NUR 550 Module 8 Evaluation Plan Example

Evaluation Plan: Measuring Whether a Delirium Prevention Bundle Was Delivered and Whether It Worked

[Student Name]

Southern New Hampshire University

NUR 550: Evidence-Based Practice and Scholarly Inquiry

Evaluation Plan

[Instructor Name]

[Date]

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

Evaluation Plan: Measuring Whether a Delirium Prevention Bundle Was Delivered and Whether It Worked

An evaluation has to answer two questions that are easily blurred: was the intervention delivered as intended, and did it produce the expected result? A bundle that fails because nurses could not deliver it on nights needs a different fix from one delivered faithfully that did not reduce delirium. This plan evaluates the delirium prevention bundle for the unit's patients aged 70 or more. It uses the RE-AIM framework to measure delivery and effect separately, corrects a baseline that would otherwise exaggerate delirium's decline and adds a comparison unit and balancing measures so the team can judge whether any change came from the bundle and whether it caused harm elsewhere.

Why RE-AIM

RE-AIM evaluates an intervention across five dimensions: reach into the intended population, effectiveness, adoption by the settings and staff expected to use it, implementation as intended and maintenance over time (Glasgow et al., 1999). It was developed because studies judged only by effectiveness often failed when programs moved into ordinary practice, which is exactly this project's concern. It also complements i-PARIHS, used for planning, by giving the team measurable questions for each part of the change.

Measures by Dimension

Table 1 defines the main measure for each RE-AIM dimension, its data source and its target.

Table 1

RE-AIM Measures, Data Sources and Targets

The Baseline Trap

The unit's earlier audit found acute confusion documented in about one older patient in six, based on chart review. Comparing that figure with delirium detected by daily 4AT screening after the bundle starts would be misleading, because the two methods detect delirium differently. Chart review misses cases that nobody documented, while daily screening with a tool of 0.88 sensitivity finds many more (Tieges et al., 2021). A post-implementation rate could therefore rise even if the bundle worked, simply because detection improved.

The fix is to measure the baseline with the same method. For four weeks before the bundle begins, all eligible patients on the unit will be screened daily with the 4AT, with no other changes to care. That screening baseline, not the chart audit, will be the comparison for effectiveness.

Design: Comparison Unit

A before-and-after comparison on one unit cannot rule out changes spreading through the whole hospital during the project, for instance a new sedation policy or a change in the patient mix. To address this, 5 East, a similar medical unit not implementing the bundle, will also screen patients 70 and older daily with the 4AT during the baseline and for the first 16 weeks after implementation, with the same training for its nurses. If delirium falls on both units, the team will not credit the bundle with the change. The design is not a randomized trial, and differences between the units will be reported, but it is considerably stronger than a single-unit comparison.

Balancing Measures

Every change can cause harm somewhere else, so four balancing measures will be tracked. Falls during assisted walks will be reported separately, since more mobility means more opportunity to fall. Use of physical restraints and of as-needed antipsychotics will be monitored, in case removing default sleep medicines leads to other sedatives being used for agitation. Nursing overtime will be tracked to see whether the bundle adds unsustainable workload. And a single sleep-quality question will be asked of alert patients each morning, since the sleep routine should improve sleep, not disturb it.

Data Collection and Quality

Most data come from fields already built into the electronic record for the project, which limits extra work. The clinical nurse specialist will run a weekly report of bundle completion, 4AT completion and positive screens, and a unit clerk will cross-check the list of eligible patients against the admission census to catch anyone missed. Once a month, the specialist will audit ten randomly chosen patient-days against the bedside, checking that a documented walk or sleep routine actually happened, because documentation can drift away from practice when a measure becomes a target. Any audit that finds documentation without delivery on more than two of ten patient-days will trigger a conversation with the shift's champion rather than a disciplinary response. Keeping the audit supportive matters, since staff who fear blame tend to document more and report less.

Analysis and Reporting

Weekly bundle completion, 4AT completion and delirium incidence will be plotted on run charts, so the team can see trends and apply standard rules to distinguish real change from ordinary variation (Provost & Murray, 2011). At 16 weeks, delirium incidence will be compared with the screening baseline on each unit, reporting the absolute and relative difference with a 95% confidence interval, and the change on the bundle unit will be compared with any change on 5 East. Falls will be compared using the unit's falls per thousand occupied bed-days. Results will be interpreted in light of the evidence, which suggests that a reduction of about 40% is achievable under trial conditions (Burton et al., 2021); a smaller reduction with high completion would suggest the adapted bundle is weaker than the original, while a smaller reduction with low completion would point to a delivery problem.

Data will be reported to staff weekly by shift, to the unit's leadership monthly and to the hospital's age-friendly steering group at 16 weeks, 6 months and 12 months.

Conclusion

This plan measures reach, effectiveness, adoption, implementation and maintenance separately, so the team will know not only whether delirium fell but why. It replaces a chart-based baseline with a four-week 4AT baseline to avoid mistaking better detection for more delirium, adds a comparison unit to separate the bundle's effect from hospital-wide change and watches four balancing measures for harm. The final project will combine this evaluation with the evidence and plan into one proposal.

References

Burton, J. K., Craig, L. E., Yong, S. Q., Siddiqi, N., Teale, E. A., Woodhouse, R., Barugh, A. J., Shepherd, A. M., Brunton, A., Freeman, S. C., Sutton, A. J., & Quinn, T. J. (2021). Non-pharmacological interventions for preventing delirium in hospitalised non-ICU patients. Cochrane Database of Systematic Reviews, 2021(7), Article CD013307. https://doi.org/10.1002/14651858.CD013307.pub2

Glasgow, R. E., Vogt, T. M., & Boles, S. M. (1999). Evaluating the public health impact of health promotion interventions: The RE-AIM framework. American Journal of Public Health, 89(9), 1322-1327. https://doi.org/10.2105/AJPH.89.9.1322

Provost, L. P., & Murray, S. K. (2011). The health care data guide: Learning from data for improvement. Jossey-Bass.

Tieges, Z., MacLullich, A. M. J., Anand, A., Brookes, C., Cassarino, M., O'Connor, M., Ryan, D., Saller, T., Arora, R. C., Chang, Y., Agarwal, K., Taffet, G., Quinn, T., Shenkin, S. D., & Galvin, R. (2021). Diagnostic accuracy of the 4AT for delirium detection in older adults: Systematic review and meta-analysis. Age and Ageing, 50(3), 733-743. https://doi.org/10.1093/ageing/afaa224