River
Y CombinatorBacked by Y Combinator
Log InStart Free
FREE TEMPLATE

Clinical Registry Data Submission Pack

Three documents and three sheets that check every required field against the registry's current specification before you submit, not after it rejects the batch.

Edit with AIDownload (.docx + .csv)

Free download  ·  No account needed

What you get

Pack of 6 files: 3 documents, 3 spreadsheets

Clinical Registry Data Submission Pack
  • Required Field RegisterSpreadsheet
  • Completeness CheckSpreadsheet
  • Submission LogSpreadsheet
  • Submission ProcedureDocument
  • Data Definition NoteDocument
  • Validation ApproachDocument
RiverClinical Registry Data Submission Pack

Completeness Check, screening view

Seven of 84 records are wrong in the same field

Illustrative rows for a fictional ambulatory surgical center, Cobalt Ridge Surgical Center, reporting colon-surgery infection surveillance to CDC NHSN for the quarter.

Record IDField CheckedValue ChartedResult
CR-Q1-2026-014Wound ClassCleanNeeds Correction
CR-Q1-2026-027Wound ClassCleanNeeds Correction
CR-Q1-2026-033Wound ClassCleanNeeds Correction
CR-Q1-2026-003Wound ClassClean-ContaminatedPass

Clean is not a valid wound class for a colon procedure under NHSN's own rule, since entering the gastrointestinal tract is never truly clean. Seven of Cobalt Ridge's 84 COLO records were charted that way this quarter. Caught before submission, all 84 stand in the denominator and the rate reads 5.95 percent. Left in, the same 5 infections divide into 77 instead of 84: 6.49 percent, from one field.

Remaining columns: ASA Physical Status, Diabetes, Duration, Corrective Action, Corrected By.

Every clinical registry publishes a specification: a defined set of required fields, a valid-value rule for each one, and a denominator built the registry's own way, not the convenient way. Almost every facility treats that specification as something read once at enrollment and followed from memory afterward. It does not stand still. NHSN republishes its Outpatient Procedure Component manual every January as a new, separately dated document. A checklist read from last year's version and never reopened against the new one is unverified the moment it publishes, changed or not.

A field that is filled in is not the same as a field that is valid, and that gap is where most rejected submissions come from. NHSN's own rule states plainly that the application does not offer Wound Class Clean as a choice for colon surgery and six other categories, since entering the gastrointestinal tract is never truly clean. A record charted that way reads complete on a check that only asks whether the field holds something. This pack runs the check first, so the record returns for correction while the cycle is open.

At Cobalt Ridge Surgical Center, a fictional surgical center, 84 colon-surgery procedures were eligible for infection surveillance in a quarter, and 7 were charted wound class Clean, invalid for that category under NHSN's rule. Caught before submission, all 84 stand in the denominator and the rate reads 5.95 percent. Left uncorrected, the specification excludes the 7 from the denominator. The same 5 true infections then divide into 77 instead of 84, and the reported rate reads 6.49 percent, 9.1 percent relatively higher, from one required field wrong on 8.3 percent of the batch.

Every document in the pack

The record that fails the register's own rule, the rate with and without it, and the two dated deadlines that can block the same month at once.

Clinical Registry Data Submission Pack
RiverThe Miscode

Required Field Register, wound class row

Seven categories can never be charted Clean

NHSN's own rule excludes one value from one field for seven specific operative categories, not a general recommendation buried in a note.

CategoryClean (C) a valid entry?
APPY, BILI, CHOL, COLO, REC, SB, VHYSNo, application-level exclusion
Every other OPC-SSI categoryYes, facility policy decides

Cobalt Ridge, Q1 2026, COLO category

COLO procedures eligible this quarter84Charted wound class Clean7 (8.3%)

The Infection Preventionist does not reinterpret the wound class from the operative note. The correction goes back to the surgical team that was in the case, per NHSN's own instruction.

RiverTwo Rates

Completeness Check, before and after

One field moves the rate 9.1 percent

 Corrected before submissionLeft uncorrected
Denominator8477
True SSI events55
Reported rate5.95%6.49%

The specification's own rule, not a judgment call, drops the 7 miscoded records from the denominator once they are submitted uncorrected. The 5 true infections do not change. Dividing the same numerator into a smaller denominator reads 9.1 percent relatively higher than what actually happened in the operating room.

None of the 5 true SSI events happened to fall among the 7 miscoded records, which is why the numerator holds steady while the denominator moves.

RiverDeadline Stack

Two dated deadlines, one calendar

A correction backlog and a survey can block the same month

RequirementDateConsequence if open
Annual Facility SurveyMarch 1Cannot open a new Monthly Reporting Plan
Wound-class correction backlogBefore month closesRecords excluded from denominator

Cobalt Ridge cleared both in 2026: the Annual Facility Survey nine days ahead of deadline, the 7 flagged COLO records corrected before the quarter closed. Either one left open blocks April's reporting on its own; both open at once do not compound the paperwork, they compound the deadline.

Version-stamping, the other half of the rule

NHSN's Outpatient Procedure Component manual is a new, separately dated PDF every January, with its own dated summary of updates. A register row with no version on it cannot be told apart from one checked against a manual that no longer applies.

What's in the pack

01

Required Field Register

Every required field for the current specification version, its valid-value rule per category, the exact chapter it was read from, and the date it was last checked against the registry's own published version. The same version-and-date discipline applies whether the register covers a registry field or a clinical trial site's essential document.

02

Completeness Check

Every record in a submission batch run against the register before the batch leaves the building, flagged by the exact field and rule it failed rather than a generic incomplete mark. Where a miscode traces back to a charting habit rather than a one-off, a clinical documentation gap review counts it across the whole record instead of just this registry's own denominator.

