At-Risk Patients Are Now 77% of What You're Trying to Collect. Most Billing Systems Were Built for the Other 23%.

Ask most RCM leaders to describe the patient they're trying to collect from, and you'll get a version of the same person. Commercially insured. Stable address. Reliable email. Logs into the portal when a statement arrives. Pays within 30 days, or close to it.

That patient exists. They just aren't where most of the outstanding money lives anymore.

Recent industry research, surveying 4,150 adults and drawing on 1.5 billion patient interactions, mapped how the financial side of healthcare actually works in 2026. The headline finding: nearly 77% of patient out-of-pocket dollars fall into difficult-to-collect cohorts. Uninsured, underinsured, digitally disengaged, or managing bills that are complicated enough to create friction on their own. These patients delay payment at twice the rate of their easier-to-collect counterparts and report 62% lower satisfaction with the billing experience.

The quieter finding, buried deeper in the same research, is the one worth pausing on. Analyzing 10 million bills, the researchers found that high external propensity-to-pay scores actually yielded 51% less than bills with no score at all. Translation: the risk models most billing systems use to prioritize collections are, in many real cases, working against the practices that rely on them.

That's worth sitting with for a moment. The billing infrastructure in use at most groups, health systems, and RCM companies was designed around a patient who is commercially insured, digitally engaged, and financially stable. That patient is now a minority of the revenue at stake.

How the Patient Population Actually Shifted

Three forces have reshaped who's paying healthcare bills, and how.

High-deductible health plan enrollment is up 65% in the past decade. A patient with a $3,000 or $5,000 deductible is technically insured but functionally paying out-of-pocket for most routine care until they hit that threshold. Their payment behavior looks more like an uninsured patient than a traditionally insured one.

Coverage churn has accelerated. 20 million individuals are experiencing premium increases following the expiration of enhanced Affordable Care Act subsidies. An estimated 10 million more are expected to lose Medicaid coverage through new work requirements. A patient who was insured when they received care in January may be uninsured by April when the bill arrives.

Uninsured balances are growing faster than any other segment. Across modern patient payment platforms, uninsured balances now represent a meaningful and rising share of collectible dollars, with year-over-year growth running well above other segments over the past several years. For behavioral health groups, the uninsured and underinsured share is often much higher than the industry average. For PT clinics with high cash-pay volume, the dynamic is similar.

The net effect: the patient your billing system is trying to collect from has changed. They're not less willing to pay. They're navigating different circumstances than the patient most billing systems were designed to serve.

Why Traditional Billing Fails This Cohort

The friction points aren't mysterious. They show up consistently in the research and, frankly, in any conversation with a billing team that's been paying attention.

Patient portals are the most obvious example. Setting up a portal account requires an email address the patient checks regularly, a password they'll remember later, and a device they can log into weeks after the encounter when the bill finally arrives. Patients dealing with housing instability, domestic situations, or multiple part-time jobs can't always maintain that continuity. The portal that was supposed to be a convenience becomes a barrier.

Paper statements have the same problem in a different form. Between 7 and 12% of Americans move in a given year, with significantly higher rates among lower-income populations. A statement sent to last month's address doesn't generate a payment. It generates a write-off.

Collection cadences calibrated for the financially stable patient miss the mark, too. Traditional collection workflows tend to assume that if the patient doesn't pay within 30 days, they're choosing not to. The 2026 research suggests something different: at-risk patients are 3x more likely to report that billing outreach feels untimely. The timing isn't wrong in an absolute sense. It just doesn't align with when the patient actually has the ability to pay.

Payment options on offer often don't match what the patient can realistically commit to, either. 30% of patients report that available payment options are unaffordable, and one in four report difficulty covering medical bills. Offering a 3-month or 6-month payment plan to someone whose income swings by $800 a month doesn't solve the problem. It just shifts the point of failure from "couldn't pay the full amount" to "couldn't pay the installment."

What Actually Works for Harder-to-Collect Populations

Research is converging on a clear set of principles.

Meet the patient where they already are. Text messaging reaches 95%+ of mobile phone users and doesn't require any account setup. A patient doesn't need a portal login, a password, or even a specific device to pay by text. They receive a branded message, verify their identity (typically with a date of birth or last four digits of an ID), see their balance, and pay. The entire interaction takes under two minutes. This works because it meets people in the channel they already use for everything else in their lives.

Offer real payment flexibility, not theoretical flexibility. Payment plans that adjust to the patient's actual cash flow (biweekly instead of monthly, variable amounts based on pay cycles, the ability to pause and resume without going to collections) work better than rigid installment plans. The patient who misses one $200 payment and then gets pulled out of the payment plan into a collection workflow ends up paying nothing. The patient who can pay $75 one week and $125 three weeks later often pays the full balance over time.

Stop treating portal login as engagement. The research is clear that surface engagement metrics like portal logins and email open rates can mask deep dysfunction for at-risk patients. A patient who opened the statement email but hasn't logged in to pay isn't "engaged." They've received the bill and aren't acting on it. The collection workflow needs to recognize that distinction and respond differently.

Assume the worst-case contact path. Build the billing experience so that a patient can pay even if they no longer have the same phone number, email address, mailing address, or insurance status they had when they received care. The more of these that are required to complete a payment, the more the system filters out the patients who are hardest to reach.

The Uncomfortable Implication for RCM Companies

If you run a billing operation or an RCM company, the research forces an uncomfortable question: how much of the traditional playbook is actually working, and how much is optimizing for the subset of patients who were going to pay anyway?

The 23% of patient responsibility that falls into easier-to-collect cohorts largely takes care of itself. Those patients have stable coverage, stable addresses, stable incomes, and the bandwidth to navigate a portal. They pay because paying is straightforward for them.

The 77% that falls into harder-to-collect cohorts is where the actual work lives, and it's where most billing operations spend their staff time on approaches that don't match the population. Collection calls to disconnected numbers. Mailed statements to old addresses. Portal-based payment experiences that assume infrastructure the patient doesn't have.

The shift isn't about lowering expectations or accepting higher bad debt. It's about recognizing that the billing infrastructure most practices and RCM companies rely on was designed for a patient population that's now a minority of the revenue at stake. Reorienting around the actual cohort means fewer required steps, more channels, more real flexibility, and less reliance on the assumption that the patient has the bandwidth and stability to come to you on your terms.

Groups and billing companies that make that shift will start collecting meaningfully more of the 77%. The ones that don't will keep optimizing around the 23% that was always going to pay, and will watch the rest of the balances quietly migrate to bad debt. We'd rather see operators working on the harder-to-collect side of their AR with tools built for it, which is most of why we built what we built.

See how it works