Data Sources & Methodology

Real-time AQI observations by zip code. We query 10001 (Manhattan), 10451 (Bronx), and 11201 (Brooklyn). Returns PM2.5, O₃, and dominant pollutant with AQI category.

⟳ Server cache: 1 hour

Free API key required (AIRNOW_API_KEY env var). Cached hourly server-side via Next.js ISR.

Annual average PM2.5, NO₂, and O₃ by 42 UHF neighborhood, 5 boroughs, and citywide. Live-queried and filtered to the most recent annual period. Used for neighborhood PM2.5 chart, borough pollutant breakdown, and citywide PM2.5 KPI.

⟳ Server cache: 24 hours

Prefers rows where time_period contains 'annual'. Falls back to latest available period if no annual rows found.

SARS-CoV-2 viral load (N1 gene copies/L) from 14 wastewater treatment facilities covering all 5 boroughs. An early warning indicator — wastewater signal typically rises ~1 week before clinical case counts. We average across all sewersheds for the citywide trend.

⟳ Server cache: 24 hours

Methodology changed April 2023 from RT-qPCR to digital PCR — values are not directly comparable across that boundary. Current chart shows most recent 6 months only.

Daily confirmed + probable cases, COVID-confirmed hospitalizations, and death certificates. Includes borough-prefixed columns (bx_, bk_, mn_, qn_, si_) enabling 90-day borough breakdown without a join.

⟳ Server cache: 24 hours

Incomplete records (same-day) excluded via `incomplete='0'` filter. Monthly trend aggregated server-side.

Weekly % of ER visits with ILI diagnosis (fever ≥100°F + cough/sore throat) by borough. Pulled from 53 sentinel hospitals. Season Wk42 2025 – Wk3 2026.

No public REST API — EpiQuery data accessed via web interface and manually updated each season. A civic data gap.

Influenza A viral concentration in wastewater from NYC sewersheds. Filtered to NY state with NYC borough FIPS codes. Averaged across sites by sample date. An early-warning signal that can detect flu trends before clinical reporting.

⟳ Server cache: 7 days

Filtered to pcr_target='fluav' and copies/L wastewater units. NYC FIPS: 36005 (Bronx), 36047 (Kings), 36061 (New York), 36081 (Queens), 36085 (Richmond). 6-month rolling window.

Adult flu vaccination rate (%) by borough. 2023–24 season survey of adults 18+. Source: NYC DOHMH Community Health Survey.

Every restaurant inspection since 2010 — violation codes, critical flags, scores, and grades (A/B/C/N/Z). We query critical violations by cuisine, average score by borough, and current grade distribution using SoQL aggregation.

⟳ Server cache: 1 hour

Every DOHMH rodent inspection result since 2010. Results include Passed, Active Rat Signs, Rat Activity, and Failed. We aggregate the last 30 days by borough.

⟳ Server cache: 1 hour

All 311 complaints since 2010. Noise complaints filtered via `complaint_type like 'Noise%'` and aggregated by borough and type over the last 7 days. Also shown on neighborhood detail pages filtered to the relevant borough.

⟳ Server cache: 1 hour

Live distribution monitoring data: free chlorine (mg/L), turbidity (NTU), fluoride (mg/L), and coliform/E. coli results from sampling sites across the city. We average the last 30 days of Operational samples to populate the water quality table.

⟳ Server cache: 24 hours

Filtered to sample_class='Operational'. Coliform is reported as '<1' for non-detects; we count numeric values ≥1 as detections.

Census-tract level data on food access, low-income populations, and supermarket proximity. Borough % of low-access tracts sourced from 2019 Atlas.

Enterococci bacteria levels at ~20 NYC beaches. EPA advisory threshold is 104 MPN/100ml — above that, swimming is not recommended. Sampled April–September by the DOHMH Beach Surveillance Program.

⟳ Server cache: 24 hours

No explicit pass/fail field in the data. We derive beach safety from the EPA Recreational Water Quality Criterion: ≤104 enterococci MPN/100ml = safe.

