The Level Index
HealthcareHow does your practice compare?
Revenue cycle, overhead, and provider productivity benchmarks for U.S. medical, dental, physical therapy, and specialty practices. Use named source context and separately labeled planning inputs; exact verification status appears beside each reference.
Reviewed October 7, 2026. Published figures retain their own observation periods and populations. Retained source-attributed inputs with no exact capture are explicitly unverified; planning references are not a measured Level industry cohort.
$5.3T
Market size
U.S. health spending (2024) (retained historical reference, exact source not reverified here)
224K+
Establishments
Physician offices in the U.S. (retained historical reference, exact source not reverified here)
135K+
Establishments
Dental practice establishments (retained historical reference, exact source not reverified here)
Method
Source and scope
Compare each defined metric with its own population and reporting window
Check your own numbers
Want to know where you actually rank? We'll show you and what to fix first.
Most owners read benchmark pages like trivia. The useful question is which number in your own P&L, payroll, billing, or operating data you trust least. Your numbers stay private, we never publish or share client data.
In the free audit, we check:
- •labor cost as a percentage of revenue
- •gross margin by job, account, or location
- •cash tied up in receivables and open work
We use this to prepare your audit before the call. Your numbers stay private.
About the data and its source basisProvenance and method for the healthcare figures on this page.
About the Data
Named source observations and separately labeled planning references answer different questions. Each verified capture retains its population and date. Previously attributed inputs without an exact retained capture are marked unverified, and do not establish a measured Level industry cohort.
Methodology
Read the evidence class and population beside each metric. A report attribution alone does not verify a value or percentile. Mixed-source table bands and planning inputs are not common quartiles; published percentiles apply only to the captured source population.
The Level CLEAR Framework
Five pillars of healthcare financial health
Every metric in the Level Index maps to one of five pillars. Together they give you a complete picture of where money is made, lost, stuck, or at risk.
Track insurance cash, patient cash, claim status and collection dates separately. Aged balances require investigation rather than automatic write-off.
Compare provider FTE, collections, clinical hours, support staffing and no-show records on the same period. Test changes before claiming a productivity effect.
Reconcile net collection, procedure costs and overhead. A hypothetical three-point overhead difference on $2M revenue equals $60K, before transition costs and other changes.
Review referral patterns, payer mix, eligibility, authorization and coding exceptions. A denial rate alone does not explain preventable revenue loss.
Review payer concentration, malpractice coverage, coding controls and compliance obligations. Model concentration scenarios with actual contracts.
Key Finding
A Marketplace denial statistic does not establish your practice denial rate.
KFF's corrected 2023 HealthCare.gov insurer analysis reports 20% of in-network claims and 36% of out-of-network claims ultimately denied. Initially denied claims later resubmitted and paid are excluded. These final insurer-level populations differ from a practice's adjudicated first submissions. Reconcile your own submitted, denied, appealed and paid claim cohort before applying a comparison.
Use eligibility, authorization and coding exception queues to test a concrete defect, then measure actual recovered cash and rework.
Source: KFF corrected 2023 Marketplace final-denial analysis (updated March 24 2026)
Use payer mix and collection cutoff before comparing net A/R days.
Days in Net A/R
Source / sample: Level planning cases, not HFMA chart values
The displayed 55, 50, 40, 32 and 28-day inputs are Level planning cases, not HFMA figures. The captured 2024 HFMA hospital/system winner group has source performance P90/P10 positions of 30.5/45.3 net A/R days; it is separate from the three physician-practice winners. Investigate eligibility, charge lag, denials and payment timing with your own records rather than diagnosing a cause from days alone.
Source basis
Definition: Net eligible patient receivables divided by matching average daily net patient-service revenue. The displayed days are planning cases; current HFMA FM-1 definitions require eligible balance and revenue inclusions/exclusions, and source-specific historical values require their own definition check.
Source: Level planning cases; HFMA 2024 selected hospital/system winner context (P90/P10 positions 30.5/45.3 days)
Point-of-service cash share depends on the collection definition and population.
POS Cash Collection (% of total self-pay cash collected)
Source / sample: 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners
The five values match the 2024 HFMA source performance positions for 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners. They measure patient POS cash as a share of total self-pay cash collected, not recovery of billed patient liability. The separate three physician-practice winners have source performance P25/P75 positions of 50.7/58.4%; neither selected group is a national practice benchmark or proof that collecting earlier produces a particular dollar gain.
