U.S. Open Data · the register

Every federal dataset on this site, and what each one cannot tell you

Open data is a subset of what LymeHQ holds. The rest is the patient’s own record, which never leaves without a consent. This page lists the public half: where it comes from, how often it changes, what licence it carries, when we last fetched it, which screen draws it, and the counting rule that travels with every figure — and, beside it, the sources this repo looked at and did not wire, in the register’s own words.

Wired
2
Verified · not wired
2
Not located in this catalog
1
Retired
1
Built · blocked upstream
1
A destination, not a source
1
Not a data source
1

How this page stays true

The rows are a module, not copy. The extractor that pulls CDC surveillance imports it, and a check re-derives each verdict every week: the wired datasets must answer with their columns, the live sums must still equal the committed snapshot, the licence on each wired row must equal what the dataset’s own metadata states, and the search that found no tick data must still find none. A verdict that stops being true turns that weekly check red rather than staying on this page.

01 · The register

Sources, with their verdicts

9 sources this repo has looked at. The date on each is when somebody looked, never when the row was edited.

  • WiredWorkingCDC · NNDSS

    Lyme disease surveillance, aggregated with geography, 1992 to 2023

    Reported cases by state and year under the CSTE case definition, which changed in 2008 (the Probable category) and in 2022 (high-incidence jurisdictions count laboratory evidence alone). Three datasets, one per era.

    Counting rule · sent with every figureReported, never estimated. Pulled under fixed published controls: the 2019 sums must equal CDC’s own published figures (23,453 confirmed · 11,492 probable) or the extractor refuses to write, and every year’s state counts must reconcile to the national total. A jurisdiction with no published count is drawn as no count — not reported, or withheld below the source’s display floor — never as zero.
    What it cannot tell youHow many people have Lyme. CDC’s claims analyses estimate roughly 476,000 may be diagnosed and treated a year against tens of thousands reported — an estimate CDC notes may be high, since billing codes do not always reflect a confirmed diagnosis (Kugeler et al., Emerging Infectious Diseases, 2021; CDC, Lyme disease data and surveillance, read 25 September 2026); this series is one side of that gap.
    The verdict · looked at 2026-08-25WIRED 2026-08-25. Socrata REST API, JSON, no auth. The line-listed pair (below) stays deliberately unwired.
    Source
    data.cdc.gov · NNDSS (opens in a new tab)
    Datasets
    84rx-ksgd · qtbi-xd4i · x5j9-wybp
    Cadence
    annual
    Licence
    Public Domain U.S. Government · stated in the dataset’s own metadata (USGOV_WORKS)
    Last fetched
    2026-08-27
    Standard
    CSTE case definition · state names
    Used on
    Lyme, countedThe Lyme Map
  • WiredWorkingCDC · NNDSS

    NNDSS Table II weekly counts, 2014 (historical)

    The weekly provisional table, kept for its seasonal shape: when in the year cases are reported.

    Counting rule · sent with every figureProvisional weekly counts, historical. The peak week is stated with its year; nothing is projected forward.
    What it cannot tell youThis year’s season. It is a 2014 series and says so on the chart.
    The verdict · looked at 2026-08-25WIRED 2026-08-25, for the seasonal pattern only.
    Source
    data.cdc.gov · NNDSS Table II (opens in a new tab)
    Datasets
    y6uv-t34t
    Cadence
    weekly, provisional · series closed
    Licence
    Not stated in the dataset’s metadata
    Last fetched
    2026-08-27
    Standard
    NNDSS weekly tables
    Used on
    Lyme, counted
  • Verified · not wiredCDC · NNDSS

    Lyme disease surveillance, line-listed without geography, 2008 to 2023

    One row per case with age, sex and status, and no geography — CDC’s own de-identification trade-off, which is the same one this platform’s research commons makes.

    Counting rule · sent with every figureNot drawn. The aggregated series carries everything the surfaces render.
    What it cannot tell youWhere a case was. That is exactly what the aggregated datasets carry instead.
    The verdict · looked at 2026-08-25VERIFIED 2026-08-25 and deliberately UNWIRED: nothing on the site needs a row per case.
    Source
    data.cdc.gov · NNDSS (opens in a new tab)
    Datasets
    9mtj-y2ba · abzs-b3gw
    Cadence
    annual
    Licence
    Public Domain U.S. Government · stated in the dataset’s own metadata (USGOV_WORKS)
    Last fetched
    never — nothing is drawn from it
    Standard
    CSTE case definition
    Used on
    no screen
  • Verified · not wiredAHRQ · HCUP

    Hospital stays and emergency visits (HCUP)

    All-payer hospital records: inpatient stays and emergency department visits, and national estimates built from them. Named by AHRQ in the 24 September 2026 session as the source for how people use hospitals.

    Counting rule · sent with every figureNothing is drawn. A figure from it would count stays and visits, never people, and HCUP withholds any count of 10 or fewer.
    What it cannot tell youThe outpatient life — the visits, tests and prescriptions most people with Lyme never go to hospital for. It sees the hospital and nothing else.
    The verdict · looked at 2026-09-25VERIFIED 2026-09-25 and NOT WIRED. The free query tool no longer searches by individual diagnosis code; it groups codes into categories, and most Lyme codes sit in a category of 189 codes covering many other infections, so no figure it returns is a Lyme figure. A Lyme count needs the purchased nationwide files, which require training and a signed data use agreement. We have neither.
    Source
    datatools.ahrq.gov · HCUPnet (opens in a new tab)
    Cadence
    annual
    Licence
    Query tool free · nationwide files purchased under a data use agreement
    Last fetched
    never — nothing is drawn from it
    Standard
    ICD-10-CM, grouped into AHRQ’s clinical categories
    Used on
    no screen
  • Not located in this catalogCDC

    Tick presence or density, by county

    Where the blacklegged tick has been recorded, by county. The kit designed a county picker on it.

