Days in AR Benchmarks and the Cost of Slow Collections

Days in accounts receivable measures how long a practice waits between delivering care and getting paid for it, and in 2026 that wait is stretching further for specific, identifiable reasons, not random bad luck. This piece walks through the formula, the current specialty-level benchmarks, and the operational causes, chiefly prior authorization and payer-specific denial behavior, that separate practices collecting on time from those slowly bleeding cash. For a practice trying to decide whether to build an internal billing function, outsource it, or adopt a platform that automates the claim-tracking work, the benchmarks and mechanisms below are the criteria that actually distinguish one option from another.

How the AR days formula works

Days in AR is a clock. It counts the days between the moment a service is rendered and the moment payment lands, and every day a claim sits beyond the benchmark for its specialty is money a practice has already earned but cannot yet use. The standard formula divides total accounts receivable by average daily charges, but the quality of that number depends entirely on what goes into the denominator. Net charges are billed charges after you subtract contractual adjustments and expected write-offs, and they give a far more honest result than gross charges, because gross charges count money a practice never expected to collect. A denominator inflated with phantom revenue makes the AR days figure look better than it is and defeats the purpose of measuring it.

Two methods dominate how practices actually run this calculation. The first, often called the monthly net charges method, takes total AR at month-end and divides it by net charges for that period, then divides again by the number of days in the period. This version is useful for spotting a trend inside a single quarter, since it reacts quickly to a bad month or a strong one. The second, recommended by MGMA for benchmarking purposes, divides total AR by annual gross charges and then by 365, which smooths out seasonal spikes, like the volume surge many practices see at the start of a plan year, and gives a steadier number for comparing performance across time or against peers. Neither method is wrong; they answer different questions. A practice tracking its own trajectory month to month wants the first. A practice trying to see where it stands against the rest of its specialty wants the second.

It helps to be clear about what this number can and cannot tell a practice. HFMA's MAP Keys framework classifies AR days as a lagging indicator, meaning by the time it moves, the underlying problem, whether that is a denial, a missing authorization, or a coding error, has already happened and the cash is already delayed. The rest of this piece works backward from that lagging number toward the leading indicators, denial rates, prior authorization volume, payer-specific behavior, that predict whether AR days will climb before the damage appears in the month-end report.

2026 benchmarks by specialty and performance tier

The optimal AR days range cited across the industry is a cross-specialty average, and MGMA and HFMA both use it as a general target, not a standard that applies evenly to every type of practice. A number that looks strong in one specialty can signal a genuine problem in another, so if you compare a practice's AR days only to the general target without accounting for specialty, you risk both false comfort and false alarm. More than 500 rendering providers sit in the 2026 benchmark dataset, and the data makes that variation explicit. Primary care is at the favorable end of the distribution, with among the lowest median AR days of any specialty tracked, a below-average denial rate, an above-average clean claim rate, and an above-average net collection rate, a combination that reflects relatively standardized coding and lower prior authorization volume. Radiology lands in the mid-range on median AR days, but it carries among the highest denial rates of the specialties in the dataset, and its clean claim and net collection rates fall notably below the top performers, because of the volume of imaging studies that need prior authorization. Behavioral health runs above the general target range on AR days, with a high denial rate and among the lowest clean claim and net collection rates of any specialty measured, because of both authorization friction and the documentation requirements specific to behavioral health claims. Oncology posts the longest median AR days of the specialties in the dataset, and its denial rate is high while its clean claim and net collection rates sit in the lower portion of the specialty range, because cancer treatment billing carries so much complexity and authorization burden.

Specialty placement is only half the picture; a practice also needs to know where it sits within the broader distribution regardless of specialty. Looking at the 2026 all-specialty combined data by percentile gives a ladder a practice can use to locate itself. At the 10th percentile, AR days are comfortably inside the general target range, so a practice here is performing well. At the 25th percentile, AR days are at the lower boundary of that target range, still healthy but with less cushion. At the 75th percentile, AR days run meaningfully above the target range, a sign that aging claims have started to accumulate faster than they are being resolved. At the 90th percentile, AR days run substantially above the target range, and that usually means a practice has lost control of its denial management, its authorization workflow, or both.

