If you look at the operating margins on the hospital side, you’ll see that they continue to decline. I’ve noticed many physician groups selling to hospitals, which is likely a defensive and smart move. However, hospitals are struggling just as much as everyone else. They have emergency rooms, more bad debt to manage, and a higher volume of patients utilizing their services, making it a challenging environment. I’ll aim to bring some optimism to this presentation—it won’t be all about the dire state of healthcare.

MedEvolve focuses on what we can control within a framework. When it comes to margin, I view it as a process: the service has been rendered, the claim is coded, billed, sent out, and eventually paid. However, there’s an economic cost tied to the labor required to get paid, along with the question of whether you’re receiving the payment you’re supposed to. No one collects 100% of net revenue—that’s just not realistic. But the real question is, how much are you writing off that could have been collected? Is it 1%, 3%, or even 10%? Most people don’t know this with certainty.

A more important question is: how many human touches are needed to collect what you do collect? Most organizations don’t track this. It’s not just about working the claim—it’s about identifying where in the revenue cycle you’re getting overly involved in fixing problems, which drives up your cost to collect and reduces reimbursement, ultimately impacting margins.

At MedEvolve, we believe in addressing this issue. We built a software platform that sits on top of our own PM system, and it can integrate with any PM system, to measure excessive touches. With this information, you can identify where in the revenue cycle problems occur and implement action plans to mitigate these issues going forward.

Billions of dollars are at stake. I was talking to a friend recently who wanted me to meet the CEO of a tech company. The company has a great idea—it sells to hospitals and offers gamification for nurses to ensure they’re happy, engaged, and committed to their jobs. This, in turn, boosts patient satisfaction. When I met with the CEO, I asked her, “Outside of the typical three-year sales cycle that hospitals require, are you aware of the current revenue cycle problems impacting your clients’ or prospects’ ability to pay for your innovative service?” She wasn’t aware. I advised her to consider not just the clinical side but also the administrative staff who need to be motivated to collect as much revenue as possible, enabling investment in new technologies like her gamification platform.

What always surprises me is that we visit doctors, undergo procedures like knee scopes, or receive treatment for illnesses, yet when I ask physicians in waiting rooms, “Do you know how many touches or people it takes to get paid for the work you’re doing on me?” they often don’t know the answer. It’s concerning that even administrators don’t have this information. But it’s not the fault of doctors or administrators—their systems were never designed to measure these metrics. These systems are purely transactional: get the claim out the door and hope it gets paid. The real issue is the billions of dollars spent on the administrative side of the revenue cycle just to collect payments.

Where are mistakes happening? They can occur anywhere from the point of intake to achieving a zero balance claim. This includes scheduled appointments, urgent care walk-ins, emergency transport, ambulance services, rehab, and more. The goal is to reach that “happy place” of a zero balance claim. Along this journey, various processes, people, and technologies—often sold by vendors at trade shows—are involved. But the key question is: where are the problems actually occurring? You might have a gut feeling about where most issues arise, but do you have objective, empirical evidence to back it up?

When we think about the questions that need to be answered, we realized we had to build a platform to address them. I’m not going to go through all of these questions now—we’ll share the deck with you—but the reality is, most of these questions can’t be answered, and that’s alarming. It’s unsettling to be in an industry facing declining reimbursement with no end in sight, payers using AI to find more creative ways to deny claims, and rising costs for supplies, materials, and labor. Yet, we’re unable to answer basic questions that other industries, like auto manufacturing, figured out over a century ago. Here we are in 2024, soon to be 2025, and we still can’t answer these questions. We need to solve this.

Automation is key. Thousands of companies are trying to automate parts of the process, but none take a holistic view. Coming from the consulting world, I prefer to look at the result first and use deductive reasoning to identify why a problem occurred. But I can’t do that if I don’t first understand where the effort is being applied. This goes beyond just a better AR management solution. It’s about holding everyone in the revenue cycle accountable for the work they’re paid to do, starting from scheduling and pre-registration.

Are you taking the necessary steps at the point of service, or even before, to get a claim out the door clean and paid without human intervention? And when manual intervention is needed, do you have the tools to assign the right work to the right AR team member? Can you measure not only their productivity but also their effectiveness? We’ll dive into some data points shortly. To succeed, you need an end-to-end view of the revenue cycle.

Generative AI is everywhere. I’ve spoken at several conferences this year, with another one coming up in a few weeks. I love attending these shows, but I’ve noticed that every vendor has AI plastered on their booth. Half of them don’t even know what it really means; it just sounds cool. It’s like the buzzwords “machine learning” five years ago or “big data” a decade ago.