Every reported dog bite incident in NYC since 2015. We aggregate the last 12 months by borough and identify the top breed (excluding Unknown/Mixed) per borough.

⟳ Server cache: 24 hours

Every EMS incident dispatch with response times in seconds. We average incident_response_seconds_qy (dispatch to on-scene) by borough for the last 12 months. Only includes validated response times (valid_incident_rspns_time_indc='Y').

⟳ Server cache: 24 hours

Response time = time from 911 call received to first unit arriving on scene. Excludes invalid/missing response times.

Age-adjusted prevalence estimates for all 5 NYC counties. We fetch datavaluetypeid=AgeAdjPrv for 8 measures: obesity (OBESITY), diabetes (DIABETES), depression (DEPRESSION), current asthma (CASTHMA), smoking (CSMOKING), physical inactivity (LPA), high blood pressure (BPHIGH), and no health insurance (ACCESS2).

⟳ Server cache: 7 days

These are BRFSS model estimates — not direct measurements. We filter server-side to NYC county FIPS codes: 36005 (Bronx), 36047 (Kings/Brooklyn), 36061 (New York/Manhattan), 36081 (Queens), 36085 (Richmond/Staten Island).

Model-based estimates for ~4,700 NYC census tracts used in the neighborhood CDC PLACES choropleth map. Five measures displayed: OBESITY, DIABETES, CSMOKING, DEPRESSION, and ACCESS2.

⟳ Server cache: 7 days

Served via internal /api/places proxy route (revalidates weekly). Tract GeoJSON bundled in public/nyc-tracts.json.

Age-adjusted death rates and counts for leading causes by sex and race/ethnicity. We filter sex='Total', group by cause, take the most recent year, and display the top 8 by death count.

⟳ Server cache: 7 days

Cause names include ICD-10 codes which are stripped for display.

Annual HIV diagnoses, death counts, and diagnosis rates per 100,000 by borough, neighborhood, sex, and race/ethnicity. We filter to borough-level totals for the most recent year.

⟳ Server cache: 7 days

Looks for rows where neighborhood='All' for borough totals; falls back to summing neighborhood rows per borough if not found.

Age-adjusted unintentional drug poisoning (overdose) death rate per 100,000 by UHF42 neighborhood of residence. 2023 data (provisional). Sourced from Epi Data Brief No. 150 (October 2025). Fentanyl involved in 80% of deaths. Range: 12.2 (Greenwich Village/SoHo) to 134.8 (Hunts Point/Mott Haven) per 100K.

No public REST API — neighborhood-level rates are published only in DOHMH Epi Data Brief PDFs. Data extracted from the data tables accompanying Brief No. 150. 2023 figures are provisional and subject to revision.

Percentage of singleton live births before 37 weeks gestational age by UHF42 neighborhood. 2020 data (most recent available at UHF42 level). Range: 5.3% (Greenpoint / Greenwich Village) to 14.0% (Northeast Bronx). Citywide average: 8.8%.

Data sourced from the NYC DOHMH Environment & Health Data Portal (EHDP), indicator ID 1 (MeasureID 2 = rate). The EHDP data is published via GitHub (nychealth/EHDP-data) and updated annually but UHF42-level data currently extends only through 2020.

Age-adjusted asthma emergency department visit rate per 10,000 by borough. 2021 data. Source: NYC DOHMH Environment & Health Data Portal.

Life expectancy at birth by borough. 2019 figures used (pre-COVID baseline). Source: NYC DOHMH Summary of Vital Statistics.

% of live births before 37 weeks gestation by borough. 2022 figures. Source: NYC DOHMH Summary of Vital Statistics.

% of K–8 students classified as obese or overweight by borough. 2022 data. Collected during physical education fitness testing.

Statewide Planning and Research Cooperative System. Top ER visit diagnoses use SPARCS discharge counts for NYC facilities. Covers all NYS hospital discharges.