The share of total AR aged past a certain threshold matters as much as the average days figure, because the oldest AR is the money most likely to be written off, with little chance of eventual collection. MGMA and HFMA benchmark targets set a ceiling on how much of a practice's total AR should fall into that oldest bucket, and a practice that stays within the general AR days target but allows a growing share of its AR to age past that ceiling is still losing money, just more slowly and less visibly. The industry's net collection rate target is set at a high threshold for the same reason: net collection rate measures how much of the revenue a practice is actually contractually owed it eventually collects, and falling below that threshold means collectible revenue is being left on the table even if the AR days number looks acceptable on its face.

AR days are getting worse across the industry, and the causes are concentrated

The 2026 benchmark data shows a clear divergence, so the industry is not declining uniformly. Specialties with high prior authorization volume, radiology, oncology, and behavioral health among them, absorbed most of the year-over-year slippage in AR days, while specialties with established peer-to-peer escalation paths and dedicated authorization staff saw far less drift. That pattern matters because it means the deterioration in collection speed is not evenly distributed bad luck; it is concentrated in exactly the practices and specialties that depend most heavily on prior authorization to get paid.

The mechanism connecting authorization volume to AR drift runs through the denial rate. A denied claim is, by definition, a claim stuck in AR: it is neither paid nor written off, and it stays in that limbo until someone on staff catches it, builds an appeal, and resubmits it. If a denial sits unworked for an extra week, that week gets added directly to the practice's AR days. Looking at what actually causes initial denials makes clear how preventable most of this is. Miscoded or incomplete claims account for a large share of denials, as do eligibility and administrative errors that a front-desk verification step would have caught. Excluded or non-covered services, missing prior authorization, and a lack of documented medical necessity round out the leading causes. None of these are exotic payer tactics; they are operational gaps, most of them fixable with better front-end processes and faster denial response. What happens after a denial lands is the deeper problem: a large share of denied claims are never resubmitted at all, and the revenue tied to them is simply abandoned.

This trend data describes physician practice billing operations, not hospital systems, and the distinction matters because hospital revenue cycle dynamics, with their larger billing departments and different payer mix, do not transfer cleanly to an independent practice weighing its own staffing and technology choices.

Prior authorization as a structural drag on AR days

Prior authorization functions as a bottleneck at both ends of the revenue cycle. At the front end, it delays the start of billing until approval comes through, and at the back end, a missing or mishandled authorization becomes one of the most common grounds for denial, which then restarts the whole collection clock. In between, it consumes staff hours that could otherwise go toward working the claims already sitting in AR.

Prior authorization carries substantial administrative weight. MGMA's Annual Regulatory Burden Report found that the vast majority of medical groups have hired new staff or reassigned existing staff specifically to handle prior authorization volume, and a majority of practices report that at least three employees are typically involved in completing a single authorization request. That is not a rounding error in a practice's staffing budget; it is a meaningful share of administrative labor going toward a task that, for most payers, still cannot be done efficiently. The 2025 CAQH Index shows that electronic adoption of medical prior authorization lags far behind adoption of claim status inquiry and claim payment transactions, so most authorization work still happens by phone, by fax, or through manual portal entry rather than through the faster electronic channels the rest of the claims process has largely moved to.

What makes this gap especially costly is that it is not just a workflow problem, it is a knowledge problem. Knowing which payer routes a given authorization through which review vendor, what documentation each one requires, and which escalation path actually gets a stuck request unstuck is institutional knowledge that a practice accumulates claim by claim, denial by denial. Most practices lose a meaningful share of that knowledge every time a biller leaves, because it rarely lives anywhere except in that person's head. You need to understand the AR days formula first, but acting on it takes real-time visibility into which claims are aging and why, and most practices lack that visibility unless they have dedicated billing staff or systematic infrastructure built to monitor AR movement as it happens. Once a practice has located itself on the performance ladder described above, the next question is whether its billing operation retains the institutional memory to notice when a payer's denial patterns shift, or when a particular claim type starts aging faster than it used to, the kind of payer-specific intelligence that compounds in value over time and evaporates the moment the staff member who built it walks out the door.

