Prospective payment integrity

Stop the error while it’s still a claim.

Medeblock reads every 837 transaction as it lands, checks it against the known issues that leak money (duplicates, NCCI coding conflicts, drug frequency) and returns a scored decision with the evidence attached. It all happens in seconds, before anything is paid.

Request a demo See how a claim is scored
  • 837I & 837P native
  • Seconds from receipt to decision
inbound · 837P · illustrative t+0.0s 
CLM*PT-40721*1840.00***11:B:1*Y*A*Y*Y~
HI*ABK:M1711~
DTP*472*D8*20260911~
SV1*HC:27447*1450.00*UN*1***1~
SV1*HC:20610*240.00*UN*1***1~
SV1*HC:J1100*150.00*UN*4***1~
LIN**N4*00641036725~
  1. Duplicate reconciliation Clear No match across member history
  2. NCCI procedure-to-procedure Edit hit 27447 / 20610 · modifier indicator 1 · no 59 or X-modifier
  3. Drug frequency Review NDC 00641-0367-25 · units vs label dosing interval
Decision Route to coding review 2 findings · evidence attached · $240.00 at risk

Why post-pay isn’t enough

$104.42M

Overpayments Medicare’s Recovery Auditors identified in 2024 and never got back. That was on top of $382M they did recover, after it had already been paid.

Source: Medicare Recovery Audit Program, 2023 results by region.

Most payment integrity still starts after the money has gone.

It intervenes late

Recovery work begins weeks or months after payment. By then, offsets are contested, providers have moved on and a share of every identified dollar is written off.

It accepts leakage as a cost of doing business

Legacy editing is tuned to what’s easy to catch. Everything that doesn’t fit the rules engine’s structure passes through and becomes someone’s recovery project later.

It scales with headcount

New concepts wait on specialist capacity. Reviewers spend time on low-yield selections, and every false positive adds to provider abrasion.

The pressure isn’t easing

$94B $57B in Medicare and $37B Medicate estimated federal improper payments in 2025
$5.3T US national health expenditure in 2024, up 7.8%
20.6% projected health share of GDP by 2034, from 17.6%

Why now

Data is finally moving at the speed the problem needs. CMS-0057-F requires impacted payers to make claims, encounter and clinical data available within one business day. Together with ACA transparency rules and HHS’s data strategy, that makes prospective, data-driven review practical at scale rather than a pilot.

Sources: CMS NHE fact sheet · HHS data strategy · US Government Accountability Office

Concepts in production

Well-known problems, handled with better data engineering.

Each concept runs as its own pipeline on facility and professional claims, with its own configurable thresholds, monitoring and audit trail.

Duplicate reconciliation

Corrected claims, resubmissions, billing slips and deliberate double-billing. Facility-specific rules and professional-claim nuances keep these alive in every book of business. We don’t force claims into a rigid structure. We re-shape them so matching is near-instant and comprehensive across bill types, overlapping service dates and providers.

Catches
Exact, near and overlapping duplicates; inpatient/outpatient overlap; bill type combinations (111, 121, 13x …)
Evidence
Matched claim, overlap window, rule set applied, similarity score by factor
Action
Auto-deny confirmed duplicates · route suspects for review

National Correct Coding

PTP edits, crosswalks and modifier exceptions, applied to the 837 transaction at the earliest possible point, before adjudication.

Catches
Column 1 / column 2 PTP pairs, same-day and cross-claim, practitioner and hospital editions
Evidence
Code pair, edit effective dates, modifier indicator, modifiers present on the line
Action
Deny column 2 line · allow in line with client preferences

Drug frequency

Pharmacy-related medical claims are one of the fastest-growing cost lines, and checking them against manufacturer dosing usually needs a pharmacist. We move that check from people to data. NDCs and HCPCS drug codes are resolved against approved labeling so every claim is checked, not a sample.

Catches
Administrations more frequent than the label allows; unit counts inconsistent with dose
Evidence
NDC, label dosing interval, prior administrations
Action
Risk-scored for onward review; clear passes go straight through

Transparent by design

Every flag can be defended.

A score is only useful if a reviewer, a provider rep or an appeals team can see why it was given. Medeblock keeps a claim-level integrity record for every decision: which concepts ran, what flagged and when.

  • Claim-level status per concept: clear, suspect, confirmed, compliant, conflict
  • Full audit trail, timestamped, for appeals and provider conversations
  • Pipeline monitoring and value dashboards for each concept
  • Thresholds tuned to each health plan’s claim mix and risk appetite
Claim integrity record
Claim59GD014F5Q
Bill type111 · INP
Service01/10 – 01/13
Review required Suspected duplicate. Hold payment pending review.
DuplicateSuspect · 1 match
NCCINot applicable
Drug frequencyClear
Audit trail
TimeCheckFinding
18:14:02overlap_checkOverlaps claim 59GD01SD5Q by 4 days (01/10–01/13 vs 01/02–01/13)
18:14:02billtype_ruleCriteria applied for 111 → 111 pair
18:14:03scoreSimilarity on member, provider, dates, revenue codes

Working together

Built to strengthen the programs you already run.

Medeblock doesn’t replace your clinical and coding teams. It gives them better selections, so their hours go to the claims that matter and need human oversight.

For health plans

  • Cost avoidance on the front end, not recoveries months later
  • Fewer post-pay disputes and less provider abrasion
  • Affordable at any plan size. You don’t need a large-payer budget
  • Runs 24/7 on cloud infrastructure that scales with volume

For payment integrity partners

  • Higher-yield selection feeding your existing review queues
  • Reusable concept assets across multiple clients
  • Evidence packaged, not just a score
  • New concepts stood up in weeks from the same data foundation

Who we are

Built by people who’ve run systems where audit is non-negotiable.

Institutional-grade engineering. Roots in financial services and telecoms mean data lineage, risk management and auditability are defaults, not features.

A delivery mindset. Global consultancy backgrounds bridge what’s technically possible and what an organisation will actually adopt.

Proven at scale. Data platforms delivered for Fortune 500 and FTSE 100 organisations, integrating with legacy estates while moving them to the cloud.

Graham Murphy

Graham Murphy

CEO & Founder

Award-winning Chief Data Officer. Aligns data maturity roadmaps with the P&L, with a track record of working within the regulatory demands of global finance.

Gill Browne

Gill Browne

Chief Product Officer

Data and AI specialist with deep experience in US healthcare payment integrity. CPC-A certified. Designs data products and MLOps that turn raw claims data into value.

David Power

David Power

CTO & Founder

Architectural lead for large-scale digital transformations. Designs resilient, cloud-native environments built for the largest organisations demands.

See it run on claims like yours.

We’ll walk through a live pipeline and show the evidence behind every flag, and we can scope a pilot on a sample of your 837 volume.

Request a demo
Graham Murphy, CEO grahamjmurphy@medeblock.com +353 86 044 4494