ICD-10 Coding Services: What to Know Before You Outsource

ICD-10-CM sits on virtually every claim your organization sends to a payer. When it is coded correctly, the revenue cycle runs. When it is not, denials accumulate, risk-adjustment revenue goes uncaptured, and audits become expensive conversations. Outsourcing ICD-10 coding is a common and often sound decision. Evaluating which vendor to use is where most organizations go wrong.

This guide is written for revenue cycle and HIM leaders who are evaluating ICD-10 coding services and want a clear frame for what to ask, what to measure, and where the risks sit. It covers accuracy benchmarks, CDI integration, coder credentials, and audit defensibility — the four areas that separate a coding partner from a coding vendor.

ICD-10 is deceptively broad. Over 70,000 codes, updated annually, with guidance layered across the ICD-10-CM Official Guidelines, AHA Coding Clinic, and payer-specific rules. Depth matters more than breadth claims in a coder’s pitch.

Why ICD-10 Coding Is Harder Than It Looks

The nominal job of ICD-10 coding is to translate clinical documentation into diagnosis codes. The harder job is to get the specificity right — every time — across thousands of codes that update every October, under payer rules that sometimes conflict, while maintaining the documentation trail that makes every code defensible in a RADV, OIG, or commercial payer audit.

Three structural challenges make this hard:

1. Specificity Has Real Revenue Consequences

An unspecified code is technically a valid code — but it can lose money, lose HCC capture, and lose a denial fight. Diabetes without complications (E11.9) and diabetes with stage 3 CKD (E11.22 + N18.30) are both defensible if the documentation supports one or the other. Coding the first when the second is documented is a coding error. Coding the first when the second could have been clarified via a query is a CDI failure.

2. The Code Set Keeps Changing

ICD-10-CM adds, revises, and deletes codes every October 1. Entire code families change structure. Guidance in the AHA Coding Clinic updates quarterly. A coding team that is not actively training on the current year’s changes is quietly falling behind by Q2. For details on the current update cycle, see CMS ICD-10-CM guidance.

3. Payer Rules Overlay the Guidelines

Medicare Advantage, commercial, and Medicaid payers each layer their own edits and coverage determinations on top of the national guidelines. A code that is clean on the Medicare side may be denied by a commercial payer for a coverage policy reason. Coders working across multiple payers need to know which rules apply where.

Accuracy Benchmarks Worth Asking About

When evaluating an ICD-10 coding service, the vendor should be able to produce current operational data on:

  • Code-level accuracy rate (target: 95%+ on audited samples).
  • First-pass clean claim rate (target: 95%+).
  • Coding-related denial rate (target: <2% of claims).
  • Query rate and agreement rate (healthy range varies by specialty).
  • DNFB/DNFC days (target: 4 days outpatient, 5–7 days inpatient).
  • HCC recapture rate for risk-adjusted populations (target: 85%+).

A vendor that cannot produce these numbers is not running a measurable accuracy program. A vendor that produces them but cannot break them down by specialty, by payer, or by individual coder is running one that is less mature than it looks. For a deeper treatment of how to build an accuracy measurement program, see Medical Coding Accuracy: How to Measurably Improve It.

CDI Integration: Non-Negotiable

ICD-10 coding without Clinical Documentation Improvement is incomplete. The majority of code specificity issues are documentation issues — the coder is accurate to what is written, but what is written does not fully describe what happened. A coding partner that does not integrate CDI is, by definition, leaving revenue and compliance on the table.

Questions to ask:

  1. What is your query rate, and what is your query agreement rate?
  2. Do you use AHIMA/ACDIS-compliant query templates?
  3. How are queries documented and preserved for audit?
  4. How does query-pattern feedback reach our providers?
  5. Do you measure CDI outcomes — CMI impact, HCC capture lift, specificity improvement — and report them?