Source basis
Definition: HFMA PA-7: patient POS payments / total self-pay cash collected for the reporting month. Qualifying current-encounter cash includes pre-service, time-of-service and up to seven days after discharge; qualifying prior-encounter self-pay cash collected before or at a new service is also included; routine payment-plan payments and refunds are excluded. Current MAP Keys definition; award-period definition version not separately captured.
Source: HFMA 2024 MAP Award Statistical Data, selected hospital/system/IDS winners (n=14)
Compare provider collections with FTE, hours and specialty before changing staffing.
Total Medical Revenue per Physician FTE
Source / sample: Previously attributed MGMA 2023 context; Data Dive versus Better Performers report identity is unverified, no certified population
The retained MGMA-attributed provider-revenue references are not independently reverified in this review. A group difference does not prove that adding support staff causes a particular revenue gain. Reconcile collections, provider FTE, clinical hours, payer mix and coding coverage, then test a bounded workflow change.
Source basis
Definition: Same-period eligible medical revenue divided by eligible physician FTE. Revenue, collections, provider FTE and clinical hours differ; source-specific MGMA population and treatment of ancillary revenue are unverified here.
Source: Previously attributed source context (exact values not reverified): MGMA Better Performers Report 2023
Define owner compensation and specialty before comparing overhead ratios.
Primary Care Overhead Ratio
Source / sample: Level planning overhead references, not a measured sample. AMA practice-expense share is a different measure and does not establish this distribution.
The displayed overhead values are retained planning references, not an AMA percentile distribution. Check whether owner compensation, rent, billing and technology costs sit in the numerator. A lower ratio can reflect different service mix or accounting treatment rather than an operating improvement.
Source basis
Definition: Defined practice operating-overhead costs divided by matching revenue. Specify clinical/nonclinical staff, owner compensation and cost classification before comparing planning cases.
Source: Level planning overhead cases; AMA PPI 2024 practice-expense share does not supply these values
Denial definitions and payer populations must match before comparing rates.
Claim Denial Rate (% of claims)
Source / sample: Retained context: Different Optum hospital and KFF ACA populations plus targets; not a common practice-percentile distribution. Exact report population/positions are not certified unless a bar is explicitly labeled published below.
The chart combines previously attributed Optum hospital and KFF Marketplace references with Level planning inputs. These are different claim populations, not one practice distribution. As an illustrative cost model only, 12,000 claims x 12% denials x $25-$118 rework per denial gives $36,000-$169,920. Verify your actual denial denominator, recoveries and rework time; the cost inputs are not established universal rates. The current corrected KFF 2023 analysis (updated March 24 2026) reports 20% final in-network and 36% out-of-network denial, excluding initial denials later resubmitted and paid. The retained 19/37 bars are historical inputs with unverified edition, not first-pass practice rates.
Source basis
Definition: Denied claims divided by eligible submitted or adjudicated claims on the chosen claim-count basis. Specify initial versus final denial and line-versus-claim counting. Optum/KFF attributions use different populations; planning targets are separate.
Source: Previously attributed Optum hospital context and retained KFF Marketplace inputs with edition unverified; separate planning targets
Aged billed receivables require payer-level review before assigning a loss.
A/R Over 90 Days (% of billed receivables, selected hospital/system winners)
Source / sample: 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners
The retained 22%, 30% and 38% values round the source performance P75/P50/P25 aged-A/R positions (22.4%, 29.8%, 38.2%) for 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners. This is a billed-receivables share, not a share of annual revenue or automatic write-off probability. The separate three physician-practice winners are a different population. Each measure reports its percentile positions separately; a winner need not occupy the same position across measures. The data does not establish a matched relationship or financing consequence.
Source basis
Definition: Aged billed A/R / total billed A/R, using a consistent snapshot and eligible debit-balance accounts. Unbilled accounts and credit balances require separate treatment; source references are selected hospital/system winner positions. Current MAP Keys definition; award-period definition version not separately captured.
Source: HFMA 2024 MAP Award Statistical Data, selected hospital/system/IDS winners (n=14)
Model and segment planning references
HealthcareIndependent retained inputs for operating scenarios. They are not measured common-cohort medians or expected outcomes.