The problem is that these AI vendors are only leveraging common data assets—the easy stuff from your claims, PM, and EMR systems. What they’re missing is the structured input from humans, your team, who actually update the status of a claim and explain what actions were taken. By combining both data sources, we can run them through a generative AI engine to produce insights and automation that exceed what the human brain can process.

AI isn’t going away, but we no longer have time for manual data mining, Excel pivot tables, or even dashboards. Instead, insights need to come directly to us: What’s the problem? Why is it happening? And what action should we take? This system needs to learn and adapt over time.

Be cautious about investing in software that relies on the same average data sets we’ve been using for the last 25 years. The promise of AI should go beyond that.

Let’s dive into some of the numbers. We’ve done extensive de-identifying and analyzed various touchpoints. We’ve built a platform that tracks every interaction—no more emails or vague metrics. Every action is tracked and validated. Denials are particularly interesting. For example, with one sizable client over a three-month period, we found that out of 58,000 touches made by their AR team, nearly half were related to denials. For this organization, the cost per touch is about $2.45. A third of their AR team is onshore, while two-thirds are offshore. However, the resolution rate that resulted in payments is only 18 cents on the dollar, which is very low, and their overall resolution rate is 39%.

The annual cost of handling denials, due to errors that could have been avoided earlier—such as at the point of service or during coding—reaches almost a million dollars. Imagine what you could do with an extra million dollars in the bank. Next, consider the types of denials: front-end, coding, and billing errors. For this client, because they didn’t consistently verify eligibility and coordination of benefits (COB), they had to rework 2,800 visits. They’ve already touched each of these cases 1.4 times on average, but they’ve only resolved 67% of the AR, with just 28% of that resulting in actual payments.

By fixing front-end eligibility and COB verification, you increase your team’s capacity, which means you either need fewer resources or can redeploy those resources to focus on more complex AR cases. These are the metrics that matter. It’s not enough to say, “We have a good first-pass rate.” We need to understand the outcome and the effort required to overturn denials.

Let’s talk about wasted touches. We see this pattern every time we take on a new client. Usually, within the first week, the same trends emerge. Let me define what a wasted touch is: it means contacting the insurance company, receiving information, but not needing to take any action. For example, being told that a claim is still in process or that it’s already paid but hasn’t been posted by your team yet.

For instance, one group had 174,000 touches over a certain period, with a cost of $3 per touch. That’s nearly a million dollars wasted. Of those touches, 63,000 were simply checking the status of a claim (like “claim in process” or “claim paid”), making up 36% of total touches within a three-month period.

Now, let’s look at denials. When you analyze overall touches, where do you find the biggest issues? Eligibility errors, missing or invalid information, and even cases where no claim is on file. Is that due to a missing secondary or primary claim? These are questions we uncover by analyzing human data. During implementations, post-implementation, and ongoing client support, we ask, “Why is this happening?” Some issues are straightforward, like missing pre-certifications, while others require deeper analysis.

Another trend we notice is escalation. If 3% of cases are being escalated to a supervisor, we need to ask why. Are staff not trained to resolve issues on the first try? Are there missing protocols? Are certain reps escalating more than others? We need structured, real-time data to make informed decisions.

We also found that resolution rates drop significantly when delegation is involved. For example, if I work on a denied Blue Cross claim for medical records, I may need someone to submit those records, which adds touches. Now, it’s not just me handling it—medical records might touch it, and it could be sent back to billing or even back to me. That’s three touches, and the resolution rate drops below 50% when delegation is needed.

This highlights the importance of streamlined communication. If AR reps need help, they can’t rely on spreadsheets or emails anymore. Every task must be linked to a visit to measure the total economic waste involved in collecting on that visit. Everything needs to be measured—that’s the reality we’re in.

Patient searches are another concern. Even with sophisticated workflow automation, some staff might bypass the system to search for accounts manually. If 36% of your claims are coming from patient searches, it indicates a lack of accountability. The goal is to provide management with actionable insights to change staff behaviors, whether they are onshore or offshore.

Let’s talk about offshore versus onshore. This is a common dilemma: “How effective are my onshore and offshore teams? Should I outsource more offshore or bring more onshore?” I’m not here to criticize offshore work. Many countries now support this, but it’s important to understand the data.

Take a look at this offshore group of insurance AR reps. For instance, reps D and E have significantly higher touches per claim, averaging 1.7. Why? The resolution rates are 22% and 24% for payments, and 39% and 54% overall.

