NU804 Unit 3 Implementation Fidelity Report Example

Thirteen of Twenty Got Every Part: A Unit 3 Fidelity Monitoring Report on an Insulin Discharge Bundle in Its First Nine Weeks

[Student Name]

Purdue University Global

NU804: DNP Project II

Unit 3 Fidelity Monitoring Report

[Instructor Name]

[Date]

The hospital, units, patients and audit results are composite and illustrative; the frameworks cited are real.

What Fidelity Means for This Bundle

Carroll et al. (2007) describe implementation fidelity mainly as adherence: whether each piece of an intervention reaches every person it was built for, in the amount and over the period the designers had in mind. Their framework adds factors that raise or lower adherence: how complicated the intervention is, what help the people delivering it receive, how well they deliver it and how the people receiving it react. For a five-part bundle, adherence is measured part by part, and the all-or-none measure, every part for every patient, is the strictest test.

Fidelity here applies to the bundle's core functions, not to every detail of how they are delivered. The sustainability plan written in the quality improvement course separates the five core functions from forms that units may adapt, such as the time of day teaching happens or whether a family member gives the supervised dose. A unit that changes a form is not drifting; a unit that drops a function is.

How Fidelity Was Measured

Every patient discharged on new insulin from the four units was included. For each, the analyst pulled the five discharge fields, the pharmacy dispensing time and the text platform record. The resource nurses audited every chart, rather than the planned five a month, because volumes were still small, using the audit tool's decision rules to confirm that each field matched the notes.

Results by Component

Supervised self-injection on the pen going home. Seventeen of 20 patients, 85 percent. Two of the misses were weekend starters taught on a trainer pen before the inpatient pharmacy's weekend take-home pens began in week four. The third was a patient discharged at 0830 before the day nurse had seen him.

Plain-language plan. Nineteen of 20 had a plan in the chart. The audit, however, found a problem the field could not show, described below.

Pen by the first insulin day. Fourteen of 20, 70 percent. Besides the two weekend starters, four patients had Medicaid managed care plans that required prior authorization for the pen brand, and approval came after the first insulin day.

Contact by day three. Eighteen of 20 had a text reply or call. The two misses had no reply to texts and could not be reached by phone until day four.

Ongoing education. All 20 were offered a class seat; seven accepted, and four have attended so far.

All five parts were delivered to 13 of the 20 patients, 65 percent, against a target of 85 percent.

A field can be complete and the care still wrong; the audit found that only by reading what the plan said.

The Drift That Mattered Most

When the resource nurses read the plans rather than checking that one existed, they found that three of the 19 listed doses that differed from the final discharge order. In each case the plan had been printed in the morning, the hospitalist had changed a dose at the discharge visit and the nurse had corrected the printed plan by hand for one dose but not another. One patient's plan listed a basal dose of 24 units when the final order was 18. The day-one text reply showed no low, and the educator corrected the plan by phone on day one, but the error could have caused harm.

This is a drift in quality of delivery rather than in adherence: the plan was present but not accurate. It was reported through the hospital's event system as a near miss and reviewed with both unit managers.

Responses

1. Plan accuracy. The plan is now generated from the final discharge medication reconciliation rather than from the active orders, and the nurse reads the doses aloud against the discharge order with the patient as part of teach-back. The audit tool has a new item: plan doses match the final order.

2. Weekend pens. The inpatient pharmacy's weekend take-home pens, started in week four, will be reported separately so their effect on the pen measure can be seen.

3. Medicaid prior authorization. The pharmacist now starts prior authorization at the first insulin order, as the proposal planned but the first weeks did not consistently do, and a daily list of new insulin starters goes to pharmacy at 0900.

4. Early discharges. On 3 West, discharges before 1000 for patients on new insulin will be flagged at the evening huddle so the supervised dose happens the night before.

5. Unreached patients. After a missed day-one reply and a missed call, the nurse now calls the emergency contact the patient named at discharge.

Classifying the Changes

Wiltsey Stirman et al. (2019) offer a framework for reporting modifications to an intervention, including whether each was planned or made in response to a problem, who decided it, what level it affects and whether it is consistent with fidelity to the core functions. Using that framework, all five responses were unplanned and reactive, decided by the student with the practice mentor and the relevant department, and all are fidelity-consistent: they protect a core function rather than removing one. Each is entered in the project's change log with its date, so that the evaluation in NU806 can read results in light of when each change began.

What the Numbers Can Say

Twenty patients is too few for confident rates, and the four units started at different times. The 65 percent all-or-none figure is a starting point, not a verdict. The pattern, though, is clear enough to act on: the weak components are the ones that depend on systems outside nursing, pharmacy timing and insurance, while the components nurses control directly, teaching and plans, are delivered to most patients. The one accuracy failure in the plan shows that nurse-controlled parts need checking too.

Chambers et al. (2013) argue that implementation should expect continuous refinement as an intervention meets its setting, rather than treating any change as decline. The first nine weeks fit that view: the bundle is being improved by what the audits find.

Next Audit

The next report, at week 14, will cover about 35 patients across the four units. Targets are 80 percent all-or-none, 90 percent pen by the first insulin day and zero plan errors on audit.

Conclusion

In its first nine weeks on four units, the bundle reached most patients with teaching and plans, fewer with the pen on time and 65 percent with every part. A plan-accuracy drift found only by reading the plans is the most important finding, and it has been fixed at its source. Five fidelity-consistent changes are logged for the evaluation, and the next report will show whether they work.

References

Carroll, C., Patterson, M., Wood, S., Booth, A., Rick, J., & Balain, S. (2007). A conceptual framework for implementation fidelity. Implementation Science, 2, 40. https://doi.org/10.1186/1748-5908-2-40

Chambers, D. A., Glasgow, R. E., & Stange, K. C. (2013). The dynamic sustainability framework: Addressing the paradox of sustainment amid ongoing change. Implementation Science, 8, 117. https://doi.org/10.1186/1748-5908-8-117

Wiltsey Stirman, S., Baumann, A. A., & Miller, C. J. (2019). The FRAME: An expanded framework for reporting adaptations and modifications to evidence-based interventions. Implementation Science, 14, 58. https://doi.org/10.1186/s13012-019-0898-y