NR 515 Week 3 Data Exchange Analysis, One Result Example

Data Exchange Analysis: One Potassium of 6.1 From a Reference Laboratory to a Physician's Inbox and a Patient's Phone, and the Exact Point Where a Hemolysis Comment Stopped Traveling

Student Name

College of Nursing, Chamberlain University

NR 515: Informatics in Healthcare

Instructor Name

Month Day, Year

Data Exchange Analysis: One Potassium of 6.1 From a Reference Laboratory to a Physician's Inbox and a Patient's Phone, and the Exact Point Where a Hemolysis Comment Stopped Traveling

The Item

The item is a single serum potassium result for a 64-year-old woman seen at our six-provider family practice for a routine blood pressure follow-up. Her blood was drawn in the practice's draw station on a Monday afternoon, held for courier pickup, and run the next morning at the reference laboratory 60 miles away. The laboratory reported a potassium of 6.1 mmol/L with a comment that the specimen was moderately hemolyzed. What the laboratory sent and what our physician and the patient saw were not the same, and this paper locates where they diverged. The case is described without identifying details.

The Route

The result passed through four parties before anyone read it:

1. The reference laboratory's information system, which sent the result as an HL7 version 2 observation result message (ORU^R01) over a secure connection.

2. The interface engine hosted by our EHR vendor, which receives messages from several laboratories, maps their fields to the EHR's data model and passes them on. The vendor configures and runs it under our contract.

3. Our EHR, which files the result to the patient's chart and places it in the ordering physician's results inbox.

4. The patient's phone, where a personal health app retrieves results from our EHR through its standards-based patient access interface, which uses FHIR. Bender and Sartipi (2013) described FHIR as a resource-based, web-oriented approach to exchange designed to be simpler to implement than earlier HL7 standards, and the Cures Act rules require certified systems to offer this kind of access to patients (Office of the National Coordinator for Health Information Technology, 2020).

Each hand-off is a different party's responsibility, and no single party sees the whole route.

What Left and What Arrived

Table 1 compares the fields the laboratory sent with what each reader was shown.

Table 1

Field-Level Comparison of One Potassium Result Along Its Route

Note. The comparison was made by the practice's informatics lead with the EHR vendor's analyst, using the raw message held by the interface engine.

Where the Two Stopped Agreeing

The value, units, range and flag arrived intact at every stop. The comment did not reach either reader's screen. Tracing the raw message with the vendor's analyst showed that the interface engine did carry the comment: it mapped the NTE segment into a result-level comment field in the EHR. The comment was in the chart. The divergence happened one step later, at the display. The physician's results inbox shows a single line per result, value, flag and range, and shows comments only when the user opens the full report. The patient app received the observation without the comment because the EHR's patient access interface did not include the result-level comment field in what it returned.

So the exact point of divergence was the EHR's inbox display configuration, and a second, separate point was its patient access mapping. The laboratory did its job, and so did the interface engine. The meaning of the result changed without any system making an error in the usual sense.

The method matters for anyone repeating this analysis. Looking only at screens would have suggested that the laboratory never sent the comment, and the first complaint our practice drafted said exactly that. Only the raw HL7 message, retained by the interface engine for thirty days, showed that the comment had left the laboratory and arrived in our system. Without that record, the practice would have blamed the wrong party and asked the laboratory to change something it was already doing correctly. Anyone tracing a result should ask for the raw message first and the screenshots second.

Why the Comment Mattered

Hemolysis releases potassium from red cells into the specimen, raising the measured value without any change in the patient. Lippi et al. (2008) described hemolysis as the leading cause of unsuitable specimens in clinical laboratories. A potassium of 6.1 in a patient taking lisinopril could reasonably prompt an urgent call and a trip to the emergency department. With the comment, it prompts a redraw. The physician saw the red flag, called the patient at 0845 and advised her to go to the emergency department; the patient had already seen "High" on her phone and was frightened. At the emergency department, a repeat potassium was 4.6 mmol/L. No harm came to her beyond a morning of worry and an emergency visit that the laboratory's own comment had advised against needing.

This kind of gap belongs to a broader problem. Callen et al. (2012) reviewed studies of test result follow-up in ambulatory care and found wide variation in how many results were not followed up, and few guidelines about who was responsible for notification. Here the result was followed up promptly; it was the context around it that went missing.

Who Owns Each Fix

The fix at the inbox belongs to our practice, since the display is a configuration choice. The EHR allows a comment indicator on the inbox line, and our office manager can turn it on with the vendor's help. The fix at the patient access interface belongs to the vendor, whose mapping decides which fields the app receives; we have asked for the comment to be included. The laboratory could also help by adding a hemolysis flag in a coded field, not only in free text, so that it survives displays that suppress comments. Until those changes are made, our nurses have been asked to open the full report for any critical potassium before calling a patient.

The analysis also produced a rule for the practice's own reviews: when a result looks alarming and out of keeping with the patient's history, the first question should be whether the laboratory said anything about the specimen. That question costs one click and would have saved this patient her morning in the emergency department.

References

Bender, D., & Sartipi, K. (2013). HL7 FHIR: An agile and RESTful approach to healthcare information exchange. In Proceedings of the 26th IEEE International Symposium on Computer-Based Medical Systems (pp. 326-331). IEEE. https://doi.org/10.1109/CBMS.2013.6627810

Callen, J. L., Westbrook, J. I., Georgiou, A., & Li, J. (2012). Failure to follow-up test results for ambulatory patients: A systematic review. Journal of General Internal Medicine, 27(10), 1334-1348. https://doi.org/10.1007/s11606-011-1949-5

Lippi, G., Blanckaert, N., Bonini, P., Green, S., Kitchen, S., Palicka, V., Vassault, A. J., & Plebani, M. (2008). Haemolysis: An overview of the leading cause of unsuitable specimens in clinical laboratories. Clinical Chemistry and Laboratory Medicine, 46(6), 764-772. https://doi.org/10.1515/CCLM.2008.170

Office of the National Coordinator for Health Information Technology. (2020, May 1). 21st Century Cures Act: Interoperability, information blocking, and the ONC Health IT Certification Program. Federal Register, 85, 25642-25961. https://www.federalregister.gov/documents/2020/05/01/2020-07419/21st-century-cures-act-interoperability-information-blocking-and-the-onc-health-it-certification