When you analyze individuals, trends emerge. Daily production matters. For inpatient, home health, or behavioral health, reps may work an entire patient’s 28-day stay, so a benchmark could be 100 claims. For neurosurgery, it might be 25. We have benchmarks for various specialties that we can share later. Understanding production is important, but outcomes are more crucial.

Looking at the onshore team, rep K performs well with 1.1 touches per claim, which is the benchmark. Rep K works 100-104 claims daily, resolving 82% of the AR the first time, with 64% resulting in payments. That’s worthy of recognition.

In this group, the offshore team resolves 72% of AR, with 48% in payments. In comparison, the onshore team resolves 51% with 32% in payments. There’s a clear quality difference between the two. Cheaper isn’t always better. Outsourcing offshore to save costs might lead to paying a higher price in the long run, so keep that in mind when considering your AR strategy.

At MedEvolve, we’ve observed the impact of human touch in the revenue cycle, particularly the number of non-actionable touches. In one case, 85% of touches over a three-month period were non-actionable. This represents a significant cost. We also found that resolution rates based on certain actions and statuses are not reliable benchmarks.

For example, in the “claim in process with payer” category, there were 22,000 claims and 23,000 touches. This accounted for 13% of the total work, but only 45% of claims were resolved. If a claim is truly in process, the resolution rate should be much higher, at least 75%.

Similarly, claims that were paid correctly still had 40% of the AR remaining outstanding. This increases the risk of AR representatives repeatedly working the same claims without resolution.

This data helps analyze the entire revenue cycle, from charge entry and coding to payment posting, tasking, billing, and AR management. It’s crucial that everyone involved understands their role in ensuring a seamless cycle. We often see other departments get involved once they understand how their accountability improves first-pass resolution rates, boosts margins, and supports the business’s sustainability.

Timely filing is another issue. While it doesn’t represent a large portion of the AR, 65% of the AR still remains unfiled after the deadline. Denied claims are also a concern, with 30 to 50% of claims denied upfront, 40% of which are front-end issues.

To address this, we developed a financial clearance module to hold every department accountable from intake. Teams like prior authorization, self-pay collections, financial counselors, and eligibility teams are all responsible for ensuring claims are properly handled without requiring back-end intervention. It’s essential to tie actions to results. If someone claims they secured a pre-cert, but the first-pass denial shows otherwise, we can trace the error back to the individual responsible.

When evaluating the health of AR, people often ask, “How’s my AR?” The answer is usually, “It’s pretty healthy,” but it’s not. It’s sort of healthy. I like to focus on AR over 60 days. For insurance AR, if it’s under 20%, you’re in good standing. However, this varies based on specialty. Work comp, legal, arbitration, and billing every 30 days can change these numbers.

How do you link AR health with the people working it? In a study we conducted, we examined locations A, B, and C. Location A’s AR was in poor shape—65% was over 60 days and 48% over 120 days. There were also a lot of touches on these claims, which is concerning. Location A had the largest AR total, the largest amount over 60 days, and 12% of the total touches during this period.

Looking at the representatives working those claims, Rep D, for example, had a 22% resolution rate and worked at locations A and B. By correlating the AR issues with the representatives, we start to see why the AR is in bad shape.

Is it all Rep D’s fault? No. Rep D might be doing their best, but issues that occurred before the claim reached them need resolution too. It could be due to harder denials or other factors. The key is linking the outcomes you see to the actions being taken. Most PM systems don’t provide this capability—they should have done it years ago, but they haven’t.

More touches don’t necessarily mean more effective work. Clients with too few touches can also face AR performance issues. If one person is handling a quarter of their workload but only processing 70 claims a week, it’s a staffing issue. We’ll discuss staffing more later, but it’s important to know if you’re correctly staffed based on your payer, facility, or location, and how your AR strategy aligns with that.

Tasking is crucial. I’ve mentioned it before, and I’ll say it again: when I can’t solve something and need help, even from physicians, it’s essential to measure that. There are two reasons for this: first, it costs money, and second, it affects turnaround time. For example, when I send a task to re-bill claims, I found that the average time to close these tasks between June and August is 19 days. Does this help speed up revenue cycle performance? No, it doesn’t.

Why are we asking someone else to re-bill claims? We need to investigate this. Could we train an AR rep to handle this task?

Adjustments are another issue. If I need someone else to make an adjustment because I’m not authorized to do so, and it takes three weeks, will I see that claim again? Likely. An insurance or self-pay rep will probably see it before the balance is cleared.