If the answer to the last question is “we send queries” without an outcomes story, the CDI program exists but is not being measured. Measurable CDI programs outperform unmeasured ones by several percentage points of accuracy and noticeable revenue lift.

Coder Credentials: What Actually Matters

Medical coding is a credentialed profession. The credentials worth paying attention to:

  • CPC (Certified Professional Coder) from AAPC — baseline outpatient coding.
  • COC (Certified Outpatient Coder) from AAPC — facility outpatient (hospital outpatient, ASC).
  • CIC (Certified Inpatient Coder) from AAPC — inpatient hospital coding, ICD-10-PCS, MS-DRG.
  • CRC (Certified Risk Adjustment Coder) from AAPC — HCC and risk adjustment.
  • CPMA (Certified Professional Medical Auditor) from AAPC — coding audit and compliance.
  • CCS (Certified Coding Specialist) from AHIMA — hospital-based coding.
  • CCS-P (Certified Coding Specialist – Physician-based) from AHIMA — physician office coding.
  • Specialty credentials (CEMC, CCC, COSC, CASCC, and others) for specialty-dense engagements.

Ask for the credential mix of the team that will code your work. “Credentialed coders” as a collective claim is not the same as “every chart coded by a credentialed coder with the right specialty credential for your work.”

Audit Defensibility

The difference between a coding vendor and a coding partner becomes visible when an auditor asks questions. A partner can produce, on request:

  • The coder and QA reviewer credentials behind every claim.
  • The source documentation that supports each code assigned.
  • Any queries that influenced the final code, including the provider response.
  • The audit sample results and remediation history for that coder and specialty.
  • Evidence of the training and quality program that governs the operation.

A vendor who cannot walk through this detail in a proposal conversation will not be able to produce it under audit pressure. The time to validate is during selection, not during a RADV review. Specialty-specific audit risk — especially in multi-specialty groups — deserves its own attention; see Multi-Specialty Medical Coding: What to Look for in a Partner.

Technology Posture: Accelerator, Not Replacement

Computer-assisted coding and AI-driven code suggestion tools are widely deployed now. Used well, they speed up high-volume coding and catch obvious errors. Used as a replacement for credentialed coders on complex or audit-sensitive work, they introduce risk. A good vendor is explicit about where they use technology, what the coder’s role is on each chart, and how QA validates the combination.

Be wary of pitches that claim fully automated coding across broad specialties. Evidence in the AHIMA and HIMSS literature consistently supports the hybrid model — AI-assisted code suggestion paired with credentialed coder review — outperforming either end of the spectrum.

Contract and Operational Questions Worth Asking

  1. Transition plan. Does the vendor run a parallel operation period before cutover?
  2. Rework accountability. Is coding-related rework at no additional cost, with root-cause analysis?
  3. Turnaround time SLAs. Are they specific to encounter type and measured, not aspirational?
  4. Reporting transparency. Can you see operational metrics at any time, not just in a quarterly review?
  5. Staffing model for volume spikes. How do they scale without accuracy drift?
  6. Security posture. HIPAA BAA, SOC 2 Type II, offshore controls if applicable.
  7. Off-boarding provisions. What happens to work-in-progress, denial queues, and documentation if you transition away?

Frequently Asked Questions

How often should ICD-10 coding be audited?

Per-coder audits should run monthly (25–30 charts per coder). Organization-level statistical audits should run at least quarterly. Specialty-dense or high-risk populations warrant more frequent focused audits.

Can we keep our current EHR with an outsourced ICD-10 coding service?

Yes. A competent coding vendor works inside your existing EHR and claim-generation workflow. If a vendor requires you to change systems to work with them, that is a red flag about their operating model.

What is the difference between AAPC and AHIMA credentials?

Both are respected. AAPC is historically more physician-office focused, AHIMA is historically more hospital and HIM focused. Either can be the right credential depending on the role. The best teams have both, matched to the work.

How fast can a coding transition happen?

30–60 days is typical for a clean transition with a parallel operation period. Pushing faster than that usually costs accuracy in the first month.

