HI595 Unit 8 Seminar Reflection Example

Worked From the Document Alone: A Unit 8 Capstone Seminar Reflection on Two Sentences I Believed Were Clear

[Student Name]

Purdue University Global

HI595: Master's Capstone in Health Informatics

Unit 8 Seminar Reflection

[Instructor Name]

[Date]

Carrowby Health, the classmate, the test cases and every detail are composite; the sources cited are real.

The Exercise

Our instructor asked each of us to give our draft deliverable to a classmate, along with three sample cases, and to say nothing while the classmate worked through them. My reader was a health information supervisor at a clinic network, careful and experienced, but outside radiology. She had version 0.2 of my specification and twenty minutes.

Case One: Worked Cleanly

A patient with a pacemaker implanted at Carrowby two years ago, scheduled for a lumbar spine MRI. The device identifier was in the operating room log, and the entry showed the manufacturer, model and MR Conditional labeling. She found the entry, set the screening status to Verified, noted that a radiologist would need to review the conditions, and moved on. The specification did what it was meant to do.

Case Two: Stalled

A patient with a cochlear implant placed at an outside hospital, whose only documentation was a phone photo of a card with the model number partly cut off. She searched the exception rules and found two: one for outside implants with a card, and one for outside implants without documentation. This case was neither. She asked aloud whether a partial card counted as a card, and I said nothing, as instructed. After four minutes she set the status to Hold, which was defensible, but she had no instruction for what to do next.

The draft told her to transcribe a card when one existed. It never said what counted as one. I had pictured a whole, legible card. She read the rule as I wrote it, and it gave her nothing for a partial one. Her choice of Hold was a reasonable reading of an incomplete rule, not an error.

Case Three: Closed the Wrong Way

A patient with a programmable shunt, implanted at Carrowby, with an entry showing MR Conditional labeling and a link to the manufacturer's conditions. She confirmed the entry, then set the screening status to Cleared and closed the case.

That was the wrong status. In my mind, Verified meant the device was identified and Cleared meant a radiologist had approved scanning under the device's conditions, which for a programmable shunt include checking the valve setting after the scan. But version 0.2 defined the statuses by name only. To a careful reader from another field, Cleared plainly meant "checked and fine," which is exactly what she had done.

I had written "Cleared" as if the word explained itself, and a careful reader showed me that it explained the opposite of what I meant.

Why the Exercise Worked

Neither failure would have shown up if I had walked her through the document. I would have explained the statuses as I explained them to myself, and she would have nodded. Working from the text alone exposed the gap between what I knew and what I had written. Requirements writers warn that terms which seem obvious to their authors are a common source of ambiguity (Wiegers & Beatty, 2013), and I now understand that warning from the receiving end. I also noticed something about myself: my first impulse, watching her stall on case two, was to explain, and the urge was strongest exactly where the document was weakest. That impulse is a useful signal for where to revise.

What Changed

Version 0.3 now defines every status with an entry condition and an exit condition. A case now becomes Verified once someone has pinned down what the device is and how it is labeled, and it becomes Cleared only after a supervising radiologist approves scanning under the stated conditions. The specification adds a sentence explaining why the two are kept apart. A new exception rule covers partial documentation: record what is legible, set the labeling to Unknown, and request confirmation from the manufacturer or implanting hospital. Radiology guidance treats an unidentified device as unsafe until proven otherwise (Harwood et al., 2024), and the new rule makes that explicit for partial cards.

What I Am Less Sure Of

I wonder how many other terms in the specification only I understand. The exercise tested two statuses and one exception rule out of a document with many more. I plan to ask a screening nurse and the informatics analyst to repeat the exercise with different cases before version 1.0. Improvement guidance encourages testing changes on a small scale with the people who will use them before spreading them (Langley et al., 2009), and a document is a change like any other.

References

Harwood, M., Fahrenholtz, S. J., Wellnitz, C. V., Kawashima, A., & Panda, A. (2024). MRI in adult patients with active and inactive implanted MR-conditional, MR-nonconditional, and other devices. RadioGraphics, 44(3), e230102. https://doi.org/10.1148/rg.230102

Langley, G. J., Moen, R. D., Nolan, K. M., Nolan, T. W., Norman, C. L., & Provost, L. P. (2009). The improvement guide: A practical approach to enhancing organizational performance (2nd ed.). Jossey-Bass.

Wiegers, K., & Beatty, J. (2013). Software requirements (3rd ed.). Microsoft Press.