7.2 NAACCR Data Exchange Standards, XML Data Specifications & Registry Certification

7.2 NAACCR Data Exchange Standards, XML Data Specifications & Registry Certification

Central cancer registries compile millions of records each year from diverse reporting sources, including hospital registries, ambulatory surgery centers, commercial pathology laboratories, radiation oncology practices, and vital statistics bureaus. Without standardized nomenclature, uniform item definitions, and harmonized electronic transmission protocols, aggregating and analyzing these disparate data streams would be impossible.

The North American Association of Central Cancer Registries (NAACCR) serves as the recognized professional consensus body governing cancer data standards throughout North America. Founded in 1987 as the American Association of Central Cancer Registries (AACCR) and later renamed when Canadian central registries joined, NAACCR coordinates data standards across all major cancer registry sponsoring agencies.

---

The Consensus Governance of NAACCR

NAACCR is not a regulatory enforcement agency; rather, it operates as an umbrella consensus organization. Through its Uniform Data Standards Committee (UDSC), NAACCR provides the formal mechanism by which key stakeholders establish shared agreements on data items, coding rules, and record layouts.

The standard-setting organizations represented within NAACCR include:

• American College of Surgeons (ACS) Commission on Cancer (CoC)
• National Cancer Institute (NCI) SEER Program
• Centers for Disease Control and Prevention (CDC) NPCR
• National Cancer Registrars Association (NCRA)
• American Joint Committee on Cancer (AJCC)
• Statistics Canada and the Canadian Partnership Against Cancer (CPAC)

When a standard-setting agency proposes a new data item—such as a novel Site-Specific Data Item (SSDI) for molecular testing—the proposal is evaluated by NAACCR's UDSC. This ensures that the item is assigned a unique item number, standard definitions, and validated edit algorithms before implementation across hospital and central registry software platforms.

---

NAACCR Volume II: Data Standards and Data Dictionary

The central technical reference maintained by NAACCR is the multi-volume series Standards for Cancer Registries. Among these, Volume II: Data Standards and Data Dictionary serves as the definitive reference manual for registry software developers and certified oncology data specialists.

Volume II provides a meticulous technical blueprint for every recognized data item in oncology data management. Each data item entry includes:

1. Standard Item Number: A unique 1- to 4-digit numeric identifier assigned permanently to that data field (e.g., Item #390 for Date of Diagnosis, Item #522 for Histologic Type ICD-O-3, Item #1750 for Date of Last Contact or Death).
2. Standard Item Name: The official human-readable label of the field.
3. XML NAACCR ID: The standardized camelCase variable identifier used in modern XML data structures (e.g., dateOfDiagnosis, histologicTypeIcdO3).
4. Field Length and Data Type: The allowable length (in characters) and data format (numeric, alphanumeric, text, or date formatted as YYYYMMDD).
5. Record Type Applicability: Specifies whether the data item is included in specific record types (e.g., Incidence, Confidential, Abstract, or Full Case).
6. Detailed Coding Instructions and Valid Code Values: Explicit, standardized coding tables defining legitimate code entries and handling of unknown or unrecorded values.

NAACCR updates Volume II on a regular revision cycle to incorporate advances in cancer diagnostics, staging revisions, and federal reporting requirements (for example, Version 25 and Version 26).

---

Evolution of Data Exchange: Fixed-Column ASCII to NAACCR XML

For nearly three decades, cancer registry data transmission relied on fixed-column ASCII flat files. In a fixed-column format, each record was structured as an enormous single string of text where every data item occupied an immutable character position. Records were many thousands of characters wide, and every new item or wider field forced software changes across the whole community.

Limitations of Fixed-Column ASCII

While fixed-column records were computationally straightforward during the 1990s, they created severe structural bottlenecks in modern health informatics:

• Inflexible Schema: Adding a single new data item or expanding a field length required altering column positions across the entire layout, breaking software parsers nationwide.
• Padding Overhead: Unused fields had to be padded with thousands of blank spaces or leading zeroes to preserve exact column positioning.
• Narrative and Genomic Data Incompatibility: Fixed-width fields could not gracefully accommodate lengthy pathology text, next-generation sequencing genomic panels, or unstructured clinical notes without truncation.

The NAACCR XML Data Exchange Standard

To overcome these constraints, NAACCR developed the NAACCR XML Data Exchange Standard. The NAACCR Board approved it in 2015, it ran alongside the fixed-width layout during a transition period, and it became the only exchange format when the fixed-width layout was retired with Version 21.

