{"main_topic":"health","sub_topic":"oasis","topic":"deaths_all_causes_tract","key":"health/oasis/deaths_all_causes_tract","version":"1.0.0","year_min":2005,"year_max":2025,"year_gaps":[],"update_frequency":"annual","detail_levels":["tracts"],"default_detail":"tracts","path_template":"health/deaths_all_causes_tract/year={year}/{detail}.parquet","grain":["year","county_fips","tract_geoid"],"title":"Deaths and Average Age at Death by Census Tract","summary":"Deaths of residents and their average age at death for each Georgia census tract, all causes, in 5-year windows 2001-2005 to 2021-2025.","description":"Deaths of Georgia residents from all causes by 2020 census tract of residence, summed over OASIS's fixed 5-year windows 2001-2005 through 2021-2025, with the average age at death of those deaths, from the Georgia Department of Public Health's OASIS Mapping Tool. Tract rows only; fewer than 5 deaths show as '0-4' (NULL here), one more tract (5 or more deaths) is NULL in some counties to shield the '0-4' tracts from county totals, and no tract rates are published.","explorer":false,"search_aliases":{"en":["deaths by neighborhood","number of deaths","death counts","people dying","neighborhood mortality","average age at death","lifespan","how long people live","small area deaths"],"es":["muertes por vecindario","número de muertes","defunciones","muerte","mortalidad por vecindario","edad promedio al morir","edad al morir","longevidad","cuánto tiempo viven"]},"source":"Georgia Department of Public Health, Office of Health Indicators for Planning (OHIP): OASIS (Online Analytical Statistical Information System) Mapping Tool, mortality map, deaths and average age at death by census tract of residence, 5-Year Aggregates, Save Data export (all causes, races, ethnicities, sexes and ages), retrieved 2026-10-24. Definitions: https://oasis.state.ga.us/gis/TrendableMaps/content/DefTrendMapMort.html","source_url":"https://oasis.state.ga.us/gis/TrendableMaps/agsMortTrend.aspx","endpoint":"/api/v1/health/deaths_all_causes_tract","columns":[{"name":"year","type":"integer","role":"year","nullable":false,"label":"Year","short_description":"","description":"End year of the 5-year window the row covers: 2005 = 2001-2005, 2010 = 2006-2010, 2015 = 2011-2015, 2020 = 2016-2020, 2025 = 2021-2025. Consecutive windows share no calendar year. See year_meaning.","valid_values":null,"example":2025,"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":"Always NULL: tract rows only. The tracts dimension gives each tract's county (also the first five digits of its GEOID).","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":"tract_geoid","type":"string","role":"fk_tract","nullable":true,"label":"Tract GEOID","short_description":"","description":"11-digit 2020 census tract GEOID of the residence of the people who died (FK to the tracts dimension): all 2,796 Georgia tracts in every window, 2001-2005 included.","valid_values":null,"example":"13001950100","unit":null,"value_min":null,"value_max":null,"null_meaning":null,"key_metric_grain_contributor":false,"metric_component":null},{"name":"num_deaths","type":"integer","role":"metric","nullable":true,"label":"Number Deaths","short_description":"Deaths of residents, all causes, in the 5-year window","description":"Deaths of the tract's residents from all causes, summed over the 5-year window (OASIS's count; divide by 5 for a yearly average). NULL where OASIS shows '0-4' (fewer than 5, possibly none), and in one more tract (5 or more deaths) of some counties, our complementary suppression to shield the '0-4' tracts from county totals. The deaths avg_age_at_death averages over. No tract population is published, so counts are not comparable across tracts of different sizes.","valid_values":null,"example":196,"unit":"count","value_min":null,"value_max":null,"null_meaning":"OASIS shows '0-4': fewer than 5 in this tract and window, possibly none. It withholds the exact number for confidentiality; it is never filled in. Also NULL in 61 tracts with 5 or more deaths (our complementary suppression): in every county and window with a '0-4' tract, one more of the county's tracts is withheld, so that the county's total in the county OASIS topics minus the published tracts is never the '0-4' tracts' own total.","key_metric_grain_contributor":false,"metric_component":"denominator"},{"name":"avg_age_at_death","type":"number","role":"metric","nullable":true,"label":"Average Age at Death","short_description":"Average age at death of the tract's residents who died in the window; reflects the tract's age mix, not life expectancy","description":"Average age at death, in years to 0.1, of the tract's residents who died in the window, all causes (OASIS: the sum of the ages at death divided by the number of deaths). It mostly reflects the tract's age structure: a tract of young adults (a military post, a college) averages low, one around a nursing home high. Not life expectancy. NULL where OASIS shows '*' (fewer than 5 deaths), and on the complementary cells (5 or more deaths, our mask; see num_deaths). An average outside 0-120 years would stop the run.","valid_values":null,"example":71.7,"unit":"score","value_min":0,"value_max":120,"null_meaning":"OASIS shows '*': the average age is not published on fewer than 5 deaths. Exactly the tracts and windows whose num_deaths is NULL: OASIS's '0-4' cells, and the 61 complementary cells (our mask) where both are withheld to shield the '0-4' tracts from county totals.","key_metric_grain_contributor":false,"metric_component":null}],"key_metric":"avg_age_at_death","numeric_filter_columns":["num_deaths","avg_age_at_death"],"usage":"Join the tracts dimension on tract_geoid for the tract's name, county and 2020 population. Compare tracts by avg_age_at_death only beside their age mix (the population's ages, for example from ACS tract age tables): a low average marks a young population as much as early deaths. Follow one tract over the five windows (year 2005 to 2025) to see change. To combine tracts, weight each average by its deaths: SUM(avg_age_at_death * num_deaths) / SUM(num_deaths) over rows where both are published, our calculation from averages rounded to 0.1 year, never a plain average of tract averages; it leaves out the '0-4' tracts. num_deaths is a 5-year total: divide by 5 for a yearly average. Never add tracts into a county or Georgia total (the '0-4' tracts are left out); use the county OASIS topics for county figures. Read directly with DuckDB over Parquet.","limitations":"Deaths of a 2020 census tract's residents from all causes, summed over a fixed 5-year window (year 2025 = 2021-2025), not one year. Average age at death mostly reflects the tract's age mix (young adults, a nursing home), not only its mortality risk; it is not life expectancy. Tract rows only: no county or Georgia rows. Deaths and average age are NULL where OASIS shows '0-4' (under 5 deaths) and in one more tract (5+ deaths) in some counties, to shield '0-4' tracts from county totals. OASIS's own definition says the average age at death is not life expectancy and, for all causes, is heavily influenced by the age distribution of the area. Tract averages run from 12.0 years (7 deaths in a tract of Chattahoochee County, where much of the Fort Benning army post lies, 2021-2025) to 87.8 (308 deaths in a Fulton County tract, 2006-2010). Tracts never add up to a county or Georgia total: the '0-4' cells are left out, and the platform builds no total that would reveal them. '0-4' means 0 to 4, so a NULL does not mean at least one death: 12 of the 15 '0-4' tracts of 2021-2025 had no residents in the 2020 Census (water or special-use areas). OASIS's county deaths and YPLL (the county OASIS topics, summed over a window) equal the sum of the county's tracts, so in each of the 61 county-windows with a '0-4' tract one more tract, with 5 or more deaths, is NULL too (our complementary suppression, the same tract in both tract topics): a NULL can hold 5 or more deaths, and a county total minus the published tracts is never the '0-4' tracts' own total. OASIS shows tract counts under 5 as '0-4' and offers only separate 5-year windows, both for confidentiality. OASIS publishes no tract rates (no tract population either), so a raw count is not comparable across tracts of different sizes. A death rate would need a tract population, our own calculation. The windows share no year, so a tract's five rows are a trend at five-year steps, but tract populations change between windows (some tracts that are populous now had under 5 deaths in 2001-2005), so a change in deaths can reflect growth, a new facility such as a nursing home or prison, or how OASIS placed older records on 2020 tracts. Every window is on 2020 census tracts: OASIS's Mapping Tool has used 2020 tract boundaries since 2023-12-01, for its earlier windows too. OASIS does not say how it placed earlier deaths on 2020 tracts; every count is a whole number, so each death sits in one tract and none is split between tracts (our reading). Deaths are placed by the tract of residence as OASIS geocodes it, not where the death happened. 2025, the newest year of the 2021-2025 window, was added to OASIS on 2026-07-29; OASIS adds a year once its data are closed out and marks no year provisional. Not comparable head to head with the county OASIS topics, which are single years with rates: use health/oasis/deaths_by_cause_nchs_rankable for county and state death rates. county_fips is always NULL: the tracts dimension gives each tract's county.","null_semantics":{"suppressed_to_null":true,"zero_is_real":true},"available_years":[2005,2010,2015,2020,2025],"schema_hash":"eb1659b231fea5fdafb1c1930d8db19f288f7f7625dbb3c9b2cd031778f8e709","tags":["health","georgia department of public health, office of health indicators for planning","deaths_all_causes_tract"],"granularity_description":"One row per year, county_fips, tract_geoid.","example_queries":[{"description":"Latest year (2025), tracts detail","query":"SELECT * FROM deaths_all_causes_tract WHERE year = 2025 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":["tracts"],"default":"tracts"},{"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}],"filter_hints":[],"non_additive":[],"partitions":[],"value_implications":[],"year_meaning":"year is the last calendar year of a fixed 5-year window (2005 = 2001-2005, 2010 = 2006-2010, 2015 = 2011-2015, 2020 = 2016-2020, 2025 = 2021-2025): each row covers all five years together, never the single year named. The windows are OASIS's own, back to back, sharing no year.","pooled_rates":[],"series_breaks":[],"year_window":{"years":5,"overlapping":false},"pooled_min_count":null,"values_endpoint":"/api/v1/health/deaths_all_causes_tract/values"}