It’s vital to ensure these areas are covered before determining who’s touching claims the most. It’s not just an AR problem; everyone in the revenue cycle is impacted when they’re asked to handle tasks outside their role.

Backlog reduction schedules are a common question: Do I have enough people? There are three key components to evaluate: daily production, resolution rate, and average days to follow up. If you’re not producing enough per day or resolving enough per touch, you’ll run into problems. The next factor is how quickly claims are followed up. Do you need to see a claim in a week, or do you just want to? These three components are critical in determining if your staffing levels are appropriate.

For example, let’s look at Rep C. They work 90 claims a day, which seems good. They have 1,300 claims assigned and have already worked 40% of them, but they are pushing claims out every 13 days. This means they have about 1.8 weeks to clear the remaining backlog, which is too short. Ideally, you want a backlog of 2 to 4 weeks. If Rep C were to push claims out 21 to 30 days, they would have more capacity to take on additional work.

Now consider Rep J. They’re working 74 claims a day, with 5,700 claims assigned, far too many. They’ve only resolved 25% of the backlog and are pushing claims out every 5 days. Will Rep J ever clear the backlog? No. Is the AR at risk? Yes. Should the workload be shifted to others? Absolutely.

Having real-time views like this allows for better decision-making and load balancing, which is essential for protecting AR when claims don’t get paid on the first pass.

In closing, it’s important to evaluate revenue cycle administrative staff from a margin standpoint. We conducted a study last year focusing on cash realization rates. This looks at claims that get paid after being worked versus claims that don’t, requiring additional follow-up.

For example, Rep F doesn’t resolve claims on the first touch, which is problematic. They cost $3.57 every time they touch a claim, making them less efficient and likely not a team member you want to keep.

Other individuals have much higher resolution rates and generate more margin per touch. Keep in mind that margin per touch can vary by specialty, payer type, or specialty within payers, so these factors should be considered.

Cash realization rate is crucial. Understanding the true cost per touch and ultimately your margin is key to success. It all begins with measuring every touch. You need to measure everything to level the playing field.

Let’s talk about benchmarks. These benchmarks aren’t found in MGMA, AAOE, or symposiums because those organizations don’t have access to the data from PM systems that can report on them. But we do, so I want to share a few important points for you to consider.

Non-actionable touches should ideally be less than 10%, and no more than 15%. Most of our clients, when they first go live, have over 50%, sometimes even 60%, within the first month. Touches related to denials should be under 10%. I prefer focusing on work effort in denials because the lower the work effort, the fewer denials you’re dealing with.

Regarding front office denials, I hate to set 20% as a benchmark, but the reality is that 45 to 55% of denials come from avoidable mistakes made before service or at the point of care. Many of our clients have decentralized pre-registration or separate leadership for front and back office functions. Everyone in your organization needs to be on board with these principles for a seamless, effective process.

The ideal average touches per claim is 1.1 or less. If your team members are excessively touching claims in a short period—one to three months—it suggests they may be following up too soon or struggling to resolve claims, and might need additional training. Also, pay attention to when claims are escalated to supervisors or team leads, as it can signal challenges.

For unworked claim weeks, I recommend 2 to 4 weeks, with 2.5 weeks being ideal. This depends on daily productivity, resolution rates, claim follow-up frequency, and current backlog, so monitor these closely.

Zero touch rate refers to claims that get resolved without human intervention, excluding clean claim pass rates. An average client typically starts with a 45-50% rate. The best clients achieve around 76% and are improving as they implement process changes and measure where issues occur before claims are filed.

Finally, turnaround time for closed tasks should be two days or less. There’s no reason it should take longer to resolve a claim when another individual or department is involved.

Let’s talk about some key recommendations. Financial clearance is probably the number one priority. It’s essential that the front office staff does their job and has a work driver to hold them accountable. You need to understand how many days out surgeries or high-dollar procedures are. You don’t want to handle these on the same day unless it’s an add-on from the doctor. Once the add-on happens, the clock starts ticking.

Financial clearance is a major gap in PM systems. Despite all the buzz around new tools like AI scheduling and online scheduling, 45% of your denials come from front office mistakes. Stay focused on improving margin in the short term.

Tasking and actions are also crucial. Don’t just rely on your AR team for data. You must track any individual or department working on a claim. Insights from even a single week are invaluable. We’ve had numerous conversations with clients on how to drive accountability and even use auto-tasking to route tasks to the right person from the start. The goal is to reduce touches and improve outcomes.