Primary Care
Previously attributed reference: Revenue per physician
$668K (lower attributed reference) / $770K (higher attributed reference)
Previously attributed MGMA reference; exact report identity, value and population not reverified
Not a measured Level sample
Surgical Specialty
Previously attributed reference: Revenue per physician
$648K (lower attributed reference) / $777K (higher attributed reference)
Previously attributed MGMA reference; exact report identity, value and population not reverified
Not a measured Level sample
Multispecialty Group
Previously attributed reference: Revenue per physician
$748K (lower attributed reference) / $879K (higher attributed reference)
Previously attributed MGMA reference; exact report identity, value and population not reverified
Not a measured Level sample
Dental, General
Planning reference: Reference gross billings/dentist
$942K
Retained planning input; exact source population unverified
Not a measured Level sample
Dental, Specialty
Planning reference: Reference gross billings/dentist
$1,146K
Retained planning input; exact source population unverified
Not a measured Level sample
Physical Therapy
Planning reference: Reference revenue per visit
$98-$101
Retained planning input; exact source population unverified
Not a measured Level sample
Advanced Healthcare Metrics
Sub-segment breakdowns, advanced operational metrics, and clearly labeled reference ranges for medical, dental, pt & specialty practices.
Reference A, center and reference C retain the existing comparison positions. They are not common measured percentiles, ascending numerical bins or proven quality ranks. Each source observation and planning input retains its own definition and population.
| Metric | Reference A | Center reference | Reference C | Note |
|---|---|---|---|---|
| Dental Net Margin | < 20% | 25-30% (before owner comp) | 30-40% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning margin bands, not verified ADA percentiles. The center explicitly precedes owner compensation; calculate the owner salary separately and use the same gross-billings or collected-revenue denominator throughout. |
| Primary Care Net Margin | < 8% | 12-15% | > 18% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning net-margin bands, not measured primary-care quartiles. Define owner compensation, clinical costs and any employed-provider subsidy consistently before comparing. |
| Specialty Clinic Operating Margin | < 15% | 18-24% | > 28% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning operating-margin bands, not verified specialty percentiles. Operating margin and net margin differ; reconcile procedure mix, owner compensation, interest and taxes. |
| Net Collection Rate | < 95% | ~96% | 99%+ | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning collection bands. Define the same eligible adjusted-charge denominator and match payment timing; these values are not certified MGMA or HFMA practice percentiles. |
| Clean Claim Rate (first-pass) | < 90% | ~95% | 98%+ | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning first-pass claim bands. Count eligible submitted claims and the first-pass acceptance or payment event consistently; a named KPI definition does not certify these numerical targets. |
| Days in A/R | > 50 days | < 40 days | < 30 days | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Retained planning bands. Separate primary context: 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners had net A/R-day positions 30.5-45.3; the displayed bands are not those measured percentiles. |
| Cost to Collect | > 4% | 2-4% | < 2% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning bands with no retained measured distribution. Current HFMA MAP Keys FM-6 divides total revenue-cycle cost by total patient-service cash collected, net of refunds, with specified inclusions and exclusions. HFMA states the definition changed in August 2025; do not treat pre-August-2025 values or percentages on other denominators as comparable. Verify the applicable definition before using these inputs. Open source |
| Subsidy per Employed Physician | > $300K | ~$307K | profitable | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Previously attributed Kaufman Hall Q4 2024 source reference; exact capture not retained here. The >$300K band overlaps ~$307K deliberately as independent sensitivity inputs, not ordered percentile bins. |
| Patient Collection Rate | < 34% | 40-48% | > 60% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Unverified retained planning recovery inputs; denominator and timing must be defined. This is not HFMA PA-7, whose denominator is total self-pay cash collected. |
| Bad Debt (% of gross patient service revenue) | > 2.2% | 1.5% | < 0.6% | Evidence: Source-attributed reference plus planning bands; not common measured quartiles Sample: 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners The 1.5% center is the captured P50 bad-debt statistic for 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners. The 0.6% and 2.2% references are source P90/P10 positions, not quartiles; the inequality bands are planning comparisons. Current HFMA AR-7 uses the income-statement bad-debt provision divided by gross patient-service revenue, not amounts written off from A/R. Match definition versions before comparison; the capture does not establish causation with aged A/R. Current definitions: https://www.hfma.org/data-and-insights/map-initiative/map-keys/ Open source |
| A/R Over 90 Days (% of billed receivables) | > 38% | 30% | < 22% | Evidence: Source-attributed reference plus planning bands; not common measured quartiles Sample: 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners Rounded 22.4/29.8/38.2% source performance positions for 14 selected 2024 HFMA MAP hospital, critical-access hospital, hospital-system and integrated-delivery-system award winners. These reference bands are not private-practice quartiles or annual-revenue percentages. Current definitions: https://www.hfma.org/data-and-insights/map-initiative/map-keys/ Open source |
| Cost per Claim Processed | > $12 | $8-$10 | < $6 | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent fully loaded cost-per-claim planning bands. Measure submission, rework and overhead separately. A transaction-only cost study does not certify the cost of the whole claim lifecycle. |
| No-Show Rate | > 10% | ~7% | < 5% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent planning no-show bands, not a verified practice average or distribution. Define eligible scheduled appointments, cancellations and the observation period. |
| Staff Cost % of Revenue | > 35% | 28-32% | < 25% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent non-provider staff-cost planning bands. Specify administrative, billing and support FTE and use the same revenue denominator; no measured Level practice cohort is supplied. |
| Technology Spend % of Revenue | > 5% | 3-4% | < 2.5% | Evidence: Retained planning or previously attributed reference; exact measurement unverified Sample: No measured common population; planning references differ Independent technology-cost planning bands. Include the same EMR, practice-management, billing and telehealth costs and revenue period; no measured Level practice cohort is supplied. |
Benchmarks for other service businesses
Simple pricing
Three tiers, one ladder.
