How the Health Score Works
We score every ZIP code in America from 0 to 100 based on how easy it is to access healthcare. Here's exactly how -- no PhD required.
Every ZIP tells a story
The Short Version
We ask one question:"How good is healthcare access where you live?"
To answer it, we compute an access index for 7 healthcare categories -- primary care, emergency rooms, hospitals, specialists, mental health, preventive screening, and dental -- using a spatial model that accounts for both how many providers exist and how many people are competing for them. Telehealth readiness adjusts the provider scores without inflating them.
We then apply a social determinants penalty that reduces the score based on real-world barriers: lack of insurance, poverty, poor transportation, low health literacy, disability rates, and age vulnerability. This penalty can cut a score by up to 65% -- but good social factors can never inflate a score beyond what the physical infrastructure supports.
The result is a score from 0 (healthcare desert) to 100 (healthcare paradise) for all 33,000 populated ZIP codes, computed from 21 free government data sources updated automatically.
The Big Idea
Most people think healthcare access means "Is there a hospital nearby?" But that's only part of the story. True access depends on three things:
Supply
How many doctors, hospitals, and facilities are near you?
Demand
How many other people are also trying to see those providers?
Barriers
Can residents actually GET to those providers? (Transport, cost, insurance)
The Pizza Shop Analogy
Imagine a pizza shop 10 minutes from your house. Sounds great, right?
But what if 100,000 other peoplealso live 10 minutes away, there's only one oven, you don't have a car, and you can't afford the delivery fee? That "accessible" pizza shop isn't so accessible anymore.
Our Health Score captures all of that -- supply, demand, and barriers -- for healthcare instead of pizza. It's the difference between "there's a doctor nearby" and "you can actually see a doctor."
The Formula
v3.1 · 2026-03-26Here's the actual equation. Every step after this section explains a piece of it.
# Master equation
HealthScore(z) = BaseAccess(z)
× SDOH_Penalty(z)
× DataQuality(z)
× AfterHours(z)
BaseAccess = weighted sum of 7 E2SFCA component scores [0-100]
SDOH_Penalty = social determinants multiplier [0.35 - 1.05]
DataQuality= completeness × freshness confidence [0.80 - 1.0]
AfterHours = urgent care availability bonus [1.00 - 1.05]
The score is multiplicative by design. Good social factors remove barriers but can't create providers. Bad data quality reduces confidence. After-hours urgent care is a small bonus.
The Score at a Glance
From raw government data to your final score. Every step is automated, transparent, and reproducible.
Excellent Access
BaseAccess × SDOH × DataQuality × AfterHours
r > 0.65
CDC PLACES
r > 0.70
County Health Rankings
> 70%
HRSA HPSA match
Simplified View
The 8 Healthcare Components
The Base Access score is built from eight components: seven directly weighted (primary care, emergency, hospital inpatient, specialist, mental health, preventive, dental) plus telehealth, which feeds back as a distance-penalty reducer rather than as a standalone weight. Each component uses the E2SFCA method (explained below) to measure supply vs. demand for that type of care.
Provider Access
55% of Base Access score
Primary CareC1~25%
Family medicine, internal medicine, pediatrics, nurse practitioners, and physician assistants near you — weighted by how many patients compete for their time.
Decay function: Gaussian· Catchment radius varies by urban/rural tier (RUCA codes)
Specialist AccessC4~15%
Cardiologists, oncologists, surgeons, and other specialists. Availability correlates strongly with cancer survival and cardiovascular outcomes.
Decay function: Logistic· Catchment radius varies by urban/rural tier (RUCA codes)
Mental HealthC5~10%
Psychiatrists, psychologists, counselors, and behavioral health providers. 55% of US counties have zero psychiatrists.
Decay function: Gaussian· Catchment radius varies by urban/rural tier (RUCA codes)
DentalC7~5%
Dentists, oral surgeons, and dental hygienists. 74 million Americans live in dental Health Professional Shortage Areas.
Decay function: Gaussian· Catchment radius varies by urban/rural tier (RUCA codes)
Hospital & Facility Access
45% of Base Access score
Emergency & TraumaC2~20%
Proximity to emergency departments and trauma centers, weighted by trauma level designation. Mortality increases 1% per 10-mile increase in trauma distance.
Decay function: Sigmoid· Catchment radius varies by urban/rural tier (RUCA codes)
Hospital InpatientC3~15%
Bed capacity, surgical services, and ICU access. Hospital closures increase heart attack mortality by nearly 6%.
Decay function: Gaussian· Catchment radius varies by urban/rural tier (RUCA codes)
Preventive & ScreeningC6~10%
Mammography, colonoscopy, immunization sites, and screening centers. Early detection directly reduces mortality across multiple conditions.
