{"main_topic":"health","sub_topic":"cms","topic":"medicare_service_utilization","key":"health/cms/medicare_service_utilization","version":"1.0.0","year_min":2014,"year_max":2024,"year_gaps":[],"update_frequency":"annual","detail_levels":["counties","states"],"default_detail":"counties","path_template":"health/medicare_service_utilization/year={year}/{detail}.parquet","grain":["year","county_fips","tract_geoid","age_bracket","service","volume_unit"],"title":"Medicare Use of Services per 1,000 Beneficiaries (CMS)","summary":"Hospital stays, visits, tests and other services per 1,000 Original Medicare beneficiaries in Georgia and its counties.","description":"How much care Original Medicare (fee-for-service) beneficiaries in Georgia and each of its 159 counties use, by service, per 1,000 beneficiaries a year: hospital inpatient stays and days, hospital outpatient visits, skilled nursing, inpatient rehabilitation, long-term care hospital and hospice stays and days, home health episodes and visits, evaluation and management visits, procedures, tests, imaging, durable medical equipment, ambulance and treatment events, and outpatient dialysis and FQHC/rural health clinic visits. Statewide, inpatient hospital stays were 276 per 1,000 in 2014 and 224 in 2024. 2014-2024, from CMS's Medicare Geographic Variation public use file. Medicare Advantage members are not counted (60% of Georgia's beneficiaries with Part A and B in 2024).","explorer":true,"search_aliases":{"en":["medicare utilization","medicare hospital stays","doctor visits","home health visits","hospice use","skilled nursing stays","imaging and tests","dialysis visits","ambulance use","medical care use"],"es":["uso de servicios medicare","estadías en hospitales","visitas al médico","atención domiciliaria","hospicio","enfermería especializada","imágenes y pruebas","diálisis","ambulancias","cuidado médico"]},"source":"Centers for Medicare & Medicaid Services (CMS), Medicare Geographic Variation - by National, State & County, data.cms.gov dataset 6219697b-8f6c-4164-bed4-cd9317c58ebc, 2014-2024 release (2014-2024 Original Medicare Geographic Variation Public Use File.csv)","source_url":"https://data.cms.gov/summary-statistics-on-use-and-payments/medicare-geographic-comparisons/medicare-geographic-variation-by-national-state-county","endpoint":"/api/v1/health/medicare_service_utilization","columns":[{"name":"year","type":"integer","role":"year","nullable":false,"label":"Year","short_description":"","description":"Calendar year of service (claims by date of service, January to December), 2014-2024 (CMS's 2014-2024 release).","valid_values":null,"example":2024,"unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":false,"metric_component":null},{"name":"county_fips","type":"string","role":"fk_county","nullable":true,"label":"County FIPS","short_description":"","description":"5-digit county FIPS code of the beneficiaries' residence (CMS BENE_GEO_CD); NULL on the Georgia state row. CMS's GA-UNKNOWN (13000) is left out.","valid_values":null,"example":"13121","unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":false,"metric_component":null},{"name":"tract_geoid","type":"string","role":"fk_tract","nullable":true,"label":"Tract GEOID","short_description":"","description":"Always NULL: CMS publishes state and county figures only. Kept for the health domain's standard key shape.","valid_values":null,"example":null,"unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":false,"metric_component":null},{"name":"age_bracket","type":"string","role":"categorical","nullable":false,"label":"Age Bracket","short_description":"'all' = every beneficiary; 'under_65' / '65_plus' (Georgia only).","description":"Beneficiary age group (CMS BENE_AGE_LVL): 'all' (every age), 'under_65' (entitled by disability or end-stage renal disease) and '65_plus'. CMS publishes the two brackets for Georgia only; county rows are 'all'. The brackets partition 'all': their beneficiaries and each service's volume (rate x beneficiaries / 1,000) add up to it exactly, and its rate lies between theirs. Pin 'all' to compare places.","valid_values":["all","under_65","65_plus"],"example":"all","unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":true,"metric_component":null},{"name":"service","type":"string","role":"categorical","nullable":false,"label":"Service","short_description":"Type of Medicare-covered service.","description":"Service category (CMS's 17 utilization categories, the same values as the spending topic), with the volume unit(s) it is counted in. 'inpatient' (stays and days) = hospital inpatient stays: acute care hospitals paid under the IPPS, critical access hospitals, inpatient psychiatric facilities and other inpatient Part A services. 'outpatient' (visits) = hospital outpatient department services (OPPS hospitals and critical access hospital outpatient departments); outpatient dialysis facilities are their own category. 'ambulatory_surgery_center' (events) = ambulatory surgery center services. 'skilled_nursing_facility' (stays and days) = post-acute skilled nursing facility stays. 'inpatient_rehabilitation_facility' (stays and days) = inpatient rehabilitation facility stays. 