HI530 Unit 8 Seminar Reflection Example

I Called Lab Mapping Clerical Until a Glucose Report Nearly Doubled: A Unit 8 Seminar Reflection on One Local Code and Two Kinds of Blood

[Student Name]

Purdue University Global

HI530: Clinical Terminologies, Vocabularies, and Information Standards

Unit 8 Seminar Reflection

[Instructor Name]

[Date]

The hospital, its report and its counts are composite; the sources cited are real.

Where I Started

In our Unit 2 discussion I wrote, "Mapping local lab codes to LOINC is a one-time data entry task; once the table is built, the hard work is done." I believed it. The six LOINC axes seemed like a careful naming scheme that someone in the laboratory would apply once, and I thought the real informatics questions lived elsewhere, in decision support or exchange.

The Case

The seminar case came from a composite community hospital that had just mapped its laboratory interface to LOINC so it could send results to a regional quality program. Its laboratory system used one local code, GLU, for every glucose result, whether the specimen was a venous blood draw run on the chemistry analyzer or a fingerstick read by a bedside meter on the unit. The mapping team matched GLU to the LOINC term for glucose measured in serum or plasma, because that is what the laboratory's main glucose test is. LOINC, however, distinguishes the specimen as one of its six axes, and glucose in capillary blood measured by a meter has its own term (Regenstrief Institute, 2025). After the map went live, the quality program's report of inpatients with a glucose above 180 rose from about 210 patients a month to about 400. No patient was sicker. Bedside readings, taken four or more times a day on every insulin patient, had joined the serum results under one term. The map did exactly what it was told; the local code had already merged two different observations, and the map carried that merge into every report downstream.

The Analyst's View

A laboratory information systems analyst in our section explained why the merge mattered beyond the count. Meter readings and laboratory values are not interchangeable: meters are less accurate at high and low extremes and can be affected by things like poor circulation in the fingertip. Clinical rules built on serum glucose, such as alerts for severe hyperglycemia, would now fire on meter readings too. She also pointed out that studies of mapping in large institutions have found errors even when experienced staff do the work, often in exactly the specimen and method axes (Lin et al., 2010). Her laboratory reviews a sample of mapped codes every quarter for that reason.

How the Error Was Found

The jump in the report was not caught by anyone in informatics. A diabetes educator noticed that the quality program's figures for her unit no longer matched what she saw on rounds and asked the laboratory why. The laboratory manager pulled a sample of results counted as high and found that almost three quarters came from bedside meters. It had taken seven weeks, and the hospital had already submitted one month of inflated data to the regional program, which then had to be withdrawn and resubmitted. I found that detail sobering. My idea of mapping as a one-time task assumed that anyone would notice a bad map quickly. In practice, the error looked like a clinical trend, and the people best placed to notice it were clinicians, not the team that built the map.

Mapping as a Judgment About Meaning

I had seen mapping as translating a name. The case showed me it is a judgment about meaning, and that a local code may already hold more than one meaning before anyone maps it. The right fix was not a better LOINC choice for GLU; it was splitting GLU into two local codes at the source, one for laboratory glucose and one for bedside meter readings, and mapping each to its own LOINC term. That fix sits in the laboratory system and the bedside device interface, not in the mapping table. I also understand now why mapping cannot be one-time. Laboratories add analyzers, change methods and adopt new point-of-care devices, and each change can quietly alter what a local code means. A map made correctly last year can be wrong this year without anyone touching it.

Checks I Now Run Before Accepting a Map

In my own capstone work on result reporting, I will check every local code for the specimen and method it actually receives, not the one its name suggests, before accepting a map. I will ask the laboratory whether point-of-care results share codes with laboratory results. And I will build a quarterly review of a sample of mapped codes into any plan I write, with the laboratory director accountable for it. Reflection is meant to surface assumptions hidden in one's own reasoning (Rolfe et al., 2001), and mine was that a code's name tells you what it holds.

References

Lin, M. C., Vreeman, D. J., McDonald, C. J., & Huff, S. M. (2010). Correctness of voluntary LOINC mapping for laboratory tests in three large institutions. AMIA Annual Symposium Proceedings, 2010, 447-451.

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

Rolfe, G., Freshwater, D., & Jasper, M. (2001). Critical reflection for nursing and the helping professions: A user's guide. Palgrave Macmillan.