Full SPARCS microdata requires a Data Use Agreement. Aggregate counts shown here are from published DOHMH reports.

Mental health emergency department visit rate per 100,000 population. 2018–2023. Source: NYC DOHMH Epi Data Brief.

Hispanic or Latino Origin by Race at county (borough) level. Variables B03002_001E (total), _003E (NH White), _004E (NH Black), _006E (NH Asian), _012E (Hispanic). Queried for all 5 NYC counties.

⟳ Server cache: 30 days

No API key required for low-volume queries (<500/day). Cached 30 days.

Poverty status in the past 12 months. B17001_001E (total) and B17001_002E (below poverty level) used to compute borough poverty rate %. Displayed on Demographics page.

⟳ Server cache: 30 days

Median household income in the past 12 months (in inflation-adjusted dollars). B19013_001E per county. Displayed on Demographics page.

⟳ Server cache: 30 days

Health insurance coverage status. S2701_C04_001E (uninsured count) and S2701_C01_001E (total civilian population) used to compute borough uninsured rate %. Displayed on Demographics page.

⟳ Server cache: 30 days

Uses the ACS Subject Tables endpoint (/acs5/subject) rather than the detailed tables endpoint.

Asian Alone by Selected Groups (citywide). Covers 20+ distinct Asian subgroups including Indian, Bangladeshi, Chinese, Filipino, Korean, Pakistani, Vietnamese, and others.

Shown as citywide NYC aggregate. Borough-level subgroup data is available but requires additional processing.

Annual telephone survey of ~10,000 NYC adults. Health disparities data by race/ethnicity: diabetes, obesity, hypertension, uninsured rate, smoking. 2022 figures used.

Survey estimates carry ±2–4 percentage point margin of error. Asian health data may under-represent newer immigrant groups.

Life expectancy at birth by race/ethnicity. 2019 pre-COVID baseline. NH Asian: 87.1y, Hispanic: 82.8y, NH White: 81.2y, NH Black: 74.5y. The 12.6-year Asian–Black gap is a key health equity indicator.

Pregnancy-associated deaths by underlying cause, race/ethnicity, and borough. Categories include cardiovascular, hemorrhage, sepsis, mental health/overdose, hypertension, embolism, and homicide. 2016–2017 data.

⟳ Server cache: 30 days

Filtered to related='Pregnancy-Related' (excludes pregnancy-associated but not caused by pregnancy). Data reveals stark racial disparities — Non-Hispanic Black women die at 8–12x the rate of white and Asian women.

Infant, neonatal, and postneonatal mortality rates per 1,000 live births by maternal race/ethnicity. Shows stark racial disparities — Non-Hispanic Black infant mortality is 3–4x that of Asian/Pacific Islander mothers.

⟳ Server cache: 30 days

Field name has a typo in the source dataset: 'materal_race_or_ethnicity' (missing 'n'). Filtered to most recent year. 'Total' and 'Other and Unknown' rows excluded from charts.

Live birth counts by method of delivery (vaginal vs. cesarean) and county of residence. Filtered to NYC counties (New York, Kings, Queens, Bronx, Richmond) to compute borough-level C-section rates.

⟳ Server cache: 30 days

NY State dataset — covers all 62 counties. We filter to 5 NYC counties and compute cesarean percentage = cesarean / (vaginal + cesarean) × 100.

Unintentional drug poisoning deaths by borough, substance, and year. 2017–2024 (2024 is preliminary estimate). Fentanyl involved in ~80% of recent deaths.

NYC-specific annual overdose trend data is not available via a public REST API. Figures sourced from published NYC DOHMH Vital Statistics reports.

% of children under 6 with elevated blood lead levels by borough and year. 2015–2023. Borough trend chart uses the ≥3.5 μg/dL CDC reference value.

Borough-level trend data sourced from published DOHMH surveillance reports. Note: the CDC lowered the reference value from ≥5 to ≥3.5 μg/dL in 2021.