Decay function: Threshold· Catchment radius varies by urban/rural tier (RUCA codes)
Telehealth doesn't get its own weight. Instead, it adjusts the Provider scores (C1, C4, C5, C7) by reducing the distance penalty for providers who offer telehealth. This prevents telehealth from inflating scores in areas with no physical infrastructure — a ZIP with only telehealth and zero in-person providers won't score well.
Filtered by the Interstate Medical Licensure Compact (IMLC). Providers in non-member states (notably California and New York) receive zero telehealth credit from out-of-state providers.
Component weights (sum to 100%)
Weights are constrained by clinical evidence and will be refined by elastic net regression against CDC PLACES utilization data as calibration data becomes available.
How the Math Works
We use a method called E2SFCA(Enhanced Two-Step Floating Catchment Area). Here's the simple version:
Count the providers around each ZIP code
For each of the 7 healthcare categories, we draw a circle around every ZIP code and count the providers inside it. But we don't treat them all equally:
- Distance decay: Closer providers count more. Each component uses a different decay curve -- emergency care drops off sharply (sigmoid), dental is more gradual (Gaussian).
- RUCA tiers: Circle size changes by urbanicity -- rural circles are bigger because people drive farther. 5 tiers from Urban Core (15 mi) to Isolated Rural (50 mi).
- Supply weights: A family medicine doctor counts as 1.0, a nurse practitioner as 0.85, a pediatrician as 0.50 (serves only children).
Factor in the competition
Here's the key insight: a doctor who serves 50,000 patients is less accessible than one who serves 5,000. So we divide each provider's capacity by the total population competing for their time. This gives us a provider-to-population ratio that reflects real-world availability, not just presence on a map.
Normalize and weight
Each component score is normalized to 0-100 using log-softcap normalization at the 95th percentile -- this preserves the full ranking while keeping the top 5% from bunching up at 100. Then we multiply each by its calibrated weight (Primary Care ~25%, Emergency ~20%, etc.) and add them up to get the Base Access Score.
Multiply by the SDOH penalty
Having a hospital nearby doesn't help if you can't afford the copay or don't have a car. The SDOH penalty is a multiplier-- it takes 6 social barrier indices, compounds them together (because disadvantages multiply, not add), and produces a factor between 0.35 and 1.05. At 0.35, an area loses 65% of its Base Access score. At 1.05, minimal barriers give a tiny 5% boost. You can't create doctors that don't exist.
Apply quality and after-hours adjustments
Two final multipliers: Data Quality(0.80-1.0) reduces confidence when underlying data is incomplete or stale -- if provider data hasn't been refreshed on schedule, the score drops. After-Hours (1.00-1.05) gives a small bonus to ZIPs with nearby urgent care facilities, recognizing that healthcare access extends beyond business hours.
Additional refinements (v3.1)
- Population-weighted centroids: ZIP locations use block-group population data, not geographic centers. Corrects 3-7 point errors for large rural ZIPs.
- Multi-site provider splitting: Providers at multiple locations are fractionally allocated — 50% to primary site, remainder split among secondary sites.
- Cross-state border penalty: Cross-state providers in the same metro area get a 15% penalty; non-metro cross-state providers get 40%, reflecting that people rarely cross state lines for care outside metro areas.
- Boundary ZIP correction: ZIPs at coastlines, state borders, or national boundaries have artificially truncated catchment areas. A geometric correction factor (capped at 1.50) accounts for this.
- State-aware telehealth: Telehealth providers are filtered by Interstate Medical Licensure Compact (IMLC) membership. Non-member state pairs receive zero telehealth credit from out-of-state providers.
Try It Yourself
InteractiveAdjust component weights and pick different locations to see how geography and social barriers interact. The SDOH penalty is applied as a multiplier -- watch how it compounds.
Interactive Score Simulator
Adjust component weights. Watch the formula in action.
Choose a location
Component Scores for Suburban Atlanta
Adjust component weights
How the score is calculated
Base Access
Final Score
Excellent Access
Suburban Atlanta
The SDOH Penalty: 6 Social Barriers
Healthcare access isn't just about geography. These 6 social factors are multiplied together -- because disadvantages compound. A ZIP with 40% uninsured AND 30% poverty gets hit harder than the sum of those penalties individually.
What percentage of residents have health insurance? The strongest single predictor of utilization — uninsured adults are 50% less likely to have a regular doctor.
Percentage of households with zero vehicles, adjusted for public transit availability. Transportation barriers cause 3-67% of missed appointments across studies.
Poverty rate and median household income. Even with insurance, copays and deductibles can be barriers. Effect partially captured by insurance and transportation indices.
Percentage with any disability in the civilian noninstitutionalized population. Disabled residents face additional physical and logistical barriers to reaching care.
Percentage without a high school diploma and limited English proficiency. Affects whether people can navigate the healthcare system and understand health information.
Percentage age 65+ and under 5, discounted by coverage rates. Areas with more dependent populations have higher healthcare demand, stretching resources thinner.