'long_term_care_hospital' (stays and days) = long-term care hospital stays. 'home_health' (episodes and visits) = home health agency services. 'hospice' (stays and days) = hospice care. 'evaluation_and_management' (events) = evaluation and management visits by physicians and other practitioners (office, hospital, nursing facility, emergency department and other E&M services; RBCS). 'procedures' (events) = procedures billed by physicians and other practitioners, anesthesia included (RBCS). 'tests' (events) = laboratory and other tests (RBCS). 'imaging' (events) = imaging: CT, MRI, nuclear medicine, ultrasound, X-ray (RBCS). 'durable_medical_equipment' (events) = durable medical equipment, supplies, oxygen, orthotics and drugs given through DME (RBCS). 'outpatient_dialysis_facility' (visits) = outpatient dialysis facility services. 'fqhc_and_rural_health_clinic' (visits) = federally qualified health center and rural health clinic visits. 'ambulance' (events) = ambulance services. 'treatments' (events) = Part B treatments: chemotherapy, other injections and infusions, radiation oncology, physician dialysis services, and physical, occupational and speech therapy (RBCS). Each service has only its own unit(s) (volume_unit). No total: services never add up.","valid_values":["inpatient","outpatient","ambulatory_surgery_center","skilled_nursing_facility","inpatient_rehabilitation_facility","long_term_care_hospital","home_health","hospice","evaluation_and_management","procedures","tests","imaging","durable_medical_equipment","outpatient_dialysis_facility","fqhc_and_rural_health_clinic","ambulance","treatments"],"example":"inpatient","unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":true,"metric_component":null},{"name":"volume_unit","type":"string","role":"categorical","nullable":false,"label":"Volume Unit","short_description":"What the figure counts: stays, days, visits, episodes or events.","description":"The unit the service's use is counted in, fixed by the service: 'stays' = covered stays: continuous, uninterrupted periods of care in one setting (Part A); 'days' = Medicare-covered days of care in those stays (Part A); 'visits' = visits (hospital outpatient department, home health, outpatient dialysis facility and FQHC/RHC visits); 'episodes' = home health episodes (periods of care): 60-day episodes through 2019, 30-day periods from 2020 (Medicare's Patient-Driven Groupings Model), so episodes before and after 2020 are not comparable (about 1.6 times as many from 2020 for the same care); use visits for home health trends; 'events' = events: each type of Part B service (an RBCS subcategory) a beneficiary received on one claim on one day. By service: stays and days for inpatient, skilled_nursing_facility, inpatient_rehabilitation_facility, long_term_care_hospital, hospice; visits for outpatient, outpatient_dialysis_facility, fqhc_and_rural_health_clinic; events for ambulatory_surgery_center, evaluation_and_management, procedures, tests, imaging, durable_medical_equipment, ambulance, treatments; episodes and visits for home_health. No total: never add across units (stays and days are two measures of the same care).","valid_values":["stays","days","visits","episodes","events"],"example":"stays","unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":true,"metric_component":null},{"name":"volume_per_1000_beneficiaries","type":"number","role":"metric","nullable":true,"label":"Use per 1,000 Original Medicare Beneficiaries","short_description":"Stays, days, visits, episodes or events (volume_unit) per 1,000 Original Medicare beneficiaries in the year.","description":"How much of the service the place's Original Medicare (fee-for-service) beneficiaries used in the year, per 1,000 of them, counted in volume_unit: the number of stays, days, visits, episodes or events divided by num_om_beneficiaries, times 1,000 (CMS's <service>_..._PER_1000_BENES, as published, 4 decimals). Over every beneficiary, not only those who used the service. Not per 1,000 Medicare enrollees: Medicare Advantage members are not counted. Compare only within one service and volume_unit. A 0 is real (nobody used the service). For several counties, pool as SUM(volume_per_1000_beneficiaries * num_om_beneficiaries) / SUM(num_om_beneficiaries).","valid_values":null,"example":224.0577,"unit":"rate_per_population","value_min":null,"value_max":null,"null_meaning":"Suppressed by CMS ('*'): fewer than 11 of the place's Original Medicare beneficiaries used the service that year (CMS then suppresses all of the service's figures): 3,830 county cells in this release, mostly long_term_care_hospital (2,334), inpatient_rehabilitation_facility (1,048), outpatient_dialysis_facility (202); never Georgia's row. CMS's 'NA' (not produced) never occurs in this column.","key_metric_grain_contributor":false,"metric_component":null},{"name":"num_om_beneficiaries","type":"integer","role":"metric","nullable":true,"label":"Original Medicare Beneficiaries","short_description":"","description":"Original Medicare (fee-for-service) beneficiaries living in the place: Part A, Part B and Original Medicare for