NAACCR XML is a modern, extensible hierarchical data exchange architecture governed by a formal XML Schema Definition (XSD). The standard replaces rigid column counting with self-describing XML tags.

┌────────────────────────────────────────────────────────┐
│                     <NaaccrData>                       │
│               (Root Element & Metadata)                │
└───────────────────────────┬────────────────────────────┘
                            │
              ┌─────────────┴─────────────┐
              ▼                           ▼
        <Patient>                   <Patient>
     (Demographics)              (Demographics)
      │            │
      ▼            ▼
   <Tumor>      <Tumor>
 (Primary 1)  (Primary 2)

Hierarchical Structure of NAACCR XML

The architecture of a NAACCR XML file mirrors the clinical reality of oncology care:

1. Root Element (<NaaccrData>): The top-level document element containing file-level metadata attributes: baseDictionaryUri (declaring the NAACCR standard version), userDictionaryUri (referencing custom local extensions), recordType (e.g., "Incidence"), and dataVersion.
2. Patient Element (<Patient>): Encapsulates patient-level demographic attributes that remain constant across all primary tumors diagnosed in that individual. Examples include patientIdNumber, dateOfBirth, sex, race1, spanishHispanicOrigin, and vitalStatus.
3. Tumor Element (<Tumor>): Nested directly within the <Patient> element. A patient diagnosed with multiple independent primary malignancies (e.g., a synchronous colon adenocarcinoma and lung squamous cell carcinoma) contains a single <Patient> container enclosing multiple discrete <Tumor> containers. The <Tumor> element contains tumor-specific attributes: primarySite, histologicTypeIcdO3, behaviorCodeIcdO3, ajccTnmClinicalStageGroup, rxSummSurgery, and dateOfDiagnosis.
4. User-Defined Extension Dictionaries: A transformative feature of NAACCR XML is the ability for central registries or research networks to define custom, localized data items in an external XML dictionary. These extensions are validated seamlessly alongside standard NAACCR items without disrupting base schema compliance.

---

NAACCR Central Registry Certification Program

To ensure that cancer registry data aggregated at the national level are scientifically robust, NAACCR conducts an annual Central Registry Certification Program. Central registries submit their data files during the annual Call for Data, where they are audited against objective, mathematically verified criteria.

NAACCR awards two formal certification tiers: Gold Certification (the highest standard of excellence) and Silver Certification.

Certification Criteria and Quantitative Thresholds

Certification audits evaluate four fundamental pillars of data quality: completeness, timeliness, internal consistency, and duplicate record resolution.

These thresholds come from NAACCR's summary of data quality criteria for the Call for Data. The duplicate criterion has been phased to tumor-level deduplication: for the December 2025 Call for Data, Gold requires fewer than 0.1% duplicates and Silver fewer than 0.2%.

---

Practice Scenario & Exam Pitfalls

Central Registry Operational Scenario

A state central registry completes its annual Call for Data submission to NAACCR for diagnosis year 2024. The audit reveals the following results: Completeness of case ascertainment is 96.2%; EDITS error rate is 0.0%; unresolved duplicates are 0.06%; missing sex is 0.8%; missing county is 1.1%; missing age is 0.5%; and missing race is 3.8%.

Certification Evaluation: Although the registry met the Gold threshold for completeness (96.2% >= 95%), timeliness, duplicates, sex, age, and county, its missing race rate was 3.8%. Because Gold Certification allows no more than 3% missing race, the registry does not qualify for Gold Certification. However, because missing race is below the 5% Silver limit, and all other metrics exceed Silver thresholds (completeness >= 90%), the registry is awarded Silver Certification.

Core Exam Pitfalls to Avoid

• Pitfall 1: Assuming NAACCR certifies hospital cancer registries. NAACCR certifies central (state, territorial, and provincial) cancer registries. Hospital cancer registries are accredited by the American College of Surgeons Commission on Cancer (CoC).
• Pitfall 2: Confusing XML Patient and Tumor element contents. Remember that demographic attributes (patient ID, race, sex, birth date) reside in the <Patient> container, whereas neoplasm-specific attributes (primary site, histology, AJCC stage, surgery) reside in the <Tumor> container.
• Pitfall 3: Inverting Gold vs. Silver completeness thresholds. Gold requires at least 95.0% completeness; Silver requires at least 90.0% completeness. A registry with 94.5% completeness cannot earn Gold, even with zero errors in other categories.