Protocols and standards are important as well. Daily productivity will vary by medical specialty. The benchmark for resolution rate is 75%. This means 75% or more of AR should be resolved within 45 days of the last action. Some clients start at just 35%, while others have already surpassed 80%. The goal is to resolve claims more efficiently, and ideally, have them paid without intervention.

Upfront protocols and goals are critical. Many bad debt write-offs occur with commercially insured plans. For elective surgeries, the patient owes the money. Where are these claims going? To bad debt. Use digital engagement, e-statements, and text-to-pay. Don’t rely on phone calls—people don’t answer the phone anymore. Don’t use paper statements either, as they are expensive. Always aim to get credit card information on file and patient estimates out before services are delivered.

Workflow and load balancing should be continually assessed. If your AR isn’t where it should be, review your workflow. Who is assigned what? Are you putting your best people on specific payers, locations, or services? This can be quickly identified and adjusted.

Another issue I’ve seen is putting too much on individuals. I prefer centralizing key functions. For example, a biller should focus solely on clearing edits and sending claims out. An insurance follow-up rep should be handling AR calls every day. Payment posters should just post payments, and underpayment reviewers should focus on that task. Centralizing these functions allows you to standardize productivity expectations.

If AR representatives have 10 different tasks, it’s difficult to measure their resolution rate or set productivity standards. You can centralize functions gradually, but start with key areas.

Don’t forget about pre-registration. It’s still too decentralized in the industry. Much of it is still done at the clinic level or by the front desk. Centralize pre-registration offsite and possibly roll it up under CBO leadership for accountability across front and back office. Pre-registration is a huge factor in denials, and no one has a perfect system yet.

Lastly, bringing non-emotional data into discussions can help level the playing field. This isn’t about finger-pointing—it’s about using data to identify why claims aren’t paid. If leadership isn’t open to this approach, they may not be a good fit for the organization. The healthcare landscape has changed dramatically, and providers are the ones struggling. The new administration won’t solve healthcare problems. Big pharma will continue profiting, and insurance companies will report strong earnings, while providers go out of business. It’s crucial to have leaders who understand the new world of measuring touches and reducing them together. This often involves people, processes, and sometimes a technology gap—or technology that’s not being used effectively. That’s why we have a consultative implementation process and ongoing relationships. Simply installing software isn’t enough. It requires active, ongoing work to make it effective.

A quick case study: Atlas Healthcare Partners. They have an amazing leader and an incredible organization, fully committed to the concept of measuring every touch. As a result, they continue to see outstanding results, and many of these statistics are even outdated. The biggest thing I noticed with Heather and her team was the increased capacity for staff. We typically see a 31-45% improvement in capacity. What this allowed Heather to do was avoid hiring when people left the organization. Those who chose to move on to other careers did so, and Heather was able to retain the committed team members without having to hire new staff. If you’re growing, that’s great. Just make sure you have the tools in place to free up your team members so you can grow without the need to hire. It might sound unbelievable, but it happens every time.

A couple of other things to note: close balance NCR went up dramatically, and they’re now over 98%, which is fantastic. Cash acceleration also improved. When we first went live with them, their current AR accelerated by 12%. We’ve seen millions of dollars come off the balance sheet of AJR, which would have likely timed out otherwise. This happened because staff members are held accountable, and the focus is on resolution rate, not just production.

In terms of reducing denials, this organization was particularly focused on our financial clearance module. They were the number one innovators for their pre-registration teams. Heather and the front office team have everything dialed in. They track every check and know exactly what hasn’t been done on a case before surgery. They are currently nine days out, and the results speak for themselves. This approach works when you embrace a culture of measuring everything and solving the underlying problems.

We live in a world where I wish I could say our PM EMR systems will do this for you, but they won’t. They are too focused on other areas, such as physician satisfaction and consumer patient satisfaction, and competing with each other. Can you imagine being a cloud-based PM EMR company right now? I don’t think those salespeople are doing well, as they’re selling a commodity back and forth. I also tell clients not to waste their time and money changing systems, because the new system still won’t give you the ability to measure every touch. We had to build our own system across our revenue cycle and PM to do that, and now we spend most of our time integrating it with other PM EMRs on the market.

The human-generated data being collected here is your valuable data asset. This data will drive generative AI, predictive analytics, and automation. It will let the machine tell you why things are happening and what to do about it. Generic data from a clearinghouse or PM system isn’t rich enough. It doesn’t tell the full story; it only tells part of it.