HIM 680 Module 2 Terminology Short Paper Example

One Answer, Three Codes: Representing Food Insecurity in LOINC, SNOMED CT and ICD-10-CM at Cedar Prairie Health

[Student Name]

Southern New Hampshire University

HIM 680: Advanced Topics in HIM I

Module Two Short Paper

[Instructor Name]

[Date]

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

One Answer, Three Codes: Representing Food Insecurity in LOINC, SNOMED CT and ICD-10-CM at Cedar Prairie Health

When a patient at a Cedar Prairie Health clinic answers yes to a question about worrying that her food would run out, that answer can be recorded in at least three code systems, each built for a different job. Health information professionals are often asked which code is right. The better question is which code is right for which use. This paper follows a single food insecurity answer through LOINC, SNOMED CT and ICD-10-CM, describes what each system captures, shows how the answer is expected to move between systems for exchange and recommends a role for each at the composite four-hospital Iowa system.

LOINC: The Question and the Answer

LOINC was designed to identify observations, the questions asked and the measurements taken, along with their answers. A standardized social needs screener can be represented in LOINC as a panel, with a code for each question and coded answer choices. Recording the food insecurity item this way preserves exactly what was asked and how the patient responded, including answers such as sometimes true or often true. That precision matters for anyone comparing screening results over time or across clinics, since a change in wording changes what an answer means. LOINC does not, however, say whether a clinician judged the patient to be food insecure, and it is not designed for claims or statistical classification.

SNOMED CT: The Clinical Finding

SNOMED CT is a clinical terminology meant to record clinical meaning in the patient record with fine detail and with relationships between concepts. Benson and Grieve (2021) describe it as the most comprehensive clinical terminology in use and stress that its value comes from its concept model, which allows software to reason about how findings relate. A clinician who reviews the screening result and concludes that the patient has food insecurity can record that conclusion as a SNOMED CT finding on the problem list. The finding carries clinical judgment, can trigger decision support such as a referral to a food pantry and can be shared with other systems. It is not built to group patients into the mutually exclusive categories that statistical reporting requires.

ICD-10-CM: The Classified Condition

ICD-10-CM groups conditions into categories for statistics, reporting and billing. Food insecurity falls under the Z59 category for problems related to housing and economic circumstances, which allows a coder to place the condition on an encounter or claim. Official coding guidelines have allowed social risk codes to be assigned from documentation by clinicians other than the provider and from what patients report about themselves after a clinician has reviewed and accepted it in the chart, a rule change that opens the door to coding from screeners at scale. The classification loses detail, since it records a category rather than the exact answer, but it makes the information countable across millions of claims. Truong et al. (2020) showed how rarely that happens in practice, with social risk codes on fewer than 2% of national hospitalizations. Gottlieb et al. (2016) argued that ICD-10 coding offers one practical way to aggregate social data for population health if extraction and documentation practices are standardized.

Moving Between Systems

Federal policy expects social needs data to travel. Its core interoperability data set, the minimum that certified record systems are expected to send and receive, includes data classes for social determinants of health covering assessments, problems or health concerns, goals and interventions (Office of the National Coordinator for Health Information Technology, n.d.). In practice, the screening assessment travels with LOINC codes, the problem with SNOMED CT and, where required, ICD-10-CM, and the referral as an intervention. A system that stores only one of these cannot fully participate in exchange, and a system that stores all three without linking them will produce numbers that do not match.

Where Each System Falls Short in Practice

Each system also fails in predictable ways at Cedar Prairie. The clinic screener was built two years ago with locally invented question codes rather than LOINC, so its results cannot be compared with other organizations or sent to the Medicaid plan in a standard form, and remapping them will take analyst time. SNOMED CT findings for social needs exist in the terminology, but clinicians rarely add them to the problem list because the problem list search returns them below dozens of medical terms, and nobody has built a favorites list. ICD-10-CM codes depend on coders seeing the screening result, yet inpatient coders work from the hospital encounter and seldom open clinic flowsheets. In each case the failure is not in the code system itself but in how the organization configured and staffed it. That is a governance problem, and it is the reason the course turns next to who owns each data element, who defines it and who checks that it is captured.

Recommended Roles at Cedar Prairie

The writer recommends a clear role for each system. LOINC-coded screening results should be the source for measuring screening reach and positive rates, because they preserve exactly what was asked. SNOMED CT findings on the problem list should drive referrals and care team work, because they carry clinical judgment. ICD-10-CM codes should be assigned on encounters and claims whenever the record supports them, so that social need is visible to payers, quality programs and population analysts. Each should be linked to the others through the record, and the data warehouse should load all three. The governance assessment in the next milestone will ask who owns each step. Linking them is what lets a single patient's answer be traced from the screen to the problem list to the claim.

Conclusion

The question of which code is right has no single answer. For the patient worried about running out of food, LOINC keeps her words, SNOMED CT keeps her clinician's judgment and ICD-10-CM makes her one of the thousands the system can count. A health information department that manages all three, and links them, turns a screening answer into care, exchange and evidence. The first practical step is modest: remap the screener's local question codes to LOINC so that every later use starts from a standard source.

References

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

Gottlieb, L., Tobey, R., Cantor, J., Hessler, D., & Adler, N. E. (2016). Integrating social and medical data to improve population health: Opportunities and barriers. Health Affairs, 35(11), 2116-2123. https://doi.org/10.1377/hlthaff.2016.0723

Office of the National Coordinator for Health Information Technology. (n.d.). United States Core Data for Interoperability (USCDI). https://www.healthit.gov/isp/united-states-core-data-interoperability-uscdi

Truong, H. P., Luke, A. A., Hammond, G., Wadhera, R. K., Reidhead, M., & Joynt Maddox, K. E. (2020). Utilization of social determinants of health ICD-10 Z-codes among hospitalized patients in the United States, 2016-2017. Medical Care, 58(12), 1037-1043. https://doi.org/10.1097/MLR.0000000000001418