship
  • Responsible use of AI in healthcare utilization management
  • How Cohere is transforming prior authorization from an administrative burden into a quality improvement process
Transitioning from building software to building companies

As a child, Siva dreamed of being an engineer. He saw his dreams through, starting his career as a Senior Software Engineer at Intel, working with programming and software compilers for microprocessors. While deep in the tech, Siva found himself drawn to the business side, wondering how he could leverage his software knowledge to find new applications for real world problems.

Siva pursued his MBA at the University of Michigan Stephen M. Ross School of Business to transform his mindset from engineering to business and new product development, and he set his sights on entrepreneurship. The idea for his first company came while Siva was the Chief Revenue Officer at MphasiS, an EDS Company.

“When I was working at EDS, I found that large payers had invested a lot of money in developing big data warehouses and then they didn’t have many use cases. That’s where I said, ‘This data is sitting around, let’s use that and develop use cases around payment and population health.’”

The resulting company, SCIO Health Analytics, provided predictive analytics and data insights to reduce unnecessary care and improve care quality for healthcare organizations. The company was a great success, selling to EXL for $250 million after 10 years of consistent growth.

Reimagining a more proactive healthcare utilization management process

SCIO provided retrospective payment integrity services, such as medical and pharmacy claims auditing and clinical reviews to help payers, health plans, and PBMs control costs, waste, and fraud. While at SCIO, Siva observed many instances when the health plan had paid the provider’s claim, but the appropriate care had not been rendered.

This inefficiency led Siva to reimagine the entire utilization management (UM) process, aligning all stakeholders before care is provided, thereby preventing disagreements and delays in payments.

“Why can’t we do that to this whole utilization management with pre-care. If things were done appropriately, everything else will be straightened out. That’s where I got the idea …Why waste the entire process and the patient was not getting the right care anyway? Why don’t we change it?”

This shift in perspective laid the foundation for Cohere Health’s mission — optimizing healthcare UM and prior authorization (PA) to ensure that the right care is delivered efficiently and promptly.

Streamlining and transforming the PA process into a quality improvement process

Siva and the Cohere team are committed to transforming UM into a patient-centric, outcome-driven approach. The Cohere platform products fall into two categories:

1 — Process Automation Solutions: Automate the intake, review, and decisioning of the PA requests that providers submit to health plans prior to care being rendered, ensuring quick approvals for patients.

Providers submit the PA request to Cohere, who receives those requests on behalf of the payers. Cohere then automatically digitizes the request, reviews it against the payer’s policy, and returns a decision on whether the request is appropriate based on the payer’s policy.

“Today, most of the prior authorization requests are submitted and reviewed manually, it takes time to make the decisions. Say there are 100 authorizations to be worked on. At the end of the day, 85–90% of them need to be approved. The first piece is how can we approve it immediately… It’s like going to the store, you buy something, you pay, and then it’s done.”

2 — Clinical Innovation Solutions: Layer AI / ML algorithms and evidence-based clinical care pathways into the process automation products to influence the patient’s care journey.

When a request is submitted, Cohere serves as a second set of eyes for physicians and health plans. Cohere identifies where the patient is in their care journey and any potential risks based on their diagnosis and medical history. If the service in the PA requests is the next step in the evidence-based care pathway and there are no potential risks, the service is immediately approved.

If the requested service is not the evidence-based recommended next step or there is a patient-specific risk, Cohere nudges the provider via their platform to initiate a discussion.

“Our focus is to make sure that if something is correct, immediately approve it. For the things that are questionable, try to work with the physician to see if you can come to an alternate path…We are able to do most of it through the platform.”

The Cohere platform is able to immediately approve roughly 90% of authorization requests, reducing the administration burden for providers and lowering the time patients have to wait for care.

Reorienting the PA approvals around the recommended care path for that patient

To ensure that the algorithms are accurate and in line with the latest clinical guidelines, the care pathways built into Cohere’s platforms are based on 3 sources:

  1. Physician Association Published Clinical Guidelines
  2. Cohere’s Clinical Advisory Board (review the guidelines regularly)
  3. Cohere Employed Physicians (review the guidelines regularly)

“We want to make sure that physicians are comfortable practicing what is prescribed by their association, which is their peer group instead of a health plan or cohere imposing something on them… Most comes from the associations and the rest is complemented by the advisory board as well as our own internal physicians.”

Cohere supports providers by distilling the latest guidelines and checking the patient’s longitudinal medical record for risks. Providers are responding positively with a 3–4x improvement in provider satisfaction scores and an NPS of 61. Cohere has also seen a 93% provider digital adoption rate.

“There have been some [flagged potential] reactions and patient safety issues. We can actually nudge them out of that saying that ‘Based on the studies, this might not be the right procedure.’ It’s very difficult for physicians to be on top of everything… The physician that sees a patient sometimes doesn’t know what all has happened to the patient. And the patient is sometimes not able to speak to the records… But since we do have a longitudinal record of the patient, we can actually say ‘Hey, they have already gone through this procedure. Don’t repeat it. Or they have already gotten three pain injections just two weeks back. No, you need to wait for another eight weeks to be able to do that.’”