every month they were entitled, new enrollees and people who died during the year included (CMS BENES_OM_CNT). Not every Medicare beneficiary: Medicare Advantage enrollees and people with Part A or B alone are left out. The denominator of volume_per_1000_beneficiaries (CMS publishes no numerator count). It belongs to the place, year and age bracket and repeats on every service and volume_unit row: never sum it across them.","valid_values":null,"example":716088,"unit":"count","value_min":null,"value_max":null,"null_meaning":"Suppressed by CMS (fewer than 11 beneficiaries). Not seen on any state or county row in the current release.","key_metric_grain_contributor":false,"metric_component":"denominator"}],"key_metric":"volume_per_1000_beneficiaries","numeric_filter_columns":["volume_per_1000_beneficiaries","num_om_beneficiaries"],"usage":"Star-schema fact table at state and county grain: state rows have county_fips NULL, county rows join the counties dimension on county_fips; tract_geoid is always NULL. Pin age_bracket = 'all' to compare places (county rows are 'all' only; 'under_65' and '65_plus' are Georgia's own rows, and they add up to 'all'). Pick one service and one volume_unit: each service publishes only its own unit(s) (stays and days for inpatient, skilled_nursing_facility, inpatient_rehabilitation_facility, long_term_care_hospital, hospice; visits for outpatient, outpatient_dialysis_facility, fqhc_and_rural_health_clinic; events for ambulatory_surgery_center, evaluation_and_management, procedures, tests, imaging, durable_medical_equipment, ambulance, treatments; episodes and visits for home_health). Compare places or years within one service and unit; for home health trends across 2020 use visits (an episode is a 60-day episode through 2019 and a 30-day period from 2020). For several counties, pool as SUM(volume_per_1000_beneficiaries * num_om_beneficiaries) / SUM(num_om_beneficiaries) within one year, service and volume_unit over the counties with a published rate (a beneficiary-weighted average; CMS's rate is uses / beneficiaries x 1,000), never a plain average of county rates; for Georgia use the state row. num_om_beneficiaries repeats on every service row of a place, year and age bracket: never sum it across services or units. Read directly with DuckDB over Parquet.","limitations":"Rates are per 1,000 Original Medicare (fee-for-service) beneficiaries with Part A and B in every month they were entitled (num_om_beneficiaries), not per 1,000 Medicare enrollees. In 2024 those beneficiaries were 35% of Georgia's Medicare beneficiaries and 20-49% of each county's (Medicare Advantage up to 76%, Clayton County), so county figures can describe a fee-for-service minority. Each service counts in its own unit (volume_unit), so never add figures across services or units. Original Medicare (fee-for-service) only: CMS counts beneficiaries with Part A, Part B and Original Medicare for every month they were entitled (num_om_beneficiaries); anyone in Medicare Advantage at any point in the year, or with Part A or Part B alone, is left out. Units: stays and days for inpatient, skilled_nursing_facility, inpatient_rehabilitation_facility, long_term_care_hospital, hospice; visits for outpatient, outpatient_dialysis_facility, fqhc_and_rural_health_clinic; events for ambulatory_surgery_center, evaluation_and_management, procedures, tests, imaging, durable_medical_equipment, ambulance, treatments; episodes and visits for home_health. Stays and days of one service (and home health episodes and visits) are two measures of the same care; one visit can also count under several services (the facility's outpatient visit and the doctor's E&M, procedure or imaging events). Home health episodes are 60-day episodes through 2019 and 30-day periods from 2020 (Medicare's Patient-Driven Groupings Model), so episodes per 1,000 jump about 60% in 2020 with no change in use: for home health trends across 2020, use visits. Georgia's outpatient dialysis visits per 1,000 Original Medicare beneficiaries roughly halved after 2020 (2,660 in 2020, 1,223 in 2024) while visits per dialysis user held steady (139 and 130 a year): that fits people on dialysis being allowed to join Medicare Advantage from 2021 (likely, not confirmed by CMS), so it reflects who stays in Original Medicare, not less dialysis per patient. Rates are CMS's published figures (4 decimals), counted over every beneficiary, not only users, so a service few people use (dialysis, long-term care hospitals) can have a high rate in a county with a few heavy users. CMS suppresses figures built on fewer than 11 beneficiaries or service users ('*'); they are NULL here and never filled in from other figures. CMS suppresses all of a service's figures when fewer than 11 beneficiaries in the place used it: 3,830 county cells here, mostly long_term_care_hospital, inpatient_rehabilitation_facility, outpatient_dialysis_facility; Georgia's row is never suppressed. A 0 is real (no user). CMS's GA-UNKNOWN row (13000), beneficiaries placed in no county, is fully suppressed and left out. Georgia's row is CMS's own figure, not a sum of the county rows: it also counts