HI530 Unit 2 Terminology Comparison Example

Recording Meaning or Counting It: A Unit 2 Terminology Comparison of Three Reference Terminologies and Three Classifications Across Six Attributes

[Student Name]

Purdue University Global

HI530: Clinical Terminologies, Vocabularies, and Information Standards

Unit 2 Terminology Comparison

[Instructor Name]

[Date]

The comparison uses publicly documented features of each code system; the sources cited are real.

The Dividing Line

Every code system answers a question about purpose before it answers any question about structure. A reference terminology tries to represent clinical meaning as precisely as clinicians express it, so that a record can say exactly what was observed, found or given. A classification groups related conditions or services into a manageable number of categories, so that cases can be counted, compared and paid for. Neither purpose is better. Desiderata for clinical vocabularies drew this line early, warning that a single vocabulary cannot serve recording and aggregation equally well (Cimino, 1998). The comparison below reads each system against that line.

Comparison Matrix

Each system is described on six attributes: purpose, unit of meaning, structure, granularity, maintaining body and release rhythm.

1. SNOMED CT. Purpose: recording clinical findings, disorders, procedures, organisms, substances and more in the record. Unit of meaning: the concept, each with a unique identifier and several descriptions. Structure: concepts linked by defining relationships in a polyhierarchy, so one concept can have several parents. Granularity: very fine. Maintaining body: SNOMED International, with a United States edition from the National Library of Medicine. Release rhythm: the US edition is released twice a year.

2. LOINC. Purpose: identifying laboratory and clinical observations, the question rather than the answer. Unit of meaning: the observation code, built from six naming axes: component, property, time aspect, system or specimen, scale and method. Structure: axis-based names with groupings for related tests. Granularity: fine enough to separate glucose measured in serum from glucose in capillary blood. Maintaining body: Regenstrief Institute. Release rhythm: twice a year (Regenstrief Institute, 2025).

3. RxNorm. Purpose: normalizing names for clinical drugs so different systems can recognize the same medication. Unit of meaning: normalized concepts at several levels, such as ingredient, clinical drug with strength and dose form, and branded drug. Structure: linked concepts connecting ingredients, strengths, dose forms and brands, with links to source vocabularies including product codes. Granularity: the clinical drug level, below which package details are left to other codes. Maintaining body: National Library of Medicine. Release rhythm: weekly updates and a monthly full release (Nelson et al., 2011).

4. ICD-10-CM. Purpose: classifying diagnoses and reasons for encounters for statistics and claims. Unit of meaning: the code, three to seven characters. Structure: chapters, categories and subcategories in a single hierarchy with instructional notes. Granularity: moderate, with residual "other" and "unspecified" codes that collect cases the classification does not distinguish. Maintaining body: the National Center for Health Statistics with the Centers for Medicare & Medicaid Services. Release rhythm: yearly, effective October 1.

5. ICD-10-PCS. Purpose: classifying inpatient procedures for hospital claims. Unit of meaning: a seven-character code, each position carrying a defined meaning such as body part, approach and device. Structure: multiaxial tables from which codes are built. Granularity: fine for inpatient procedures, but only for the purpose of payment and statistics. Maintaining body: the Centers for Medicare & Medicaid Services. Release rhythm: yearly, effective October 1.

6. CPT. Purpose: reporting professional services and procedures, mainly for payment. Unit of meaning: a five-character code describing a service. Structure: sections by type of service, with guidelines and modifiers. Granularity: set by how services are paid, not by clinical detail. Maintaining body: the American Medical Association. Release rhythm: yearly, effective January 1, with more frequent additions for some code categories.

The three reference terminologies differ from one another in what they describe, but all three share a structure built to add detail, while the three classifications share a structure built to set detail aside.

How Structure Allows Detail

SNOMED CT's defining relationships let a concept such as a fracture of a specific bone carry its site, its laterality and its morphology as relationships, and let software infer that it is also a kind of bone injury. LOINC's six axes separate tests that a laboratory might casually call by the same name, which is what lets results from different laboratories be compared. RxNorm's normalized names connect a brand, its generic equivalent and the product codes on a pharmacy claim to a single clinical drug. Integrating systems such as these is the purpose of the National Library of Medicine's Metathesaurus, which links terms that mean the same thing across many vocabularies (Bodenreider, 2004).

How Structure Sets Detail Aside

ICD-10-CM deliberately places many specific conditions in shared categories so that statistics remain stable and comparable across years. Its "unspecified" codes exist because claims must be filed even when documentation is incomplete. ICD-10-PCS is detailed, but only along the axes that matter for inpatient payment and analysis, and it says nothing about a procedure's findings. CPT describes what a clinician did in the terms by which the service is paid, which is why a single code can cover a range of clinically different visits. These are design choices, not defects.

One Patient Through All Six

A composite 64-year-old woman with type 2 diabetes is admitted for a below-knee amputation. Her record, if built well, uses all six systems at once. Her diabetic foot ulcer and peripheral vascular disease sit on the problem list as SNOMED CT concepts, with site and laterality captured. Her hemoglobin A1c and her wound culture are LOINC-identified observations. Her insulin and her new antibiotic are RxNorm clinical drugs, linked to the product codes on the pharmacy's dispensing records. When she is discharged, coders assign ICD-10-CM codes for her diagnoses and an ICD-10-PCS code for the amputation, and the surgeon's professional claim carries a CPT code. Each system does one job. Problems begin only when one is asked to do another's job, such as using the claim's diagnosis codes to decide which patients need foot checks.

What the Comparison Means for Practice

A health system needs both kinds. The record should capture meaning in reference terminologies, problems in SNOMED CT, results in LOINC and medications in RxNorm, while claims and statistics use the classifications. The two are connected through maps, which always lose some meaning when they move from a fine system to a coarse one. Interoperability texts describe exactly this division of labor between terminologies for recording and classifications for reporting (Benson & Grieve, 2021). Choosing one system for every purpose forces either recording to lose detail or reporting to lose comparability.

Limits of the Comparison

Each attribute is summarized, and every system has exceptions, such as extensions, local editions and code categories with different release schedules. Release rhythms change over time, so a health system should confirm current schedules with each maintaining body before planning updates.

References

Benson, T., & Grieve, G. (2021). Principles of health interoperability: FHIR, HL7 and SNOMED CT (4th ed.). Springer.

Bodenreider, O. (2004). The Unified Medical Language System (UMLS): Integrating biomedical terminology. Nucleic Acids Research, 32(Suppl 1), D267-D270. https://doi.org/10.1093/nar/gkh061

Cimino, J. J. (1998). Desiderata for controlled medical vocabularies in the twenty-first century. Methods of Information in Medicine, 37(4-5), 394-403. https://doi.org/10.1055/s-0038-1634558

Nelson, S. J., Zeng, K., Kilbourne, J., Powell, T., & Moore, R. (2011). Normalized names for clinical drugs: RxNorm at 6 years. Journal of the American Medical Informatics Association, 18(4), 441-448. https://doi.org/10.1136/amiajnl-2011-000116

Regenstrief Institute. (2025). LOINC users' guide. Regenstrief Institute.