Service

Medical coding by certified specialty coders

Coding decides whether you are paid correctly, paid late, or flagged for audit. Our coders hold active AAPC or AHIMA credentials and are assigned by specialty.

Get a free A/R analysis Book a 15-minute call

The problem this fixes

Two coding errors cost practices money in opposite directions. Undercoding — billing a 99213 when documentation supports a 99214 — quietly reduces revenue on every encounter. Upcoding creates audit exposure. Most independent practices drift toward undercoding out of caution, and never learn what that caution costs them.

Then there's specificity. ICD-10 rewards precise diagnosis coding, and unspecified codes increasingly trigger denials or downstream medical necessity challenges.

The cost is invisible from inside the practice, because the people doing the coding are the people who would have to catch it. A provider who has coded a 99213 for the same presentation for six years has no signal telling them the documentation supports more, and no denial arrives to prompt a review. Undercoding is the only revenue problem in a practice that generates no error message at all.

What goes wrong when this is done badly

Four failure modes we see repeatedly, and each one is a process problem rather than an effort problem.

Coding from the superbill alone

A superbill records what was ticked, not what was documented. Coding from full documentation routinely captures services the form did not have a box for, and catches services documented but never billed at all.

Coding to habit after a rules change

E/M selection changed to medical decision making or time, and a lot of practices are still coding to patterns formed under the old documentation rules. The code still exists and still pays, which is exactly why nobody notices.

Unspecified diagnosis codes by default

An unspecified code is faster to select and weaker in every downstream way: it fails medical necessity policies, it undermines authorisation requests, and in risk-adjusted contracts it under-represents the population.

No feedback loop to the provider

A coder who fixes the same documentation gap every month without telling anyone is doing rework forever. Recurring gaps have to go back to the person writing the note.

What's included

Every item below is a deliverable, not a capability statement.

ICD-10-CM diagnosis coding

Coded to the highest specificity the documentation supports.

CPT and HCPCS Level II procedure coding

Full procedure and supply coding against the encounter record.

Modifier assignment and validation

Modifiers 25, 59, XE/XS/XP/XU, 51, 50, LT/RT, 24, 57, 76, 77, 78 and 79.

E/M level assignment

Under current guidelines, based on medical decision making or time.

Surgical coding

Including global period tracking and bundling review.

NCCI edit and MUE compliance checking

Applied before the claim leaves your practice.

Charge reconciliation against the schedule

Catching encounters that were never billed at all.

Coder-to-provider feedback

Documentation gaps reported back so they stop recurring.

Annual code set updates

Applied on schedule, every cycle.

How it works

1 Documentation review Each encounter is read against what was actually documented.
2 Code assignment By a specialty-matched certified coder, not a general pool.
3 Second-level QA review On a [CLIENT INPUT]% sample of coded encounters.
4 Provider query Raised where documentation doesn't support the service billed.
5 Feedback loop Recurring documentation gaps reported monthly.

Code set calendar

Code set	Update cycle	Effective
ICD-10-CM / ICD-10-PCS	Annual	1 October
CPT	Annual	1 January
HCPCS Level II	Annual, with quarterly interim updates	1 January

We apply each update on its effective date. You do not need to ask.

What you get back

Coded encounters returned inside your agreed turnaround window
A monthly documentation-gap report by provider
Compliant provider queries, logged and auditable
Annual code set updates applied without a change request

Who this is for

Practices whose providers code their own encounters
Practices that suspect they are undercoding but cannot prove it
Multi-specialty groups needing coders matched per specialty
Practices preparing for or recovering from a payer audit

What we measure

The numbers this service actually moves. You see all of them weekly, in your own system, with the benchmark alongside.

Coding accuracy rate The share of audited encounters where the code billed is supported by the documentation, measured on a QA sample rather than asserted. Reported by coder and by provider.
E/M level distribution by provider Your distribution against specialty peers. Skewed low is undercoding, and it costs money on every encounter. Skewed high is audit exposure. Both are visible in the same chart.
Query rate and query turnaround How often documentation does not support the service performed, and how quickly providers respond. A rising query rate on one provider is a training conversation, not a coding problem.
Coding-attributable denial rate Denials caused by code, modifier or specificity problems, tracked separately from front-end causes so the two do not hide each other.

Benchmark ranges are sourced from MGMA and HFMA MAP Keys. MedBil's own figures are published only with a stated measurement window, per specification section 12.7.

Where this matters most

The specialties where this part of the cycle carries the most risk and the most recoverable revenue.

Gastroenterology Screening vs. diagnostic colonoscopy, modifier 33/PT, auth Dermatology Lesion coding by size and site, cosmetic vs. medical, pathology Physical Therapy Timed vs. untimed units, the 8-minute rule, therapy caps, plan of care Urgent Care High volume, S-codes, place of service, self-pay mix OB/GYN Global obstetric package, antepartum/postpartum splits, ultrasound Wound Care Debridement by depth, skin substitutes, HBOT, documentation of measurements

Proof

Start with your own numbers

We measure your denial rate, A/R ageing and net collection rate against industry benchmarks before you commit to anything. You keep the analysis either way.

90 days of claims data, no patient identifiers needed
A written analysis, not a sales deck
No contract, and no obligation to proceed
Client results published with a stated measurement window, once available

Questions about medical coding

What credentials do your coders hold?

Active CPC, CPB, CPMA, CIC or COC from AAPC, or CCS, RHIA or CDIP from AHIMA.

Do you code from the chart or from a superbill?

Both, though coding from full documentation captures more.

Will you query me about documentation?

Yes. Compliant queries are how we avoid both undercoding and audit risk.

Can you code for multiple specialties in one practice?

Yes — coders are assigned per specialty.