Clinical console

VAERS Alternative

A morning-report desk for the public vaccine-safety files. Read the system before you read the number.

A VAERS report is not a confirmed injury and not proof of cause.

How to read

Five rules before the chart

Use this when someone brings a screenshot, a dashboard, or a state map. It does not say whether a vaccine is indicated for the patient in front of you. It does not calculate a rate.

  1. 01

    Three different words

    A report is a submission. An injury is a diagnosis you have confirmed. Cause is a judgment that survives a comparator. VAERS gives you the first word only.

  2. 02

    Name the system first

    A passive report, a solicited check-in, a chart system with a denominator, a compensation ledger, and a school census answer different questions. Do not let a logo upgrade one into another.

  3. 03

    No denominator, no rate

    Doses, children, person-time, or assessed kindergartners — said in one sentence. A different year’s doses do not count. If you cannot name it, you have a count.

  4. 04

    Age bands have to match

    Kindergarten MMR is not the infant cohort in a SIDS file, and it is not the age mix inside pediatric VAERS. A state outline is not a join key.

  5. 05

    Know what confirmed felt like

    Two childhood signals had a window, a comparator, and a product-specific action: RotaShield and intussusception; first-dose MMRV and febrile seizure. If the chart on the table has none of those, it is not that kind of evidence.

A screenshot on rounds

A colleague forwards a state map. The caption treats kindergarten MMR and infant deaths as one story. Move the control. The page opens the shelf that answers that step.

The caption says the vaccine caused those infant deaths.

A report, an injury, and a cause are three words. This picture has not earned the second or the third.

Opens How to read.

Stack

Which system answers which question

Click a system. The note is the clinical use, then the open link to the primary page. Passive reports, check-ins, chart systems, a compensation ledger, query software, and a school census are not substitutes for each other.

CDC and FDA, public files via HHS

VAERS

Clinical use

Passive surveillance. Anyone may file — clinician, patient, manufacturer, or lawyer. Use the public files to notice an unexpected cluster of a symptom, a product, and an onset window. Say “reports,” not “cases.” The file holds coincidences, incomplete stories, stimulated reporting, and duplicates.

Does not support

Not a confirmed injury. Not proof of cause. Not an incidence. Do not divide a report count by a denominator this file does not contain, and do not put a VAERS total in a note as a risk.

Open datasetsOpen VAERS WONDER

Sources

Library, shelved by what it is

Official pages are the citable record. Index sites redraw a public file. Contested shelves are arguments a family may bring. Method critique names why a chart fails. Open the original before a number leaves the room.

Official

VAERS data sets

vaers.hhs.gov

Clinical use

Primary public release of the report, vaccine, and symptom files. Every index site on this page redraws this release or a lagging copy of it. If a count will be cited, this is the file to name, with the drop date.

Do not cite it for

Not confirmed cases and not causes. The HHS page states that a report is not proof the vaccine caused the event. Keep that sentence attached to any number you lift.

Open original

Charts

Classification, not incidence

These figures say what a system is for, and how far a childhood topic can be pushed. They do not display injury rates. None were computed.

Figure 1. What each system is for

A filled mark is the question that system was built to answer. An empty mark is a misuse. WONDER the software also queries VAERS; that job is the VAERS row. The death-certificate modules are a different row. Nothing here is a rate.

What each system is forVAERS is marked only for spontaneous reports. v-safe is marked only for solicited check-ins. VSD and BEST are marked only for an excess against a comparator. VICP is marked only for a legal ledger. WONDER death files are marked only for death certificates. School coverage is marked only for a school census. Empty marks are questions that row does not answer. Marks are categories, not rates.ReportCheck-inExcessLedgerCensusDeathsVAERSv-safeVSDBESTVICPWONDER deathsSchool coverage
Figure 2. How far a childhood reading can go

Four steps: describe the file, a defined onset window, a comparator with a denominator, and a product-specific action. Filled dots are steps the file can support. A rust ring is a claim placed past the last supported step. Position is a category. It is not an injury rate. No rates are drawn.

