How I got here
I moved to Nashville for music and stayed for healthcare. I started in nursing school, then found my way into healthcare IT at Community Health Systems in 2011, and spent the next decade and a half inside the revenue cycle at CHS, HCA, and Parallon. I learned the systems the money moves through, Artiva, Epic, Cerner, Medhost, the HL7 feeds between them, and I learned what happens to a hospital's margin when claims come back denied and nobody has time to work them.
I taught myself to write software because the workflows I was handed were broken and nobody was going to fix them for me. In 2016 I started Paradexo to do that work for other health systems. It is still running, bootstrapped, and it has paid for itself many times over. Along the way I cofounded two more companies: a travel business in Peru that still operates today, and a healthcare cybersecurity company that is still in stealth, where I ran operations and helped close the first round.
In 2023 I took a sales job on purpose. I had built products and run operations, but I had never carried a quota, and I did not want to raise money for a company whose founder could not sell. Two years selling revenue cycle software to hospital CFOs taught me that, and it taught me something else: every one of those conversations ended up on denials.
AppealsOne is the company I was always going to start. Denials are the problem I know best, from the provider side, in every role I have held. The piece below is the argument I make in those conversations.
The front end will not stop denials
For fifteen years the industry has invested in the front end of the revenue cycle. Eligibility checks, claim scrubbers, coding audits, pre-bill edits. I have built some of it. It is good work and it is necessary. And denials keep going up anyway.
Roughly 15 percent of claims submitted to private payers are denied on first pass.1 Medicare Advantage plans denied post-acute care at rising rates between 2019 and 2022, and a Senate investigation found the three largest plans using predictive algorithms to drive those decisions.2 ProPublica reported that one national insurer's medical directors rejected over 300,000 claims in two months using a system that let them deny in batches without opening the file, about 1.2 seconds per claim.3
I am not here to say payers are villains. This is a complicated system and both sides are responding to incentives. But the mechanics matter. When a payer can deny at machine speed and a provider can only appeal at human speed, the outcome is decided by bandwidth, not by who is right.
A denial does not have to be correct to work. It only has to be cheaper to write off than to fight.
That is what happens. Up to 65 percent of denied claims are never resubmitted or appealed.4 Not because they are wrong. Because the denials team has a queue they will never get to the bottom of. When providers do appeal, they win more often than not: over half of denied commercial claims are eventually paid,1 and Medicare Advantage plans overturned three out of four of their own denials when beneficiaries pushed back.5 Hospitals are already spending close to $20 billion a year fighting denials, and about half of that is spent on claims that should have been paid the first time.1
Put those numbers together and the picture is simple. Something on the order of $160 billion a year in denied claims is recoverable and most of it is never pursued.6 That money is owed to providers. It gets written off because nobody has the staff to go get it.
This is why I think the durable position in this market is on the response side, wherever the payer says no. Claim scrubbers and coding edits catch the provider's own mistakes, and they should. But the denials that are growing are not the provider's mistakes. They are the payer's decisions, made by software, before the service as a prior authorization denial or after it as a claim denial. Both land on the same overworked team. Both require someone to read the decision, pull the record, find the policy language, write the response, send it, track it, and escalate it. That work is tedious, repetitive, and well defined. It is exactly the kind of work that software, and now AI, can do at the same speed and scale the payers already operate at.
The prior auth market looks crowded from the outside. From inside a physician practice or a small billing company it does not, because the tools that exist were built for hospitals and plug into hospital EHRs. The practices I talk to are still doing auths by phone and fax, and still letting denials age out because the queue never ends.
Providers do not need another dashboard telling them their denial rate went up. They need the auths submitted and the appeals written and sent. That is what AppealsOne does.
- Premier, Trend Alert: Private Payers Retain Profits by Refusing or Delaying Legitimate Medical Claims, 2024. Survey of 516 hospitals: 15% initial denial rate, 54% of denied claims eventually paid, $19.7B spent on appeals, $10.6B of it on claims that should have been paid initially.
- U.S. Senate Permanent Subcommittee on Investigations, Majority Staff Report on Medicare Advantage, October 2024.
- ProPublica, How Cigna Saves Millions by Having Its Doctors Reject Claims Without Reading Them, March 2023.
- HFMA, cited in Health Insurance Claim Denial Statistics: up to 65% of denied claims are never resubmitted.
- HHS Office of Inspector General, Medicare Advantage Appeal Outcomes and Audit Findings Raise Concerns About Service and Payment Denials, 2018. MA organizations overturned 75% of their own denials on appeal.
- Change Healthcare Healthy Hospital Revenue Cycle Index: roughly $262B in claims initially denied, with about 63% judged recoverable, which works out to about $165B. Figures are from 2017 and the denial rate has risen since.