NURS 6051 Week 8 Informatics Project Proposal Example

Joining Three Timestamps: An Informatics Project Proposal to Test Whether Meal Tray Timing Explains Hypoglycemia on a Medical Unit

Student Name

College of Nursing, Walden University

NURS 6051: Transforming Nursing and Healthcare Through Technology

Instructor Name

Month Day, Year

Joining Three Timestamps: An Informatics Project Proposal to Test Whether Meal Tray Timing Explains Hypoglycemia on a Medical Unit

Problem Statement

On a composite 28-bed medical unit, nurses have noticed that patients receiving mealtime insulin seem to have more low blood glucose readings after lunch than after other meals. Several nurses suspect a timing problem: insulin is given when the glucose check is done, but lunch trays sometimes arrive much later because of the dietary department's delivery route. When rapid-acting insulin is given and the meal is delayed, the insulin can lower glucose before food is absorbed, which is a recognized cause of hypoglycemia in hospitalized patients (Moghissi et al., 2009). Current guidance recommends coordinating mealtime insulin with the actual delivery of meals (American Diabetes Association Professional Practice Committee [ADA], 2024).

The suspicion is plausible, but it is still a suspicion. The unit does not know whether lunch trays are actually later than other trays, whether insulin is actually given well before trays arrive, or whether those gaps are associated with low readings. The purpose of this project is to use data the hospital already collects to confirm or dismiss the hunch before any practice change is made.

Data Elements

Three data elements, held in three different systems, are needed.

Point-of-care glucose results: value, time, and patient identifier, held in the laboratory information system through the glucose meter interface. Pulled by the laboratory's point-of-care coordinator.

Meal tray delivery times: time each tray is scanned as delivered to the room, with patient identifier and meal, held in the dietary department's meal ordering system. Pulled by the dietary supervisor.

Insulin administration times: drug, dose type (mealtime versus basal or correction), time scanned at the bedside, and patient identifier, held in the electronic medication administration record. Pulled by the clinical informatics analyst.

The informatics work is joining these three sources by patient and time for every insulin-treated patient over three months, then calculating, for each meal, the interval between insulin administration and tray delivery and whether a glucose reading below the hospital's hypoglycemia threshold occurred in the following hours.

Stakeholders

Charge nurses: they see the problem on the unit and will need to change how insulin and trays are coordinated if the hunch is confirmed. Their interest is workflow that is realistic on a busy day.

Diabetes educator: brings clinical knowledge of insulin action and hypoglycemia risk, and will help interpret whether observed intervals are clinically meaningful.

Pharmacist: understands insulin types and timing, and can advise whether some patients are on regimens where timing matters less.

Dietary supervisor: controls tray delivery routes and times and holds one of the three data sets; any solution involving delivery timing depends on this role.

Nurse informaticist: translates the clinical question into a data request, knows which systems hold which data, and can build the joined report (American Nurses Association [ANA], 2022).

The Informaticist as Translator

The nurse informaticist's role in this project is translation in both directions. The clinical team asks a question in clinical language: are we giving insulin too long before lunch arrives? The informaticist turns that into a technical specification: extract barcode administration events for mealtime insulin, match them by patient and date to tray scan events for the nearest subsequent meal, and flag glucose results below a threshold within a defined window after the insulin. Then the informaticist translates the results back: a table of intervals by meal, a comparison of hypoglycemia rates when the interval is short versus long, and a note on data limitations, such as trays scanned late or insulin given without a scan. Without that translation, the clinical team would have a question and three databases it cannot join, and the analysts would have data without a question (McGonigle & Mastrian, 2022).

Technology Needed

No new software is needed. The project uses existing systems and a joined report built by the informatics analyst in the hospital's reporting tool, refreshed monthly. If the hunch is confirmed and a practice change follows, the same report becomes the monitoring tool for that change. Keeping the technology modest keeps the project within the unit's reach and avoids a purchasing process.

Timeline, Privacy, and Limits of the Data

The project runs over four months. In the first month, the informaticist and the three data owners agree on definitions, such as which insulin orders count as mealtime doses and which tray scan counts as delivery, and the informaticist builds and tests the joined report on one week of data. The second and third months cover the three-month retrospective extract and the analysis. In the fourth month, the results go to the unit practice council with a recommendation.

Because the project joins identifiable data from three systems, it is reviewed by the hospital's privacy officer and, since it is a quality improvement project using existing data for internal purposes, by the quality department's determination process rather than as research. Only the analyst sees identifiers; the council receives aggregated tables.

The data have known limits. Tray scans depend on dietary staff scanning at the bedside rather than at the unit door, insulin scanned late will look late even if it was given on time, and glucose checks happen at scheduled times, so a low that recovers between checks will be missed. The informaticist will estimate how often each problem occurs in the test week and report it with the results, so that the council can judge how much weight the findings will bear.

Expected Outcome and What Would Disprove the Hunch

If the hunch is right, the data will show that intervals between mealtime insulin and tray delivery are longer at lunch than at other meals, and that hypoglycemia is more frequent when the interval is long. That would justify a practice change, such as scanning the tray before giving mealtime insulin, and the expected outcome would be fewer hypoglycemia events after lunch, measured against the unit's own three-month baseline.

The proposal states in advance what would mean the hunch is wrong. If lunch intervals are no longer than other intervals, or if hypoglycemia rates do not differ by interval, then tray timing is not the explanation, and the team would look elsewhere, for example at correction insulin given at lunch or at patients' intake. The project succeeds either way, because the unit will have replaced a suspicion with an answer.

References

American Diabetes Association Professional Practice Committee. (2024). 16. Diabetes care in the hospital: Standards of care in diabetes 2024. Diabetes Care, 47(Suppl. 1), S295-S306. https://doi.org/10.2337/dc24-S016

American Nurses Association. (2022). Nursing informatics: Scope and standards of practice (3rd ed.). Nursesbooks.org.

McGonigle, D., & Mastrian, K. G. (2022). Nursing informatics and the foundation of knowledge (5th ed.). Jones & Bartlett Learning.

Moghissi, E. S., Korytkowski, M. T., DiNardo, M., Einhorn, D., Hellman, R., Hirsch, I. B., Inzucchi, S. E., Ismail-Beigi, F., Kirkman, M. S., & Umpierrez, G. E. (2009). American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Diabetes Care, 32(6), 1119-1131. https://doi.org/10.2337/dc09-9029