Is offshore ICD-10 coding reliable?

It can be, with governance. HIPAA BAA, SOC 2 Type II, access controls, U.S.-based QA leadership, and clear data residency rules are the baseline. The quality question is about governance, not geography.

The Bottom Line

ICD-10 coding is a depth game, not a volume game. The vendor that produces the best numbers is the one with credentialed specialty-matched coders, disciplined CDI integration, measurable accuracy and denial rates, and genuine audit defensibility. Volume discounts are not a substitute for any of that. If you are evaluating ICD-10 coding services, Qway Healthcare is glad to walk through our accuracy, denial, and CDI benchmarks on a no-obligation basis and show you what the improvement math looks like for your operation.


Explore more of our Medical Coding & FQHC Billing cluster:

External References

  1. Centers for Medicare & Medicaid Services. “ICD-10-CM Official Guidelines for Coding and Reporting.” https://www.cms.gov/medicare/coding-billing/icd-10-codes
  2. American Hospital Association. “AHA Central Office — Coding Clinic.” https://www.codingclinicadvisor.com/
  3. American Academy of Professional Coders (AAPC). “Medical Coding Certifications.” https://www.aapc.com/certification/
  4. American Health Information Management Association (AHIMA). “AHIMA Certifications Overview.” https://www.ahima.org/certification-careers/certifications-overview/
  5. AHIMA / ACDIS. “Guidelines for Achieving a Compliant Query Practice.” https://acdis.org/resources/guidelines-achieving-compliant-query-practice
  6. National Center for Health Statistics (CDC). “ICD-10-CM Files.” https://www.cdc.gov/nchs/icd/icd-10-cm/index.html

    FQHC Billing and Coding Services: The Complete Guide

    Federally Qualified Health Centers do not bill like the rest of healthcare, and that is the first thing most outsourcing vendors get wrong. FQHC payment runs on a different framework, the documentation rules tie directly to HRSA funding conditions, and the math on...

    Medical Coding Accuracy: How to Measurably Improve It

    "Our coding is accurate" is the most common and least useful claim in the revenue cycle. Accurate against what benchmark? Measured how? On what sample size? Accuracy is a number, not a feeling, and the path to improving it runs through a disciplined measurement...

    HCC Coding Services in the USA for Risk-Adjusted Plans

    Hierarchical Condition Category coding is the single largest revenue lever most risk-adjusted providers leave unpulled. Medicare Advantage plans, certain ACA commercial products, and a growing set of accountable care and risk-bearing arrangements all pay based on...

    4 Proven Methods to Optimize Risk Adjustment

    Imagine a healthcare landscape where providers are fairly rewarded for the quality of care they deliver, rather than just volume. This vision hinges on the crucial process of risk adjustment, which ensures that compensation for healthcare providers reflects the...

    Transformations in Evaluation & Management (E&M)

    The world of healthcare is anything but static; it is a dynamic environment that continuously adapts to new challenges, especially in medical coding and billing. The Current Procedural Terminology (CPT) is a crucial player in this landscape, a comprehensive code...

    CMS HCC Coding: Top Mistakes and How to Prevent Them

    Last time, we broke down the CMS HCC model and showed how it helps match payments to the real care patients need. Now, it’s time to get into making it better, the most common HCC coding mistakes that silently drain your revenue, plus how to avoid them. HCC coding...

    Healthcare has come a long way from what it used to be.

    Imagine a patient walking into the doctor’s office, handing over a few dollars in cash, and walking out with no bills, claims, or paperwork. That was healthcare in America not so long ago. Then came employer-sponsored insurance, followed by the launch of Medicare...

    The Turning Point in Global Healthcare 

    The healthcare world is always grappling with growing complexities—more patients, evolving diseases, tighter regulations, and an overwhelming surge of data. However, at the heart of this storm is a decades-old system, ICD-10. Reliable, yes. But outdated,...