Limitations of “goldcarding” and an alternative solution to overhaul the PA process

In an effort to decrease the volume of PA requests required and the administrative burden on providers, there has been recent legislation passed allowing physicians to be “gold carded.” This removes the PA requirement for certain services if the physician has a track record of high PA approvals. Siva expressed mixed feelings about gold carding regulations, cautioning against static approaches.

I think sometimes the regulations start to prescribe solutions, which don’t keep up with what is happening in the technology…They have prescribed for X months, you will look at the utilization, score them and then give them a gold card for six months…They have prescribed a number ~80–90%… If a procedure is not required, even 90% is not a good score. In 10% of the cases, it can be a patient safety issue.”

Siva highlighted Cohere’s vision for a better alternative — a dynamic, real-time authorization process called green lighting. Instead of providers submitting a PA request, Cohere’s platform would be integrated into the EHR and monitor orders submitted by the physician. If the order is in line with the evidence-based care path and there are no patient-specific risks, there would be no action required from the provider.

The interoperability capabilities required for this background monitoring, or “touchless authorization,” are not currently in place across the industry. As an interim solution, providers enrolled in Cohere’s greenlighting program upload a PDF of their order to the platform — no PA request form required. Siva shared that they are seeing positive results from an ongoing program with the American Academy of Orthopedic Surgeons that allows joint replacement registry participants to join this greenlighting program.

“Physicians will continue to do their work, we will just be listening in on the orders. If the service is ordered and right, we will just let it go. There is no interference. When there is an issue, we will point that out. That’s very dynamic. That’s where we want this to go in the next three to five years… We are addressing the patient issue and making sure that we are not interfering with the physician’s workflow.”

Responsible AI applications in healthcare utilization management

There has been some criticism of the impact of using AI in the PA process. In March 2023, ProPublica published that over a two year period in 2022, Cigna physicians denied more than 300,000 claims as a part of their claims review process that used AI. The article highlighted that Cigna doctors spent an average of 1.2 seconds on each case. With this little time spent by clinical experts on each case, AMA and other organizations have expressed concerns over how automation could negatively impact patient outcomes and keep patients from being reimbursed from the care that they actually need.

Siva addressed these concerns directly, emphasizing that he believes that AI should not be relied on to deny PA requests. Denials are not automated in Cohere’s platform. If a request is flagged, the provider is contacted for a discussion with a clinical specialist.

“We are not interfering with the patient or the provider. We are saying that if the patient needs to get it, let’s use AI to get it faster… Denial is never done using AI…It should not be used for that. It should be reviewed by a specialist who knows the subject matter and who can actually do that discussion. There is a discussion always offered before a denial is done. That’s actually one of the main reasons why our denial rate is very, very low.”

This human involvement safeguards against potential negative impacts on patient outcomes. The platform’s low denial rate demonstrates a commitment to patient safety and satisfaction. Cohere’s platform has led to a 15% decrease in medical cost savings with fewer denials.

Improving the patient experience with faster care approvals and quality guardrails

Cohere’s intelligent prior authorization platform reduces the administrative back-and-forth required to get health plan approval prior to a patient’s procedure or service, helping patients get the care faster. It also provides patients peace of mind that they are getting evidence-based care, avoiding any provider variation and time wasted on ineffective treatments.

In less than a year post-implementation, Humana saw significant patient care quality improvements while working with Cohere.

  • Patients receive care 4 days sooner on average
  • 81% faster patient scheduling
  • 15% shift of inpatient to outpatient setting (arthroplasty and spine)
  • 35% reduction in medically unnecessary surgeries (arthroscopy)

Additionally, Cohere’s data captures a longitudinal view of the patient’s care to date across providers, lifting the burden off of the patient to have to communicate every piece of their care journey. Cohere monitors what the physician is doing and can quickly flag if it conflicts with something in the patient’s care history. Cohere is also looking into how its quality and outcomes data may help providers better inform their members when selecting where they receive care.

Creating an intelligent prior authorization journey for health plan customers

The industry transition to intelligent prior authorization is still in its early stages. Cohere is currently prioritizing transitioning the many health plans and providers still submitting PA requests, status requests, etc. manually. However, Cohere continues to look forward, creating a journey for their customers over the next five years.

For example, Cohere is partnering with Geisinger Health System to reduce the administrative burden in their value based care programs — automating their internal PA process and bundled authorizations based on diagnosis. With bundled authorizations, Cohere removes a large portion of the administrative burden for surgeons, encouraging more physicians to participate in value-based care risk arrangements.

Siva is also focused on taking advantage of the market tailwinds, upcoming regulations and the increase in health plans and providers exploring AI applications, to continue to grow their customer base. Cohere also plans to expand its care pathway products to additional specialties for a more comprehensive view of the patient journey (currently Musculoskeletal, Cardiology, Pain Management, Physical Therapy, and Imaging).

Published On: December 19th, 2023Categories: News, Podcast

Share:

About the Author: Siva Namasivayam

In his third entrepreneurial healthcare venture, Siva Namasivayam is passionate about building companies that are focused on improving the healthcare system. Prior to co-founding Cohere Health and serving as its CEO since 2019, Siva was a founding partner of SCIO Health Analytics which served over fifty Fortune 500 healthcare organizations. Siva has more than 20 years of experience in utilizing technology and data to improve healthcare processes. He holds a master’s in computer science from the University of Pittsburgh, as well as an M.B.A. from the University of Michigan.