# NEISS Emergency Room Injury Records 2025

**Every way Americans hurt themselves badly enough to visit the ER in 2025, in the hospital's own words.** Each emergency-room visit in CPSC's NEISS sample carries the coder's one-line story of what happened, the product involved, the injury, and tags from the Clef model: three emoji, how it happened, what the patient was doing, and a score for how strange it is. Sort by `weird_score` to find the…

NEISS Emergency Room Injury Records 2025 is an open, machine-readable dataset published on Mostly Right as immutable Parquet versions, with a documented column schema, a JSON query API and a direct download. Its first rows are free to read and it is open to AI agents.

## Facts

- Slug: `neiss-emergency-room-injury-records-2025`
- Canonical page: https://mostlyright.md/datasets/neiss-emergency-room-injury-records-2025
- Publisher: mostlyrightmd (https://mostlyright.md/publishers/mostlyrightmd)
- Published: 2026-10-16T11:36:14.872Z
- Tables: 1
- Formats: Apache Parquet (download), JSON (query and sample)
- Cost: the first 500 rows of each table are free to everyone; querying and downloading need a free account and count against a monthly free allowance unless the workspace is subscribed
- Metadata API: `GET https://mostlyright.md/api/v2/public/datasets/neiss-emergency-room-injury-records-2025`
- Topics: emergency room, injuries, consumer product safety, neiss, cpsc, accidents, injury narratives

## How to use it

1. Read a table's schema — no account needed.
2. Fetch its published sample rows — no account needed.
   The first 500 rows of every table, in stored order, are free to everyone, with or without an account. Without an account you get that sample and nothing else: no filter, sort, search, download or page past row 500.
3. Sign in at https://mostlyright.md/login, connect this dataset to a workspace on its page, and create an `mr_use_` key.
4. Query bounded JSON, or download the complete Parquet snapshot and read it locally.
   A free workspace may read 100,000 rows and run 1,000 queries a UTC month past the sample. The allowance is shared by everyone in the workspace, and a free workspace may have any number of members. Reads count against the workspace whose data or key they use. A subscribed workspace is never counted, for anyone working in it.

Agents can do all of this over MCP at `https://mostlyright.md/api/mcp` — searching, reading schemas and sampling rows work with no credential at all.

## Table: injuries

- Table id: `4605f6f5-c69c-45a8-be25-8eef6d9507fc`
- Slug: `injuries`
- Immutable version: `c6c62c95-ca4d-5f60-9ed8-7fdcb8026d97`
- Schema: `GET https://mostlyright.md/api/v2/public/tables/4605f6f5-c69c-45a8-be25-8eef6d9507fc/schema`
- Sample rows, no key: `GET https://mostlyright.md/api/v2/public/tables/4605f6f5-c69c-45a8-be25-8eef6d9507fc/preview`
- Bounded JSON query: `POST https://mostlyright.md/api/v2/public/tables/4605f6f5-c69c-45a8-be25-8eef6d9507fc/query` with `x-api-key: mr_use_…`
- Full Parquet snapshot: `GET https://mostlyright.md/api/v2/public/tables/4605f6f5-c69c-45a8-be25-8eef6d9507fc/current` with `x-api-key: mr_use_…` (application/vnd.apache.parquet)

### Columns (42)

| Column | Type | Description |
| --- | --- | --- |
| `case_number` | integer | CPSC case number, unique per ER visit in the NEISS sample. |
| `treatment_date` | date | Date the patient was treated in the emergency department. |
| `narrative` | string | The hospital coder's free-text account of what happened, as published by CPSC. |
| `emoji_1` | string | The first best-fitting emoji from Clef's forty, kept only when its score is at least 0.5. Empty when fewer fit. The vocabulary: 🪜 ladder, 🛏️ bed, 🛋️ furnitu… |
| `emoji_1_p` | number | Clef's uncalibrated yes score for emoji_1, from 0 to 1. |
| `emoji_2` | string | The second best-fitting emoji, kept only when its score is at least 0.5. |
| `emoji_2_p` | number | Clef's uncalibrated yes score for emoji_2, from 0 to 1. |
| `emoji_3` | string | The third best-fitting emoji, kept only when its score is at least 0.5. |
| `emoji_3_p` | number | Clef's uncalibrated yes score for emoji_3, from 0 to 1. |
| `emojis` | string | emoji_1 to emoji_3 joined with commas, for search and filtering. |
| `mechanism` | string | Clef's tag for how the injury happened, from the narrative alone: fall, struck_by_or_against, cut_or_pierced, burn, caught_or_crushed, overexertion_or_twist, b… |
| `activity` | string | Clef's tag for what the patient was doing: sports_or_exercise, play, household_chores, cooking_or_eating, home_repair_or_yard_work, bathing_or_personal_care, s… |
| `holiday` | string | Clef's tag for a holiday the narrative names: none, new_year, fourth_of_july, halloween, thanksgiving, christmas, other_holiday. |
| `weird_score` | number | Clef's uncalibrated score, 0 to 1, that this is an unusual or absurd way to get hurt. Sort by it to find the strange ones. |
| `stunt_score` | number | Clef's uncalibrated score, 0 to 1, that a stunt, dare, prank or showing off caused the injury. |
| `age_years` | integer | Patient age in years. Children under two are 0, with their age in age_months. Empty when unknown. |
| `age_months` | integer | Age in months for patients under two years old. |
| `sex` | string | Patient sex as coded by CPSC. |
| `race` | string | Patient race as coded by CPSC. N.S. means not stated. |
| `race_other` | string | Free-text race when the code is OTHER. |
| `hispanic` | string | Hispanic ethnicity as coded by CPSC. |
| `body_part` | string | Main body part injured. |
| `diagnosis` | string | Main diagnosis. |
| `diagnosis_other` | string | Free-text diagnosis when the code is OTHER. |
| `body_part_2` | string | Second body part injured, when recorded. |
| `diagnosis_2` | string | Second diagnosis, when recorded. |
| `diagnosis_2_other` | string | Free-text second diagnosis when its code is OTHER. |
| `disposition` | string | What happened to the patient: released, admitted, transferred, held for observation, left, or died. |
| `location` | string | Where the injury happened, such as HOME, SPORTS or STREET. |
| `fire_involvement` | string | Whether a fire was involved and whether the fire department attended. |
| `product_1` | string | Main consumer product involved, in CPSC's product category wording. |
| `product_2` | string | Second product involved, when recorded. |
| `product_3` | string | Third product involved, when recorded. |
| `product_1_code` | integer | CPSC product code for product_1. |
| `product_2_code` | integer | CPSC product code for product_2. Empty when none. |
| `product_3_code` | integer | CPSC product code for product_3. Empty when none. |
| `alcohol_involved` | boolean | CPSC's flag that alcohol was involved. |
| `drug_involved` | boolean | CPSC's flag that drugs or medication were involved. |
| `stratum` | string | NEISS sampling stratum of the hospital: S small, M medium, L large, V very large, C children's. |
| `psu` | integer | NEISS primary sampling unit, the hospital's sample identifier. |
| `weight` | number | Statistical weight. Sum it to estimate national ER visits. |
| `tag_model` | string | The Clef checkpoint that answered for this row. |

## Citing this dataset

Cite the dataset by name and canonical URL (https://mostlyright.md/datasets/neiss-emergency-room-injury-records-2025), the table, and the immutable `version_id` the rows came from. A query response carries `table.version_id` and `table.content_digest`; a download carries a strong ETag and a `Content-Digest` header.