03

Submission Log

Every cycle's denominator, numerator, pre-submission completeness result and the specification version it was checked against, so a clean cycle and a corrected one are both on the record. The same before-submission discipline applies whether the registry is disease-specific or the qualified registry a MIPS measure names as its own collection type.

04

Submission Procedure

The step-by-step sequence this facility actually follows, from where each field is captured and by whom, to who has authority to release a batch once it clears the check and how a post-submission finding gets logged.

05

Data Definition Note

Where the specification leaves a judgment call to facility policy, such as who assigns a field the registry itself does not define, this is where that policy gets written down, dated, and checked against the current specification.

06

Validation Approach

How the facility confirms a value is actually true, not just present, through internal sampling on a stated schedule and, where a count is generated electronically, a manual cross-check against it before the electronic count is trusted alone.

07

Submission Deadline and Completeness Watch

A periodic pass that ranks any flagged record by how little of the reporting cycle is left to fix it, and separately flags any register row whose specification version predates the registry's current one. Where a payer's own value-based contract requires this registry's outcomes as a term of payment, a value-based contract performance review reads what a clean submission is actually worth once a reconciliation statement arrives.

How to use it

  1. 1

    Open the pack, or download it

    Open the pack in River and send the registry's current specification directly, or download the blank Word and CSV files and build the register from the manual on your own.

  2. 2

    Send a real extract

    The agent runs your most recent cycle's actual records against the register, not a hypothetical batch, and returns which ones fail a rule and on which exact field.

  3. 3

    Correct while the cycle is open

    Every flagged record names the field, the rule it failed, and the register row to check the correction against, with enough of the cycle left to fix it and re-check it.

  4. 4

    Submit, log the version, repeat

    Log the specification version the batch was checked against on the Submission Log, and the watch flags the register again the moment the registry publishes a newer one.

Frequently asked questions

Is this template free?

Yes. Download the three documents and three sheets as Word and CSV files with no signup and no card. "Edit with AI" is a separate, optional path for facilities that want the agent to build the register from their own specification. The rest of the library is at the template library.

What format are the downloaded files?

Word documents (.docx) for the Submission Procedure, the Data Definition Note and the Validation Approach, and CSV (.csv) for the three sheets, zipped into one download. They open natively in Word, Pages, Google Docs, Excel, Numbers and Sheets.

Why version-stamp the register instead of fixing it when something breaks?

Because a break is a submission that already went out wrong. NHSN republishes its Outpatient Procedure Component manual every January as a new, separately dated document, and a rule can change between versions with no announcement to anyone not reading the new one. A dated register row is treated as unverified the moment a newer version publishes, whether or not it actually changed.

Does the Completeness Check replace the registry's own validation?

No, and it answers a different question. Completeness asks whether a field is present and holds a valid value; it says nothing about whether that value is true. NHSN recommends internal validation annually, at minimum, re-abstracting a sample of records from source documentation and comparing each one against what was originally entered.

What if the specification leaves who assigns a field unclear?

That is what the Data Definition Note resolves and dates, rather than guesses at. NHSN's own manual leaves wound class assignment to facility policy, so this pack records which role assigns it and when that policy was last confirmed accurate, instead of treating "we all know how we do it" as a definition.

Is a rejected submission the only cost of an outdated register?

No. NHSN's own Annual Facility Survey is due March 1 each year, and after that date a facility cannot open a new Monthly Reporting Plan until the survey is complete. A correction backlog and an open survey can compound on the same date, blocking more than the one flagged record.

What if a submitted rate needs to actually improve, not just report correctly?

That is a different project from this pack, and a good next step once a submission is clean. A quality improvement project template freezes a baseline before testing a change to the underlying rate, then checks each new period against it with the same run-chart discipline this pack applies to registry versions rather than trend lines.

Sources

  1. CDCFAQs: Surgical Site Infections (SSI) Events (opens in a new tab)
  2. CDCData Validation Guidance (opens in a new tab)
  3. CDCSurgical Site Infections (SSI) (opens in a new tab)

Written by

Chandler Supple

Co-founder & CEO, River

Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.

LinkedInGitHubX

Reviewed by

Brian Joseff, Co-founder, River

LinkedInX

Related workspaces

MIPS reporting checklist template

For choosing the six measures a group reports, including which collection type feeds each one.

Value-based care contract performance review

For what a registry's own outcomes are worth once a payer contract gates payment on them.

Clinical documentation improvement review

For counting a charting gap like this one across the whole record, not just one registry's denominator.

Clinical trial site regulatory binder pack

For the same version-stamped register discipline, applied to a trial site's own binder and its IRB continuing-review deadline instead of a registry specification.

More free templates

Clinical Quality Improvement Project Pack

Freezes a 12-week baseline median before testing a change, then checks each new point against it with the four run-chart rules, not a slope.

Clinical Trial Site Regulatory Binder Pack

Tracks each essential document's version and expiry date, and flags the IRB continuing review 45 days before it lapses, not after a monitor finds it.

MIPS Reporting Checklist Template

Picks the six measures from what your record can already evidence. Under 75 percent completeness a measure scores zero, and the one you excel at is capped.

Patient Satisfaction Survey Analysis Pack

Cross-references comment themes against the patients who scored your lowest dimension worst, not against everyone who left one, before ranking by count.

Check the register before the registry does

Download the blank pack as Word and CSV files, or open this exact pack in River and let the agent build the Required Field Register directly from your registry's current specification.

Edit with AI