How to Find Money in Medical Bill Errors - Zoe Holderness - Co-Founder of Slingshot

The podcast features an interview with Zoe Holderness, CTO and co-founder of Slingshot, a software company dedicated to auditing medical bills for self-funded employer groups. The discussion centers on the prevalence of medical billing errors, the administrative complexity of the U.S. healthcare system, and how Slingshot leverages software engineering and data analysis to achieve significant cost savings for both employers and their employees. Holderness details her personal journey, which began after successfully fighting an overcharged gynecology bill, leading her and her co-founder to recognize a scalable business opportunity in automating the auditing and recoupment process.

Slingshot’s core value proposition is its ability to perform continuous, post-adjudication audits of every claim that flows through an employer’s plan data stream. By integrating with data sources like TPAs or partners like navMD, Slingshot ingests claims data (including CPT codes, diagnosis codes, and procedure codes) and runs it through proprietary algorithms to identify objective errors, such as incorrect modifiers, unbundling (charging for components separately when a single, cheaper code exists), or pattern abuse (e.g., a provider consistently billing maximum severity levels). This post-adjudication approach is critical, as it eliminates "noise" and stress for the member; the provider has already been paid, and Slingshot’s case managers only contact the member when a check is being sent back, resulting in a high Net Promoter Score. Slingshot reports a 70% recoupment rate on identified errors, translating to an average savings of 5% to 10% of the medical plan spend.

Beyond claims auditing, Slingshot addresses another major administrative failure: the underutilization of financial assistance policies at non-profit hospitals. Non-profit hospitals are legally required to offer community benefits, often including financial assistance for patients whose household income falls below a certain threshold—sometimes as high as 400% of the federal poverty level (up to $120,000 for a family of four). Slingshot uses employer income data to proactively identify members who qualify for this assistance. They then manage the complex application and follow-up process with the hospital, ensuring the member’s out-of-pocket portion is written off. This service is enhanced by saving member financial information, allowing for rapid re-application if the member or a family member requires subsequent hospital visits, significantly reducing the administrative burden on employees.

Holderness emphasizes that the ability to perform this service at scale is a recent development, largely driven by the transparency requirements introduced by the Consolidated Appropriations Act (CAA), which mandates that employers have access to their claims data. Her long-term "moonshot" vision is to leverage transparency and technology to reduce the cost of healthcare for the U.S. consumer by 30%, viewing Slingshot’s current audit and assistance services as foundational tools necessary to achieve this systemic change. The company’s success hinges on combining technical expertise (software engineering and machine learning) with the willingness to perform the necessary human follow-up and dispute resolution with providers.

Key Takeaways:

• Leveraging Post-Adjudication Audits for Member Experience: Slingshot operates retrospectively (post-adjudication), meaning claims are audited after they have been paid. This strategy minimizes member disruption, as the first contact a member receives is often notification of a refund check, leading to a positive perception of the solution and eliminating the stress of balance billing.
• Quantifiable Savings from Coding Errors: Slingshot’s rule-based algorithms identify objective billing errors, such as "unbundling" (using multiple codes for a procedure that should be covered by a single, cheaper code) or incorrect modifier usage. These audits yield a high success rate, with approximately 70% of disputed claims resulting in recoupment, saving an average of 5% to 10% of total medical plan spend.
• Data Access as a Prerequisite for Action: The ability for self-funded employers to access and utilize their claims data (often via CSV files or TPA feeds) is the essential first step for any actionable cost-saving solution. The Consolidated Appropriations Act (CAA) has been instrumental in making this data accessible, enabling data-driven solutions in healthcare.
• The Power of Continuous Auditing: While retrospective audits (looking back up to a year) are possible, continuous auditing (receiving and processing claims data weekly) yields the highest success rate because providers are more likely to correct recent errors than those from many months prior, which improves the 70% recoupment rate.
• Financial Assistance Policies are Underutilized: Non-profit hospitals are mandated to offer financial assistance, often covering patient out-of-pocket costs for families earning up to 400% of the federal poverty level (e.g., $120,000 for a family of four). This resource is poorly advertised, and employers should proactively seek solutions to help members access it.
• Automating Financial Assistance Applications: Slingshot automates the complex process of identifying eligible members, confirming their income, applying for assistance, and performing the necessary follow-up with the hospital. This transforms a task that typically takes an individual 20 hours into a low-work confirmation process for the employee.
• The Need for Human Case Management in Dispute Resolution: While software identifies the errors and patterns, the critical step of dispute resolution—contacting the provider, presenting the CPT guideline rationale, and ensuring the money is sent back—requires experienced case managers and can take two weeks to three months.
• Focus on Objective, Rule-Based Errors: Slingshot focuses exclusively on errors that are highly objective and rule-based (e.g., NCCI rules, CPT guidelines). This ensures the audits are defensible and leads to a higher recoupment rate, as opposed to subjective disputes.
• The Moonshot Vision for Healthcare Cost Reduction: The long-term goal is to leverage transparency and technology to achieve a 30% reduction in healthcare costs for the U.S. consumer, suggesting that the current level of administrative waste and error represents a massive, addressable portion of healthcare spending.
• Machine Learning for Abuse Detection: The platform uses machine learning algorithms to detect patterns of abuse, such as a provider consistently billing for the highest level of severity (Level 5 E/M codes) when the median level of care (Level 3) would be expected, providing actionable intelligence to the employer.

Tools/Resources Mentioned:

• navMD: A data integration partner that provides existing relationships and data streams for claims data access.
• Y Combinator: The accelerator program that provided funding, mentorship, and business education to Slingshot's founders.

Key Concepts:

• Unbundling: A billing error where a provider uses multiple procedure codes to describe services that should have been covered by a single, comprehensive code, often resulting in overcharging.
• Post-Adjudication Audit (Retrospective Audit): Reviewing medical claims after they have been processed and paid out by the payer (TPA/insurer). This approach minimizes member disruption and stress.
• 400% of Federal Poverty Level (FPL): A common threshold used by non-profit hospitals for determining eligibility for financial assistance, which can cover the patient's out-of-pocket portion of the bill.
• Consolidated Appropriations Act (CAA): Legislation that, among other things, strengthened employers' fiduciary duty and access to their claims data, enabling data-driven solutions like Slingshot.

Examples/Case Studies:

• Founder's Personal Experience: The company was founded after Zoe Holderness successfully reduced her own gynecologist bill by 40-50% after requesting an itemized bill, reviewing CPT codes, and requesting four separate audits.
• Hamburger Analogy for Unbundling: Unbundling is described as being charged for the bun, lettuce, and tomato separately, rather than just the single price of a hamburger, illustrating how multiple codes are used to describe a single service.
• ER Level 5 Severity Pattern: Slingshot’s machine learning algorithms look for patterns, such as an Emergency Room provider consistently billing for the highest level of severity (Level 5), which is statistically unlikely given the median level of care typically required in an ER setting.