Revenue integrity · claim pre-check

Shield. See rejection risk before the claim is submitted.

Pre-claim revenue integrity platform identifying reimbursement and audit risks before submission. A risky line item goes to the specialist's queue with its reason; the specialist decides whether it is submitted.

Request a demo or pilot Try the example

Invoice

F-2026-0413

Pre-check 3/5

Out-of-package supply₺1,280In review
Chest CT₺2,340Verified
Outpatient visit₺640Verified
Physiotherapy session₺410Checking
Biochemistry panel₺520Waiting

Submission record

Verified items

F-2026-04124 items · ₺4,980

When the check runs: Pre-claim
Risk score: 0–100
Who decides: Specialist
Audit log: Append-only

Problem

A rejection found after submission doubles the work.

A large share of rejected claims come from preventable errors: missing documents, code mismatches, exceeded limits. These can be caught before submission, by a rule, with a reason.

Rejections arrive late

A rejection usually comes after the claim has been sent. Appeal, correction and resubmission delay payment and take up the team's time.

The rules change often

SUT (Turkey's Health Implementation Communiqué), package definitions and payer contracts are updated often. Tracking them from personal experience means the same item is handled differently by different people.

A score without a reason is not useful

A score that does not say why an item is risky does not tell the team what to do, and cannot be defended in an audit.

How it works

A check layer in front of your billing system.

Shield does not replace your billing system. It runs just before submission, checks each line item against rules and shows the questionable ones to a specialist.

01
Claim and authorisation data
Service lines, diagnosis and procedure codes, pre-authorisation and document status are collected before submission.
- claim lines
- ICD-10
- SUT annex
- authorisation status
02
Pre-check
Versioned rules check code consistency, missing documents, limits and duplicate lines.
- versioned rule set
- reproducible result
03
Risk score
The score is the sum of the contributions of the rules that fired. Each contribution is shown with its rule ID.
- Σ rule contributions
- 0–100
04
Review queue
Items above the threshold go to the specialist's queue. Shield never decides on submission by itself.
- threshold ≥ 40
- HITL
05
Decision
The specialist approves, sends for correction or holds the submission. The decision and reason are stored with the item.
- approve
- correct
- hold
06
Record
Score, rule version, who decided and when are written to an append-only log.
- append-only
- digest chain

Risk score and review queue

A readable reason behind every score.

The risk score is the sum of the contributions of the rules that fired. The specialist sees how much each rule moved the score and decides on that basis.

Select a claim in the queue; the rules behind its score are listed.

Approve it, send it for correction or hold the submission.

Your decision is added to the audit log as a new row and cannot be undone.

Pre-check queue

Claims before submission; select a row to see the details.
Claim	ICD-10	SUT annex	Amount	Risk	Status
CLM-24817Chest CT	M54.5Low back pain	EK-2BFee-for-service	₺4,820	78	In review
CLM-24822Appendectomy package	K35.8Acute appendicitis, other	EK-2CDiagnosis-based package	₺38,150	61	In review
CLM-24830Biochemistry panel	E11.9Type 2 diabetes, without complications	EK-2BFee-for-service	₺1,240	34	Below threshold
CLM-24836Emergency care	R07.4Chest pain, unspecified	EK-2BFee-for-service	₺2,960	12	Below threshold

Rule engine matches · CLM-24817

ruleset v3.8.1

R-112Diagnosis–procedure mismatch+34
R-047Prior authorisation missing+26
R-203Procedure repeated within 30 days+18

Rejection risk (Σ rule contributions)78/100

Specialist decision

Risk threshold 40 exceeded; the submission decision is left to the specialist.

Audit trail

Append-only · each entry includes the previous hash

Seq	Time	Actor	Event	Claim	Hash
1044	09:12:06	system	BELOW_THRESHOLD	CLM-24836	84ccf5a1…4818
1043	09:12:05	system	QUEUED_FOR_REVIEW	CLM-24822	0aa4aeeb…e0b4
1042	09:12:05	system	QUEUED_FOR_REVIEW	CLM-24817	65bd85e9…a866
1041	09:12:04	system	SCORE_COMPUTED	CLM-24817	d90f8ccb…c066

Example · fictional claims. ICD-10 codes are real; the SUT column shows the relevant annex of Turkey's Health Implementation Communiqué.

Audit log

Who decided what, under which rule.

To defend a decision, you have to show in full which data, which rule version and which person it was made with.

Append-only

Once written, an entry cannot be changed or deleted. Even a correction is added as a new entry.

Full context

Each entry holds the rule set version, a digest of the input, the score components, who decided and when.

Reproducible

A past decision can be reproduced exactly, with that day's rule version and data, and shown in an audit.

Exportable

Entries can be exported in a structured format for internal and external audit.

Access and isolation

The institution's data stays with the institution.

Each institution works in its own space. Queries, queues and logs cannot reach another institution's data, and roles decide who can see what.

Isolation: Each institution's data, queue and log are kept separate
Query scope: Institution ID required on every query
Access: Role-based, least privilege

Example roles

Institution admin Sets up users, roles and thresholds	Admin
Revenue analyst Sees the queue, reports and trends	Read
Review specialist Decides on items in the queue	Decide
Auditor Reviews logs and decision history	Read-only

Roles are configured to fit the institution. Shield does not diagnose or make treatment decisions; it works only in billing and pre-authorisation.

Demo and pilot

See Shield on your own claims.

We can run a demo on the rejection reasons you see most often. For a pilot, we write the scope together in a non-binding Letter of Intent (LOI).

Request a demo or pilot

Hospital billing and revenue cycle units
Pre-authorisation and SGK reimbursement teams
Healthcare groups with several hospitals