Why getting paid keeps getting harder, with Roshan Patel of Arrow

Getting paid for services already rendered shouldn't be this hard. But for most physician practices, navigating insurance payments has become one of the most frustrating and resource-intensive parts of running a business.
In this episode, Roshan Patel, founder and CEO of Arrow, joins Medical Economics Managing Editor Todd Shryock to break down why health care payment friction keeps getting worse — more prior auth requirements, more denials, more fragmented systems and a chronic shortage of trained billing staff to manage it all. Patel walks through the three routes practices typically take when payment processing gets out of hand, explains where AI is genuinely helping in revenue cycle management versus where it's still more hype than substance, and argues that denials management is the single highest-impact area for technology right now.
 
Music Credits:
MUCHOS BESOS by Bopper Beats  - stock.adobe.com
A Textbook Example by Skip Peck - stock.adobe.com

Editor's note: Episode timestamps and transcript produced using AI tools.

0:00 – 0:28 | Sponsor message Copic medical liability insurance.

0:28 – 0:54 | Cold open Patel previews the episode's central tension: in health care payments, one side wants to get paid and one side doesn't want to pay — making it almost a zero sum game by design.

0:54 – 1:48 | Introduction Austin Littrell introduces the episode and previews the conversation with Patel.

1:48 – 2:42 | The biggest pain points in medical payment processing Patel describes the current state: constantly changing payer rules, more prior auth requirements, rising denial rates and increasingly fragmented systems — all adding up to mass frustration across the industry.

2:42 – 4:14 | Why hasn't this been fixed? Patel explains the structural problem — misaligned incentives, fragmented payer systems, information that's nearly impossible to access without calling the insurer directly, and a staffing pipeline that keeps burning out the specialized people needed to manage it all.

4:14 – 5:28 | Three routes practices take Patel lays out the options: hire more in-house staff, outsource to a medical billing company or adopt technology. Each has tradeoffs, and knowing which fits your practice depends heavily on size and growth stage.

5:28 – 8:34 | How AI actually fits into revenue cycle management Patel distinguishes between two AI models — full automation for lower-stakes tasks like scribing, and a co-pilot approach for higher-stakes work like billing where a human needs to stay in the loop. He walks through specific use cases, including clean claims checks and AI-assisted appeal letter writing, where tasks that took hours now take minutes.

8:34 – 9:26 | P2 Management Minute Keith Reynolds shares practice management tips and invites listeners to submit their own workflow ideas.

9:26 – 10:45 | Patient payments and the system-of-record problem Patel addresses the gap between insurance and patient payment systems, explaining why most practices can't answer basic questions about their own collection rates — because there is no single source of truth, and the EMR was never really built to be one.

10:45 – 12:10 | How to modernize your payment systems Patel's advice: skip the Google search and ask peers what's actually working. Smaller practices should look for one or two tools that do most of the work; larger practices can afford point solutions. Conferences are an underrated place to vet vendors in person.

12:10 – 13:15 | What metrics to demand from vendors Patel says practices should know their own collection rate and average time to payment before approaching any vendor — and then hold that vendor accountable to moving those specific numbers, not just general promises.

13:15 – 14:24 | Where technology makes the biggest difference Denials management. It's the most labor-intensive part of revenue cycle — vague denial reasons, phone calls to insurers, appeal letters, follow-up — and the area where Patel sees the clearest case for technology.

14:24 – 15:51 | The five-to-ten year outlook Patel pushes back on the idea that AI will replace medical billers, arguing that the specialized institutional knowledge experienced billers carry is something AI can't replicate. His prediction: billers stick around but become significantly more productive — and health care payments never fully loses its friction.

15:51 – 17:10 | Outro Shryock closes the interview. Littrell thanks listeners and reminds the audience to subscribe and visit MedicalEconomics.com and PhysiciansPractice.com.