The fight to protect hospital margins in 2027 may begin long before a claim is submitted.
With insurance coverage becoming less stable and payer scrutiny intensifying, health systems face growing risks of preventable denials, delayed reimbursements and lost revenue. The stakes are already high: 43% of U.S. hospitals are operating at a loss, according to Kaufman Hall, while a November analysis from Kodiak Solutions estimated that bad debt and final insurance denials cost hospitals $132.2 billion in net revenue in 2025.
Those pressures could deepen in 2027 as new Medicaid work requirements, more frequent eligibility checks and changes to ACA coverage threaten to push more patients into self-pay or uninsured status. Hospital leaders are already preparing for coverage disruptions and rising uncompensated care. Meanwhile, insurers are deploying AI and automation to review claims at a speed and scale many providers struggle to match, creating new urgency around revenue cycle efficiency and financial sustainability.
For Nichol Cobaugh, vice president of revenue protection at Ensemble Health Partners, the answer lies in shifting from reactive revenue recovery to proactive risk prevention. Ms. Cobaugh spoke to Becker’s about why hospitals must rethink eligibility verification, how AI can help identify denial patterns and protect reimbursement, and why health systems should build their 2027 financial strategies around emerging risks versus historical performance.
Question: As hospitals prepare for potential Medicaid coverage losses, ACA changes and rising uncompensated care in 2027, how significant is the financial threat? Are health systems underestimating it?
Nichol Cobaugh: I think a way to frame the 2027 financial picture is you’ve got coverage disruption, you’ve got payer automation and you’ve got shifting review rules all hitting at the same time. They show up in a few different ways: more avoidable self-pay, slower cash and more denials. Organizations are really in the position right now of figuring out not how to work harder, but how do you move your intelligence upstream and react to that? Coverage, reimbursement and hospital margins are all shifting right now, so that means a lot of things to a lot of organizations and different pieces within the organization.
Q: What are the most proactive health systems doing now to protect their revenue and margins heading into 2027?
NC: Looking at leading indicators and not lagging ones. That means exploring what are going to be potential renewal failures, your pending Medicaid eligibility, what’s your ACA concentration, your self-pay conversions, your charity approvals, initial denial rates. A big one is requests for information volumes and days to pay. Take that data and really look at it prospectively, breaking it down by facility, by service line and — in your larger organizations — by payer plan, because a portfolio average can hide exposure. One hospital with a heavy Medicaid expansion population, for example, can look fine inside a health system average until it isn’t.
Q: What are you seeing around rising self-pay, and how should health systems respond?
NC: I certainly think it’s going to affect access to care. We’re already seeing higher deductibles and stricter coverage requirements.
But self-pay is often a label and not an answer. It means that something broke in the coverage process. For example, there was missed paperwork, a pending Medicaid application, coverage you couldn’t find or financial assistance eligibility.
Coverage gets lost, and I think the more sophisticated organizations — and the ones that are really trying to be proactive around this — are the ones really modeling out what these changes mean to them. A lot of the coverage loss is going to be procedural because they’ve not verified the work requirement or are missing a six-month requalification, not really a choice to go uninsured.
The answer for those organizations is always-on eligibility. You’re screening everyone. You’re revalidating throughout the care journey. You’re connecting the patient intelligence with the clinical intelligence you have, all to feed it back into: How can we get this patient covered? Have we seen a decline in services provided? I don’t think we’ve seen the full extent of what could happen in 2027 if these issues go unchecked.
Q: How is payer automation changing the denial landscape for hospitals?
NC: I don’t think it’s ever an intention to delay and deny care. However, payers are certainly getting more sophisticated, and it’s not what I’m hearing; it’s what I’m seeing.
Denials are getting more clinical. They are more automated. They are coming in at a pace that we’ve never seen before. They are certainly more payer-specific. You’ve heard about some of the nuanced payer rules that have come out recently. We’ve had level-of-severity language change or readmission requirements change.
Payers are reviewing claims at machine speed, while many providers are still on fragmented data and manual rework processes.
[Request-for-information] denials are a great example. Every payer with EHR data exchange still sends RFI denials. We saw volume increase in 2025. We see it continuing to increase in 2026, and nearly nine out of 10 of those pay. Is there substance behind the claim that there’s a delay to pay?
Certainly, there is. More than three-quarters of our final denials are clinical. So payers are overriding physician judgment at scale, and of course they’re using AI and automation.
Q: On the other end of the spectrum, how is Ensemble using AI and automation to help providers respond?
NC: We’re doing it on many fronts, and we’ve got an enormous amount of data.
Number one, we’re engaging with patients in many different ways. That looks different than it historically has, whether that’s preemptively through voice, text or having on-site advocates in place for patients.
From a clinical perspective, we’re tightening documentation requirements. We pair AI intelligence with our clinical expertise, and we’ve seen great outcomes as a result of that. On clinical appeals, for example, generative AI drafts an appeal, but a clinician makes the call.. We’re overturning clinical denials at a rate up to ~7+ percentage points higher when we leverage our AI-backed clinical reasoning engine. It’s not about replacing a clinician, but it lets them spend their time on the argument and not necessarily the paperwork.We also do a lot with large language models and predictive analytics. Every appeal that we win, we feed it back into the model. We flag potential authorization denials, and for clients that are live with our predictive modeling, we’ve seen a significant decrease in authorization denials in that space.