How payer behaviors add to AR days

Payers do not behave uniformly, and the gap between a claim that clears quickly and one that ages past the point of likely collection often comes down to whether a practice knows which plan uses which review process and which escalation path actually works. Prior authorization denial rates vary sharply by payer type. KFF's analysis of 2025 data found that Medicare Advantage insurers denied 12% of standard prior authorization requests, Medicaid managed care insurers denied 14%, and ACA Marketplace insurers denied 18%, the highest rate among the three categories. Even within a single payer type, the variation is wide: among the six Medicare Advantage insurers KFF examined individually, the share of standard prior authorization requests denied ranged from 5% at Elevance to 17% at UnitedHealth Group, a more than threefold difference between plans a practice might treat as interchangeable.

That variation is not evenly distributed by accident. A federal watchdog review found that the three largest Medicare Advantage organizations by enrollment denied prior authorization requests for long-term acute care and inpatient rehabilitation at some of the highest rates among their peers, and those denials piled up in exactly the post-acute care categories where a delayed or denied authorization can strand a patient's claim in AR for months. Some major plans route specific categories of prior authorization, advanced imaging, certain surgical specialties, oncology services, through third-party review vendors with their own proprietary criteria and their own submission portals, separate from the payer's general claims system. A practice that submits those requests through the wrong channel, or that does not know a particular service line has been carved out to a specialty vendor, loses time it did not need to lose.

The exposure is not limited to the point of initial authorization, either. Payer takebacks after payment, where the payer pays a claim and then later reverses or reduces that payment, rose in 2026. Denials are increasingly arriving after the money has already been counted as collected. A practice that monitors only its initial denial rate will miss this second wave of exposure entirely, because by the time the takeback happens, the claim has already cleared the AR aging reports that would normally flag a problem. None of this payer-specific routing information, which vendor handles which service line, which escalation path a given plan actually responds to, lives in a policy manual a new biller can read on day one. It accumulates through claim history and direct experience, and because that knowledge is specific to each payer relationship, it resets to zero every time the person who built it leaves the practice. Since denial rates drive AR aging from the start, catching denials the moment they land and working them immediately, rather than batching them for a weekly or monthly review, is where a practice recovers the largest share of otherwise-lost revenue. Services such as Altair Health integrate with a practice's existing patient-records and clearinghouse systems to surface denials in real time without requiring the practice to change its existing workflow, and they are built around exactly that gap between when a denial happens and when a human finally notices it.

What the 2026 CMS prior authorization rule changed

CMS-0057-F gives billing teams a new set of tools for pushing back against slow or vague prior authorization decisions, though the rule only helps a practice that understands what it now requires payers to do. The underlying regulations took effect April 8, 2024, but the operational provisions that matter most to day-to-day billing work became effective January 1, 2026. As of that date, Medicare Advantage, Medicaid, and CHIP payers, along with Qualified Health Plan issuers on the Federally Facilitated Exchanges for the specific denial-reason and API and metrics requirements, must accept electronic prior authorization requests, respond within a short window for urgent requests and a longer defined window for standard requests, and provide specific, actionable denial reasons in place of generic boilerplate.

That last requirement is the one that changes the most for billing teams in practice. Before January 2026, a denial could arrive stamped "not medically necessary" with no further detail and no way for a biller to know what evidence or documentation would actually overturn it. Under the new rule, a payer has to state the specific reason for the denial, which gives a billing team a direct line to the correct resubmission or appeal argument instead of guessing at what the payer wanted the first time. The 2026 Final Rule also prohibits Medicare Advantage organizations from retroactively denying or downgrading inpatient care that was previously authorized, closing off one of the pathways payers had used to claw back payment after the fact. For a billing team that knows the rule's specific deadlines and documentation requirements, this is leverage that did not exist before January 2026. For a team that does not, it is just another regulation sitting in a binder.