    Counting rule · sent with every figureNothing is drawn, because this repo has not located a source it can wire.
    What it cannot tell youAnything, yet. A county lookup here would be invented until a source is located and wired.
    The verdict · looked at 2026-09-19NOT LOCATED IN THIS CATALOG, re-probed 2026-09-19. Six terms — tick · ticks · tick surveillance · tickborne · Ixodes · acarological — each return zero datasets on data.cdc.gov’s Socrata catalog, with the filter proven live by the same run’s difference control (nine for “lyme”, zero for a nonsense term). ⚠ That is a finding about ONE CATALOG under SIX TERMS. It is not a finding about the world: an agency can publish outside a Socrata catalog, and this repo has not looked there. The earlier verdict on this row read “DOES NOT EXIST as open data” and was an unscoped conclusion drawn from a single bare-token search.
    Source
    api.us.socrata.com · six terms (opens in a new tab)
    Cadence
    —
    Licence
    —
    Last fetched
    never — nothing is drawn from it
    Standard
    —
    Used on
    no screen
  • RetiredRuled outCDC · NCHS

    NHANES as a population baseline

    Two prototypes cited it as a live serology baseline.

    Counting rule · sent with every figureNothing is drawn.
    What it cannot tell youAnything about Lyme: no released cycle carries Borrelia content.
    The verdict · looked at 2026-08-25RETIRED 2026-08-25. No released NHANES cycle carries Lyme, Borrelia or tick content; the API path the prototypes cited returns 404 and appears never to have existed. Re-open only with a primary source in hand.
    Source
    cdc.gov/nchs/nhanes (opens in a new tab)
    Cadence
    —
    Licence
    —
    Last fetched
    never — nothing is drawn from it
    Standard
    —
    Used on
    no screen
  • Built · blocked upstreamBlocked upstreamCMS · Medicare

    A beneficiary’s own Medicare claims over Blue Button 2.0

    The patient authorises LymeHQ to read their own Part A, B and D claims (ExplanationOfBenefit) into their record. Not open data: a patient’s own record, arriving from elsewhere with their consent.

    Counting rule · sent with every figureImporting is not sharing. Claims arrive labelled by source and stay closed to providers and research by promise.
    What it cannot tell youCosts outside Medicare, and today, any claims at all: no data has ever moved.
    The verdict · looked at 2026-08-25BUILT 2026-08-25 to the CMS boundary, sandbox only. The sandbox’s synthetic beneficiary logins fail at medicare.gov SSO, as does CMS’s own sample client; reported.
    Source
    bluebutton.cms.gov · api-documentation (opens in a new tab)
    Cadence
    on authorisation
    Licence
    CMS developer terms · patient-authorised
    Last fetched
    never — nothing is drawn from it
    Standard
    FHIR 4.0.1 · ExplanationOfBenefit
    Used on
    Where the connections are
  • A destination, not a sourceRuled outNIH · NCATS

    CURE ID

    A registry clinicians and patients submit case reports to. HHS’s 2026 Lyme initiative added a chronic Lyme form and urged patient submissions.

    Counting rule · sent with every figureNothing is drawn from it; there is no API in either direction.
    What it cannot tell youFeed this site anything. The honest verb is to prepare a submission, and that is not built.
    The verdict · looked at 2026-08-25A DESTINATION, not a source (2026-08-25). Browsing is gated to licensed clinicians; submission is a web form.
    Source
    cure.ncats.io (opens in a new tab)
    Cadence
    —
    Licence
    —
    Last fetched
    never — nothing is drawn from it
    Standard
    —
    Used on
    Where the data goes
  • Not a data sourceHHS · LymeX

    LymeX Living Guidelines

    Cited by a prototype as a guideline feed.

    Counting rule · sent with every figureNothing is drawn.
    What it cannot tell youAnything: lymex.org is a parked domain.
    The verdict · looked at 2026-08-25NOT A DATA SOURCE. lymex.org is a parking lander; LymeX itself publishes at lymeinnovation.org and has no data API. A 200 from a parked domain counts toward nothing.
    Source
    lymeinnovation.org (opens in a new tab)
    Cadence
    —
    Licence
    —
    Last fetched
    never — nothing is drawn from it
    Standard
    —
    Used on
    no screen

02 · The other half

What is not open data on this site

The patient’s record

Symptoms, labs, medicines and household records are the patient’s own, held under row-level access rules and released to nobody until a switch is turned on. Research output leaves only as counts of five or more, at three-digit ZIP and ten-year age bands, enforced inside the database.

Not assessed

A source that is not on this page has not been looked at. The design kit named USGS land cover, NOAA climate normals, Census block groups, the FDA device registers and NIH All of Us as designed-for; none has been probed here, so none is listed as anything. Each would join this register with a verdict, or not at all.

Where to go next