Then we take all of the evidence and data that we have and arm our clients with the ability to go in and have contract language discussions, whether it be on authorizations, turnaround times on RFIs or enforcing two-midnight rules. We’re making sure payers are abiding by prompt-pay rules and having the ability to easily and quickly escalate patterns instead of single accounts, as a lot of organizations have historically.
Combining all of the technology with the expertise that we have has really been able to move us forward over the last few years. We’re able to match the sophistication that payers are putting out there in the market.
Q: How significant is the two-midnight rule as a revenue cycle challenge today?
NC: The two-midnight rule is certainly still a challenge. We’re seeing payers really overriding clinical decisions at scale right now, and the two-midnight rule is a big part of it.
The two-midnight rule is made up of the period of time the patient is intended to be in the hospital in that particular setting, in addition to medical necessity, and payers do not interpret it the same way as a clinician interprets it. That is very apparent in the stats that are out there publicly, and it continues to be a challenge.
To respond to that, you’ve got to have strong clinical documentation. However, it can get exhausting as an organization to have to respond to these issues, so you’ve got to build an infrastructure that allows you to respond to each and every one of them consistently.
I also think you have to collaborate with payers. Providers and payers are clearly reviewing things clinically two different ways. Is there an opportunity for us to come to the table and review cases together and see if we can come up with a better outcome for our patients?
We’ve had some payers that have been responsive to that and some that haven’t. But I would argue that it is one of the most challenging things and is still prevalent within the industry for organizations that are really looking at compliant statuses.
Q: At Becker’s CEO+CFO Roundtable, Mark Cuban encouraged hospital leaders to use AI to analyze healthcare contracts and identify where they may be losing money, asking, “Where are they ripping me off?” How can health systems apply that same approach to payer contracts and claims data to uncover denials, underpayments and other sources of lost revenue? And how can smaller organizations with limited resources effectively leverage these tools?
NC: I think it is absolutely an effective strategy. For smaller organizations, it’s really about being able to focus. Take your top three denials and see what you can do by payer, by provider, by physician, even diagnosis. It will get you much further, much quicker.
At Ensemble, we’re constantly running what I would call “AI drills” to identify potential problems that we can investigate further. It puts us much further ahead than having 10 associates on an accounts receivable follow-up team trying to identify issues and conduct root-cause analysis.
We’ve also been automating some of that root-cause analysis, and it’s been very effective. The key is identifying where to focus your time and resources. You need a targeted plan that prioritizes your biggest risk areas and allows you to address those issues strategically. These tools can be especially beneficial for smaller organizations with limited resources.
Q: As care continues shifting into new settings, how is Ensemble thinking about its broader revenue cycle strategy?
NC: Ensemble’s growth strategy centers on maintaining best-in-class revenue cycle operations and strengthening our partnerships with clients. We’re focused on working closely with our clients and partners to protect their revenue and ensure they’re paid for the care they provide.
There are emerging trends in healthcare — particularly around site of care — that we’re closely monitoring. But our priority remains delivering white-glove service and ensuring we execute at the highest level in everything we do.
We’re also focused on maximizing the expertise within our organization while considering how changes in care delivery may shape the future of healthcare. Those factors inform our broader strategy, but our primary focus remains protecting our clients’ revenue and delivering exceptional results.
Q: What can health systems learn from Ensemble’s work converting self-pay accounts back to insured coverage?
NC: I think the biggest challenge is that eligibility has historically been checked at the front end of care and then forgotten. The industry can no longer operate that way. That’s why always-on eligibility is so important for organizations of all sizes. You need to continually verify coverage throughout the entire patient care journey.
At Ensemble, we screen 100% of our patients daily and revalidate eligibility throughout their care journey. We follow up through phone calls, text messages, digital channels and on-site advocates. To date, we’ve had more than 12 million coverage-related touchpoints with patients.
The key is maintaining that process all the way through claim payment because unexpected coverage changes can arise. For example, a patient with an exchange plan may miss a premium payment, potentially resulting in a retroactive change to their coverage. The more you can anticipate those situations, put safeguards in place and work with patients to maintain coverage, the better positioned your organization will be to protect its revenue.
Eligibility verification can’t be a one-time process. It needs to continue throughout the entire patient care journey.
Q: What is the biggest revenue protection message health system leaders should be thinking about as they prepare for 2027?
NC: Health systems need to see coverage and payer risk earlier, and have to start acting before it just becomes rework that is definite at-risk revenue. You’ve got to protect the revenue that keeps the care available.
For 2027, with all the changes that we’re seeing across the industry, it’s going to be so important that you keep an eye on that. You’re not building a budget off lagging metrics. Organizations can’t rely solely on lagging metrics when developing their budgets. They need to account for emerging financial risks and the challenges they expect to face in 2027, especially as those pressures are likely to intensify in the years ahead.