CDC NNDSS Weekly Outbreak & Disease Anomaly Monitor

Turn the CDC's National Notifiable Diseases Surveillance System (NNDSS) weekly provisional case tables into a clean, deduplicated feed of outbreak signals. For every disease, reporting area, and MMWR week the actor computes a spike ratio (this week vs. the prior 52-week peak), the year-over-year cumulative change, and flags anomalies — new 52-week highs and case spikes — so you can catch emerging outbreaks the moment CDC publishes them.

Data source: CDC NNDSS weekly data on data.cdc.gov (Socrata resource x9gk-5huc). Keyless — no API token required.

Who uses this

• Pharma & vaccine demand forecasting — track measles, pertussis, influenza-associated conditions, and other vaccine-preventable diseases by region and week.
• Public-health & epidemiology analytics — a ready-to-model surveillance feed with spike and YoY metrics already computed.
• Newsrooms & data journalists — surface where a disease just hit a 52-week high.
• Health-tech & risk teams — feed outbreak signals into dashboards and alerts.

Example input

{
  "diseases": ["Measles"],
  "areas": ["Colorado", "New England"],
  "year": "2026",
  "anomalyThreshold": "2.0",
  "onlyAnomalies": false,
  "maxResults": 1000
}

Leave diseases, areas, and year empty to pull every notifiable condition for all available weeks (bounded by maxResults). Disease and area matching is case-insensitive substring — "Measles" matches "Measles, Total" and "Measles, Imported".

Output fields

One row per disease x reporting area x MMWR week:

field	meaning
disease	NNDSS condition label
reporting_area	State / region / national reporting area
year, week	MMWR year and week
current_week_cases	Cases this week (NNDSS m1; null if suppressed)
max_52wk	Max weekly count over the prior 52 weeks (m2)
cum_ytd	Cumulative cases year-to-date (m3)
cum_prior_ytd	Cumulative cases, same period prior year (m4)
spike_ratio	current_week_cases / max_52wk (zero-safe; null when max is 0)
yoy_cum_change_pct	(cum_ytd - cum_prior_ytd) / cum_prior_ytd * 100 (zero-safe)
anomaly	True for a new 52-week high or a spike at/above the threshold
anomaly_reasons	new_52wk_high and/or spike_ratio>=<threshold>
sort_order	NNDSS stable sort key

Filters

• diseases[], areas[] — case-insensitive substring, OR-ed within each list.
• year, weekFrom, weekTo — time window.
• minCases — minimum current-week count.
• anomalyThreshold — spike multiple (default 2.0).
• onlyAnomalies — return only flagged rows.
• maxResults — hard cap.

Use as an MCP tool

This actor is available to AI agents (Claude, Cursor, etc.) via mcp.apify.com — an agent can call it to fetch weekly outbreak anomalies and chain the structured rows into a larger public-health or forecasting workflow.

FAQ

How do I detect a measles outbreak by state? Set diseases: ["Measles"], add your areas, and set onlyAnomalies: true to get only the disease-weeks that hit a new 52-week high or spiked.

Where does the data come from? CDC NNDSS weekly provisional case tables published on data.cdc.gov (Socrata x9gk-5huc), keyless.

What does spike ratio mean? Current-week cases divided by the highest weekly count in the prior 52 weeks. A ratio at or above 1 means this week matched or beat the year's peak.

Why are some case counts null? CDC suppresses small or non-notifiable cells; those arrive as flags with no numeric value and are returned as null.

How current is the data? It tracks CDC's weekly publishing cadence (provisional MMWR weeks), typically updated within days of each MMWR week.