Why multiplicative?The arithmetic mean lets a high score in one area offset a catastrophic score in another. If a ZIP has excellent transportation but 40% uninsured, an arithmetic mean says "not that bad." The multiplicative model correctly reflects that insurance is a hard barrier regardless of transportation -- you either have coverage or you don't. Each gamma (elasticity) controls how aggressively that factor penalizes the score, set from peer-reviewed effect sizes in the health disparities literature.
# SDOH Penalty formula
SDOH(z) = 0.35 + 0.70 × ∏(1 - Vj)γj
where Vj = vulnerability percentile [0,1] for each sub-index
output range: [0.35, 1.05]
Range: 0.35 to 1.05.A wealthy area with great insurance and cars can't score 90 if there are only 2 doctors for 100,000 people. The maximum 5% bonus for minimal barriers is intentionally tiny. The design is asymmetric: disadvantage can destroy access (up to -65%), but advantage barely inflates it (+5% max).
11 Scores per ZIP Code
Every ZIP code gets an Overall Health score plus 10 condition-specific scores. Each condition score uses the same E2SFCA methodology but filters for providers and facilities relevant to that condition.
Overall Health
Alzheimer's
Arthritis
Asthma
Breast Cancer
COPD
Hypertension
Lung Cancer
Prostate Cancer
Type 1 Diabetes
Type 2 Diabetes
For example, the Breast Cancer score weights mammography facilities, oncologists, and breast cancer treatment centers more heavily than general primary care. Explore condition scores
What the Scores Mean
Every score comes with a human-readable label so you instantly know what it means:
Multiple providers in every category within short distance. Top-tier hospitals. Full specialist coverage.
Most healthcare needs met locally. Some specialists may require a short drive.
Basic care available but gaps exist. Some services require significant travel.
Significant gaps in healthcare availability. Many services require long drives.
Severe shortage of healthcare services. Residents face major barriers to basic care.
Where the Data Comes From
Every data point comes from free, public government sources. No paid data, no private databases, no guessing. All 21 sources are updated automatically.
| Source | What It Provides | Updates |
|---|---|---|
| CMS NPPES | 4 million+ doctor profiles | Weekly |
| CMS Hospital Compare | Hospital quality ratings | Monthly |
| CMS Care Compare | Nursing homes, dialysis centers | Quarterly |
| SAMHSA Locator | Mental health & substance abuse facilities | Quarterly |
| Census ACS 5-Year | Demographics, income, insurance, disability | Annually |
| Census TIGER | ZIP code map boundaries | Annually |
| CDC PLACES | Community health measures (validation) | Annually |
| HRSA HPSA | Official shortage area designations | Quarterly |
| USDA RUCA Codes | Urban vs rural classification | On change |
| FCC Broadband | Internet availability for telehealth | Annually |
| FDA MQSA | Certified mammography facilities | Quarterly |
| CMS POS | Hospital master list with details | Quarterly |
Plus 9 additional sources for validation, calibration, and geographic reference data.See full registry
How We Know It's Accurate
A score is only useful if it reflects reality. We validate our scores against three independent benchmarks:
r > 0.65
CDC PLACES
Correlation with community health utilization measures (checkups, screenings, ER visits)
r > 0.70
County Health Rankings
Correlation with the Robert Wood Johnson Foundation county health data
> 70%
HRSA HPSA
Concordance with federally designated Health Professional Shortage Areas
Translation: when we say an area is a "Healthcare Desert," at least 70% of the time the federal government has also designated it as a shortage area. And ZIP codes we score highly tend to have measurably better health outcomes in the CDC data.
Academic Foundation
Our methodology is built on peer-reviewed research. The E2SFCA method has been cited over 2,000 times in academic literature since its publication.
Core method:Luo, W., & Qi, Y. (2009). "An enhanced two-step floating catchment area (E2SFCA) method for measuring spatial accessibility to primary care physicians." Health & Place, 15(4), 1100-1107.
SDOH framework:Syed, S. T., Gerber, B. S., & Sharp, L. K. (2013). "Traveling towards disease: transportation barriers to health care access." Journal of Community Health, 38(5), 976-993.
Rural adaptations:McGrail, M. R. (2012). "Spatial accessibility of primary health care utilising the two step floating catchment area method." Applied Geography, 32(2), 309-320.
SDOH compounding:Marmot, M. (2005). "Social determinants of health inequalities." The Lancet, 365(9464), 1099-1104.
Weight calibration:Zou, H., & Hastie, T. (2005). "Regularization and variable selection via the elastic net." Journal of the Royal Statistical Society: Series B, 67(2), 301-320.
Our full formula specification cites 36 academic papers. The complete bibliography is available in our technical documentation.
See your ZIP code's score
Now that you know how the score works, find out what healthcare access looks like where you live.
Search your ZIP code