$500+/mo
Bookkeeping
The clean data layer: monthly books, reconciliations, and organized financials AI can work with.
$1,500-$5,000/mo
Scale
The full AI operating layer: custom agents, weekly actions, and benchmarks to grow margin per hour.
Custom
Platform / Multi-Office
Multi-branch benchmarking and scorecards for PE-backed and multi-location groups.
How does your practice compare?
Compare your own source records and accounting definitions with the selected references, then test a bounded operating decision.
No commitment. Real numbers, not generic advice.
Frequently Asked Questions
What is a good net collection rate for a medical practice?
The 95-99% bands here are retained planning references, not verified practice quartiles. Define eligible charges minus contractual adjustments before comparing payments. A hypothetical three-point difference on a $2M adjusted-charge denominator is $60,000; raw charges are a different denominator.
How many days in A/R is acceptable?
The captured HFMA 2024 group contains 14 selected hospital, critical-access hospital, hospital-system and IDS award winners, not a national practice cohort. Its net A/R-day reference P90/P10 positions are 30.5/45.3 days. The displayed 28-55-day chart is a separate Level planning model. The three physician-practice winners are separate; match payer mix, eligible balances and cutoff before comparing.
What should my practice overhead ratio be?
Use the retained overhead bands as planning inputs, not measured specialty quartiles. Define owner compensation, clinical and administrative costs, net revenue and time period. Review contracts and utilization before claiming a saving.
How much does a claim denial actually cost?
Measure actual rework time, loaded labor and recovered payment by denial type. In an illustration, 10,000 claims at an assumed 12-15% denial rate and $25-$118 cost per denial produce $30,000-$177,000 modeled rework cost. Those assumptions are not a verified universal practice rate.
What revenue should each provider generate?
The provider-revenue ranges are retained source-attributed references with unverified exact report values. Compare specialty, FTE, collections, clinical hours and support costs. Revenue divided by compensation alone does not establish profitability.
Sources
- • MGMA DataDive (medical group medians) (only values with an exact captured source are verified; other attribution is historical context)
- • HFMA 2024 MAP Award statistics (14 selected hospital/system/IDS winners, 3 separate physician-practice winners); current MAP Keys definitions, with FM-6 updated August 2025 (only values with an exact captured source are verified; other attribution is historical context)
- • ADA Health Policy Institute Survey of Dental Practice 2024 (only values with an exact captured source are verified; other attribution is historical context)
- • AMA Physician Practice Index 2024 (only values with an exact captured source are verified; other attribution is historical context)
- • BLS Occupational Employment & Wages 2023-2024 (only values with an exact captured source are verified; other attribution is historical context) (background only; no displayed value exactly verified here)
- • Retained KFF Marketplace 19/37 inputs, exact edition unverified; separate corrected 2023 context is cited in the owner decision.
- • Level operating guidance; no measured Level healthcare-practice cohort is supplied. The contractor research universe is separate and does not certify these practice statistics.
- • KFF 2023 Marketplace claims analysis, corrected March 24, 2026: 20% final in-network / 36% out-of-network; separate from retained 19/37 chart inputs.
The Level Index combines named third-party observations with separately labeled Level planning guidance. Each published figure retains its own source population, period and definition. Planning ranges are illustrations, not a measured Level industry cohort. Individual comparisons depend on business model, accounting definitions and source coverage. © 2026 Level. All rights reserved.
Read the source basis and limits beside each reference. Definitions and sample sizes are shown when captured from the reviewed source. Download this dataset (JSON, free to cite).