How far a childhood reading can goRotaShield intussusception and first-dose MMRV febrile seizure reach all four steps: describe, window, compare, and action. Kindergarten MMR joined to SIDS, to IDEA counts, or to pediatric VAERS reaches only the describe step. Claims that VAERS deaths are caused deaths, that VICP payments are an incidence, or that a borrowed per-dose figure is a comparison sit past the file. Dots are not magnitudes.Calibrated signalRotaShield — intussusceptionDescribeWindowCompareActionMMRV dose 1 — febrile seizureDescribeWindowCompareActionEcologicalKindergarten MMR × SIDSDescribeWindowCompareActionKindergarten MMR × IDEADescribeWindowCompareActionCoverage × pediatric VAERSDescribeWindowCompareActionContestedVAERS deaths read as causedDescribeWindowCompareclaimVICP payments read as incidenceDescribeWindowCompareclaimBorrowed per-dose figureDescribeWindowclaimAction

Do not move the two signals onto other products

  • RotaShield — intussusception. Cluster after dose 1 was confirmed with a comparator. That product was withdrawn in 1999. Later rotavirus vaccines are not this product.
  • MMRV dose 1 — febrile seizure. More seizures in days 5–12 than when MMR and varicella were given separately, in the second year of life. This is not a finding against MMR alone.
  • Kindergarten MMR × SIDS. Two public series can be plotted on one map. They do not share a person, an age, or a vaccination status.
  • Kindergarten MMR × IDEA. Coverage and special-education counts can be described side by side. Eligibility rules are not a biologic outcome.
  • Coverage × pediatric VAERS. Uptake and report volume can both be stated. Report volume has no dose denominator and is not an injury rate.
  • VAERS deaths read as caused. You can describe how many reports have a death checkbox. The claim that those deaths were caused sits past the file.
  • VICP payments read as incidence. You can describe petitions and dollars. Calling that an attack rate skips the denominator and the legal standard.
  • Borrowed per-dose figure. A secondary site can print a quotient. Unless the denominator is the same ages, products, and dates, it is not a rate.

Walk one topic

Drag the control. Two childhood signals go through all four steps. A map join, and any claim that skips a denominator, stops at the first step. Nothing here is a count.

The question names one product and one outcome: this rotavirus vaccine, and intussusception. Reports opened the question. They were not the last word.

Geography

Joins that look tempting

Pick a file people try to color onto a state. Read what that file can support, and the unit mismatch that blocks the join. An unmatched join is a result.

Outline only. No values.

Every state is the same ink. There is no color scale and no rate. The picture is here so a tempting map has a shape to refuse.

Outline of the states with no dataThe states, the District of Columbia, Alaska, and Hawaii are drawn in one fill and one stroke. No state is shaded by coverage, deaths, services, reports, or payments.

Assessed kindergartners, one school year

does not join

Not the infants on a death certificate

Kindergarten MMR

CDC SchoolVaxView. Share of assessed kindergartners documented as up to date for MMR, plus exemptions, by state and school year. The Washington Post repo is a separate, partial school-level cut — do not treat it as this file.

Can support

  • Describe uptake and exemption geography among kindergartners who were assessed.
  • Support an outbreak-readiness conversation about measles in school cohorts.
  • Show that states do not measure or publish the assessment the same way.

Cannot support

  • Cannot support an injury rate, an autism rate, or a mortality effect.
  • Cannot stand in for infant doses. School entry is years after the first MMR.
  • Cannot prove that a gap is refusal rather than missing paperwork.

Unit mismatch. The unit is an assessed kindergartner in a school year. It does not join cleanly to infant deaths, to IDEA eligibility, or to VAERS reports filed at other ages.

Open SchoolVaxView

Ops

Phases 0–4

The work order for anyone extending this desk. This page ships the inventory and the index as a reading surface. It does not build denominators, colored maps, or rates.

Phase 0

Inventory

Done when

Every file you might cite has an owner, a refresh, a license, and a column list taken from the file itself.

Do

Open the primary download or query page and write down what is actually in the release. Note lag. Note which ZIP a dashboard last ingested.

Do not

Do not treat an About page as the inventory. Do not load restricted microdata into a clinical tool. This console stops at public pages.