the handful of beneficiaries CMS places in no county (GA-UNKNOWN), so the 159 counties need not add up to it exactly. age_bracket 'under_65' (beneficiaries entitled by disability or end-stage renal disease) and '65_plus' exist for Georgia only; county rows are 'all'. The two brackets add up to 'all': beneficiaries, and each service's stays, days, visits, episodes or events (rate x beneficiaries / 1,000). CMS also publishes how many beneficiaries used each service and their share; those count people, not volume, and are not in this table. From CMS's 2014-2024 release (file updated 2026-04-29); each release restates earlier years, so the figures here are replaced when the data is refreshed. tract_geoid is always NULL.","null_semantics":{"suppressed_to_null":true,"zero_is_real":true},"available_years":[2014,2015,2016,2017,2018,2019,2020,2021,2022,2023,2024],"schema_hash":"0870c5205bdab185fc494ca90d6483b5d0cd96edd28ee0c19c41473481f7ada5","tags":["health","centers for medicare & medicaid services","medicare_service_utilization"],"granularity_description":"One row per year, county_fips, tract_geoid, age_bracket, service, volume_unit (geography columns are NULL at higher aggregation levels).","example_queries":[{"description":"Latest year (2024), counties detail","query":"SELECT * FROM medicare_service_utilization WHERE year = 2024 LIMIT 100"},{"description":"Filter by age_bracket = all","query":"SELECT * FROM medicare_service_utilization WHERE age_bracket = 'all' AND year = 2024 LIMIT 100"}],"foreign_keys":[{"column":"county_fips","target_object":"counties","target_columns":["county_fips"],"attribute_columns_added":["county_name"],"scope":"global"},{"column":"tract_geoid","target_object":"tracts","target_columns":["tract_geoid"],"attribute_columns_added":["tract_name"],"scope":"global"}],"filters":[{"param":"year","kind":"year_exact","type":"integer","multi_value":false,"enum_enforced":false,"values":null},{"param":"year_min","kind":"year_range","type":"integer","multi_value":false,"enum_enforced":false,"values":null},{"param":"year_max","kind":"year_range","type":"integer","multi_value":false,"enum_enforced":false,"values":null},{"param":"detail","kind":"detail","type":"string","multi_value":false,"enum_enforced":true,"values":["counties","states"],"default":"counties"},{"param":"county_fips","kind":"foreign_key","column":"county_fips","type":"string","multi_value":true,"enum_enforced":false,"values":null},{"param":"tract_geoid","kind":"foreign_key","column":"tract_geoid","type":"string","multi_value":true,"enum_enforced":false,"values":null},{"param":"age_bracket","kind":"categorical","column":"age_bracket","type":"string","multi_value":true,"enum_enforced":true,"values":["all","under_65","65_plus"],"has_total":true,"total_value":"all","requires_single_value":false,"nullable":false,"null_filter_token":null},{"param":"service","kind":"categorical","column":"service","type":"string","multi_value":true,"enum_enforced":true,"values":["inpatient","outpatient","ambulatory_surgery_center","skilled_nursing_facility","inpatient_rehabilitation_facility","long_term_care_hospital","home_health","hospice","evaluation_and_management","procedures","tests","imaging","durable_medical_equipment","outpatient_dialysis_facility","fqhc_and_rural_health_clinic","ambulance","treatments"],"has_total":false,"total_value":null,"requires_single_value":true,"nullable":false,"null_filter_token":null},{"param":"volume_unit","kind":"categorical","column":"volume_unit","type":"string","multi_value":true,"enum_enforced":true,"values":["stays","days","visits","episodes","events"],"has_total":false,"total_value":null,"requires_single_value":true,"nullable":false,"null_filter_token":null}],"filter_hints":[{"primary":"volume_unit","dependent":"service","note":"Each service is published only in its own unit(s): stays and days for inpatient, skilled_nursing_facility, inpatient_rehabilitation_facility, long_term_care_hospital, hospice; visits for outpatient, outpatient_dialysis_facility, fqhc_and_rural_health_clinic; events for ambulatory_surgery_center, evaluation_and_management, procedures, tests, imaging, durable_medical_equipment, ambulance, treatments; episodes and visits for home_health. Set volume_unit to one the service has."}],"non_additive":[{"column":"service","reason":"Services are different kinds of care counted in different units (stays, days, visits, episodes, events), and one visit can count under several services (the hospital's outpatient visit and the doctor's E&M, procedure, test or imaging events). num_om_beneficiaries repeats on every service row. A sum across services is not a total of anything: compare services side by side."},{"column":"volume_unit","reason":"Units count different things: stays and days (or home health episodes and visits) are two measures of the same care, so adding them double counts it, and num_om_beneficiaries repeats on each unit's row. Use one unit at a time."}],"partitions":[],"value_implications":[],"year_meaning":null,"pooled_rates":[],"series_breaks":[],"year_window":null,"pooled_min_count":null,"values_endpoint":"/api/v1/health/medicare_service_utilization/values"}