Neighborhood-level childhood blood lead data. Uses field bll_5_total_pct — the percentage of tested children under 6 with blood lead levels ≥5 μg/dL. Shown on individual neighborhood profile pages.

⟳ Server cache: 30 days

Important: this dataset uses the older ≥5 μg/dL threshold, NOT the current CDC reference value of ≥3.5 μg/dL. Rates would be higher if measured at the lower threshold. Filtered to most recent year and UHF42 geography.

Composite Heat Vulnerability Index score (1–5) by UHF42 neighborhood. Score 1 = lowest vulnerability, 5 = highest. Combines surface temperature, green space, air conditioning access, poverty rate, and proportion of non-Latino Black residents (a proxy for structural racism and health disparities).

⟳ Server cache: 30 days

The HVI is a composite index — not a direct temperature measurement. It identifies neighborhoods where residents face the greatest risk during extreme heat events due to the combination of environmental exposure and social vulnerability. Scores ≥4 are flagged as high risk in neighborhood profiles.

Three RSS queries merged and deduplicated: (1) Local NYC media outlets — Gothamist, NY1, amNY, WNYC, Crain's, City Limits; (2) NYC agency news — DOHMH, NYC Health Department; (3) NYC-specific health alerts and outbreaks. Returns up to 12 headlines with source, timestamp, and link. Priority flag applied to CDC/outbreak/emergency items.

⟳ Server cache: 30 minutes

RSS XML parsed server-side with regex (no external package). Google News RSS is free and requires no API key. Results from all 3 queries are merged, deduplicated by URL, and cached for 30 minutes. Items with 'CDC', 'alert', 'outbreak', 'emergency', 'advisory', or 'warning' in title/source are surfaced first and highlighted in amber.

Age-adjusted asthma emergency department visit rate per 10,000 residents by UHF42 neighborhood. 2019 figures used. One of the primary health burden indicators in neighborhood profiles.

Life expectancy at birth by UHF42 neighborhood. 2019 figures (pre-COVID baseline). Ranges from 76.4y (Hunts Point/Mott Haven) to 89.4y (Upper East Side) — a 13-year gap across the city.

Obesity and diabetes rates mapped to UHF42 neighborhoods via census tract aggregation. 2023 PLACES release. Used for obesity%, diabetes% fields in all 42 neighborhood profiles.

⟳ Server cache: 7 days via /api/places

Poverty rate (% below federal poverty line) by ZIP code cluster, mapped to UHF42 boundaries. 2022 ACS 5-year estimates. Used for poverty% in neighborhood profiles.

Polygon boundaries for all 42 United Hospital Fund neighborhoods, simplified and bundled at public/uhf42.json (45KB). Used by the Leaflet choropleth map on the Neighborhood Index page. Properties: GEOCODE, GEONAME, BOROUGH.

UHF42 was designed to align with ZIP code clusters for hospital discharge data. It is the standard geographic unit for NYC public health surveillance and differs from Community Districts and NTAs.

Citywide prevalence estimates for NYC high school students (grades 9–12). Tracks adolescent obesity, smoking, binge drinking, daily soda consumption, and physical activity. 2011–2021 biennial trend data.

⟳ Server cache: 30 days

Filtered to rows where prevalence contains 'Prevalence' (excluding CI rows). All values are strings, parsed to floats. Citywide aggregates only — no borough or school-level breakdown available.

National Health and Nutrition Examination Survey. 58+ blood biomarkers including vitamin D and iron by race/ethnicity and sex. 2017–2020 cycle used. National data — not NYC-specific.

NYC-level biomarker data does not currently exist. NYC HANES (last conducted 2013–14) is the only NYC-specific nutrition survey and is significantly out of date.

Blood biomarker survey of ~3,500 NYC adults covering vitamin D, iron, HbA1c, and more. Conducted 2013–14. This is a genuine civic data gap — no equivalent dataset exists for post-2014 NYC.

NYC HANES needs to be repeated. A 2025 version would enable neighborhood-level nutritional surveillance that currently does not exist anywhere.