Clinical Trial Results Table QA API

Structural QA of two normalized tables. Product page and free test form.

Endpoint

Paid	`POST https://www.spreadrun.com/api/v1/clinical-trial-table-validator`, API key required, $0.25 per completed audit
Demo	`POST https://www.spreadrun.com/api/demo/clinical-trial-table-validator`, no key, 512 KB, 10 runs per day
Content-Type	`application/json`

Request

A JSON object with exactly two string fields, each holding a UTF-8 CSV table with a header row. Extra fields are rejected.

{
  "studiesCsv": "trial_id,condition,phase\nNCT90000001,Type 2 diabetes,PHASE3\n...",
  "outcomesCsv": "trial_id,outcome_id,outcome_type,primary_endpoint,result_status,results_first_post_date\n..."
}

Field	Required columns
`studiesCsv`	`trial_id`, `condition`, `phase`. One row per trial.
`outcomesCsv`	`trial_id`, `outcome_id`, `outcome_type`, `primary_endpoint`, `result_status`, `results_first_post_date`. One row per outcome measure.

Rejected with HTTP 400 and not charged: a body that is not this JSON object, a missing or empty table, missing required columns, duplicate headers, more than 40 columns, rows with a different number of cells than the header, malformed CSV, more than 10,000 rows in a table, more than 5 MiB of CSV in total, or a request body over 4.4 MB (HTTP 413). The last limit comes first in practice.

Report

Field	Meaning
`status`	`PASS` when no rule fired, otherwise `FAIL`.
`rowCounts`	Data rows read from each table.
`fieldCoverage`	Share of rows with a non-empty value, per required column, 0 to 1.
`issueCount`, `issueCounts`	Total findings and findings per code.
`issues`	Up to 1,000 findings: `table`, `record` (1-based data row, header excluded), `field`, `code`. Cell values are never echoed.
`issuesTruncated`	True when more than 1,000 findings exist.
`inputSha256`	SHA-256 of each submitted table, so you can tie a report to an exact input.
`scope`	What the audit does not certify.

The audit is deterministic: the same tables always produce the same report.

Finding codes

Code	Meaning
`MISSING_VALUE`	A required column is empty in this row.
`INVALID_NCT_ID`	trial_id is not "NCT" followed by exactly 8 digits.
`DUPLICATE_KEY`	Repeats an earlier row: trial_id in the studies table, trial_id plus outcome_id in the outcomes table.
`ORPHAN_OUTCOME`	The outcome points to a trial_id that has no row in the studies table.
`INVALID_OUTCOME_TYPE`	outcome_type is not PRIMARY, SECONDARY or OTHER_PRE_SPECIFIED.
`INVALID_ISO_DATE`	results_first_post_date is not a real calendar date written as YYYY-MM-DD.

Example response

A paid audit of the synthetic sample tables with defects (studies, outcomes). Generated by running the real endpoint code.

{
  "requestId": "7f3c2a1e-5b8d-4c6f-9e0a-1d2b3c4d5e6f",
  "api": "clinical-trial-table-validator",
  "mode": "paid",
  "charged": true,
  "priceCents": 25,
  "balanceCents": 950,
  "report": {
    "schemaVersion": 1,
    "status": "FAIL",
    "rowCounts": {
      "studiesCsv": 6,
      "outcomesCsv": 7
    },
    "fieldCoverage": {
      "studiesCsv": {
        "condition": 1,
        "phase": 1,
        "trial_id": 1
      },
      "outcomesCsv": {
        "outcome_id": 1,
        "outcome_type": 1,
        "primary_endpoint": 0.857143,
        "result_status": 1,
        "results_first_post_date": 1,
        "trial_id": 1
      }
    },
    "issueCount": 7,
    "issueCounts": {
      "DUPLICATE_KEY": 2,
      "INVALID_ISO_DATE": 1,
      "INVALID_NCT_ID": 1,
      "INVALID_OUTCOME_TYPE": 1,
      "MISSING_VALUE": 1,
      "ORPHAN_OUTCOME": 1
    },
    "issues": [
      {
        "table": "studiesCsv",
        "record": 4,
        "field": "trial_id",
        "code": "INVALID_NCT_ID"
      },
      {
        "table": "studiesCsv",
        "record": 6,
        "field": "trial_id",
        "code": "DUPLICATE_KEY"
      },
      {
        "table": "outcomesCsv",
        "record": 4,
        "field": "primary_endpoint",
        "code": "MISSING_VALUE"
      },
      {
        "table": "outcomesCsv",
        "record": 5,
        "field": "outcome_id",
        "code": "DUPLICATE_KEY"
      },
      {
        "table": "outcomesCsv",
        "record": 3,
        "field": "results_first_post_date",
        "code": "INVALID_ISO_DATE"
      },
      {
        "table": "outcomesCsv",
        "record": 6,
        "field": "trial_id",
        "code": "ORPHAN_OUTCOME"
      },
      {
        "table": "outcomesCsv",
        "record": 7,
        "field": "outcome_type",
        "code": "INVALID_OUTCOME_TYPE"
      }
    ],
    "issuesTruncated": false,
    "inputSha256": {
      "studiesCsv": "34a162e906102e1701f4f90de8eea9c57fade81ac11de5310c11b44d46ef4ae9",
      "outcomesCsv": "4e355d3c00765d6bb0e63e7352f923cf5fda2d8a9d255ae1b57c0dc0639cca96"
    },
    "scope": "Structural QA only; no source authenticity, clinical accuracy, freshness, regulatory compliance or reuse-rights certification."
  }
}

Errors

See the shared error table. Example input error (HTTP 400, not charged):

{
  "error": {
    "code": "input_error",
    "message": "studiesCsv: unique headers and required columns are mandatory (maximum 40).",
    "requestId": "7f3c2a1e-5b8d-4c6f-9e0a-1d2b3c4d5e6f",
    "charged": false
  }
}

Code samples

curl -X POST https://www.spreadrun.com/api/v1/clinical-trial-table-validator \
  -H "Authorization: Bearer $SPREADRUN_API_KEY" \
  -H "Content-Type: application/json" \
  --data "$(jq -n --rawfile s studies.csv --rawfile o outcomes.csv '{studiesCsv: $s, outcomesCsv: $o}')"