NU802 Unit 3 DNP Evidence Synthesis Example

Strong for Teaching, Thin for the Phone Call: A Unit 3 Evidence Synthesis for the Five Parts of an Insulin Discharge Bundle

[Student Name]

Purdue University Global

NU802: DNP Project I

Unit 3 Evidence Synthesis

[Instructor Name]

[Date]

The project and hospital are composite; every study appraised is real and is described from its published report.

Question and Search

The practice question is whether, for adults who begin insulin while in the hospital, a bundle combining a watched first self-injection on the pen they will use at home, a one-page plan in everyday words, that pen dispensed before they leave, contact within three days and a link to outpatient education reduces returns for glucose problems within 30 days, compared with usual discharge. I searched three databases, CINAHL, PubMed and Cochrane, for work published from 2010 through 2025, combining terms for insulin, diabetes, discharge, transition, education, teach-back, follow-up and readmission. The search found no trial of the bundle as a whole, so it was widened to evidence for each component. Sources were graded with the Johns Hopkins hierarchy, from level I for systematic reviews of randomized trials to level V for expert opinion, and rated for quality (Dang et al., 2022).

The Evidence for Each Part

Inpatient education and the supervised injection. Wexler et al. (2012) randomized 31 adults hospitalized for reasons other than diabetes, all with type 2 diabetes and an A1C over 7.5 percent, to inpatient diabetes management with education and an improved handoff to outpatient care, or to usual care. Across all participants, the difference in A1C at one year was small and not significant. Among patients newly discharged on insulin, however, A1C fell 2.4 percentage points more in the intervention group, a significant difference, while patients already on insulin did not benefit. This is level II evidence of modest quality: a randomized design, but a very small sample and a subgroup finding. Healy et al. (2013), in a retrospective cohort of patients with A1C above 9 percent, found that those who received inpatient diabetes education were readmitted within 30 days less often, with an adjusted odds ratio of 0.66. This is level III evidence of good quality, limited by the possibility that patients who were seen by an educator were healthier or more engaged in ways the statistical adjustment could not see.

Teaching method. Talevski et al. (2020) reviewed 20 studies of teach-back, a check in which patients restate or perform, in their own words and hands, what they have just been shown. Teach-back was effective in 19 of them, on outcomes from knowledge and self-care to readmission and quality of life, although only half the studies reported how teach-back was put into practice. This is level III evidence, a systematic review of mixed designs, and it supports making the supervised injection a demonstration by the patient, not by the nurse.

The best evidence supports the part of the bundle nurses already value most: the patient doing the injection while a nurse watches.

Pen in hand before discharge. Stedge et al. (2023) looked back over the first year of an Atlanta hospitalist service's program that brought filled prescriptions to patients' rooms before they left. Readmission fell in some groups, most clearly among patients with no comorbidities and those with a heavy comorbidity burden, and the effect was larger when prescriptions were subsidized for uninsured and underinsured patients. In the middle comorbidity group, delivery without a subsidy was associated with more readmissions, not fewer. This is level III evidence of fair quality from a retrospective design, and its mixed results suggest that delivery helps most when cost is also addressed.

Contact after discharge. Jayakody et al. (2016) systematically reviewed interventions using telephone follow-up to reduce 30-day readmission in cardiovascular disease, chronic respiratory disease and diabetes. Of ten studies, five reduced readmission, and every one combined the call with other parts, most often support before discharge. Study quality was generally poor, and the authors judged the evidence inconclusive. This is level II evidence of low quality. It suggests that the three-day check-in should not be expected to work alone, and it supports keeping it tied to the teaching and planning that come before discharge.

Connection to diabetes education. Drincic et al. (2017) compared 30-day readmission for patients with diabetes in the 18 months before and after a tiered program that expanded diabetes educators' work into case management and trained staff nurses as diabetes resource nurses. Across more than 66,000 discharges, readmission fell from 20.1 to 17.6 percent, and the patients educators saw stayed the group least often readmitted throughout. This is level IV evidence, a before-and-after comparison, of good quality given its size, though other changes in the hospital over the same period could explain part of the drop.

National standard. The American Diabetes Association's hospital standards recommend structured discharge planning that starts at admission, education in survival skills such as insulin use and recognizing lows and follow-up arranged before discharge (American Diabetes Association Professional Practice Committee, 2025). As a consensus guideline grounded in evidence review, it supports every part of the bundle in principle but does not show which parts matter most.

Reading the Evidence Together

Read together, the sources point in the same direction, and none points against the bundle. The strongest support is for inpatient education delivered to patients new to insulin, with teach-back as the method, where a small trial, a large cohort and a systematic review agree. Support for dispensing the pen before discharge is moderate and conditional on cost. Support for follow-up after discharge is weak when the follow-up stands alone but stronger when it is part of a package that begins in the hospital. Support for connecting patients to diabetes educators comes from a large before-and-after study and from the national standard.

The overall strength of evidence for the bundle is therefore moderate. No study tested these five parts together, and the evidence cannot say which part produces most of the benefit. That gap is a reason for the project's design: the bundle is implemented as a whole because the evidence supports a multi-part transition, and process measures for each part let the project describe which parts reached which patients.

What the Evidence Changes in the Proposal

Three changes follow. First, the supervised injection will use teach-back explicitly, with the patient demonstrating the full dose on the pen going home, as the strongest evidence suggests. Second, the pen-in-hand part will be paired with the pharmacist's early start on coverage and assistance for uninsured and Medicaid patients, since delivery without help with cost had mixed results. Third, the proposal will describe the three-day check-in as part of a package rather than a stand-alone safety net and will not claim that it reduces returns on its own.

Limits of This Synthesis

The search was limited to English-language studies and three databases, and one reviewer selected and graded the studies, which a second reviewer would strengthen. Most studies came from academic centers larger than this hospital. Two of the six are small or retrospective. The synthesis therefore supports implementing the bundle as a quality improvement and evaluating it carefully, not claiming in advance what size of effect it will have.

Conclusion

The bundle is supported by moderate evidence overall, strongest for teaching patients who are new to insulin and weakest for follow-up calls standing alone. The evidence argues for a bundle delivered as a whole, with teach-back at its center, cost addressed alongside delivery and follow-up linked to what happens before discharge. Those conclusions shape the methods section that follows.

References

American Diabetes Association Professional Practice Committee. (2025). 16. Diabetes care in the hospital: Standards of care in diabetes-2025. Diabetes Care, 48(Suppl. 1), S321-S334. https://doi.org/10.2337/dc25-S016

Dang, D., Dearholt, S. L., Bissett, K., Ascenzi, J., & Whalen, M. (2022). Johns Hopkins evidence-based practice for nurses and healthcare professionals: Model and guidelines (4th ed.). Sigma Theta Tau International.

Drincic, A., Pfeffer, E., Luo, J., & Goldner, W. S. (2017). The effect of diabetes case management and Diabetes Resource Nurse program on readmissions of patients with diabetes mellitus. Journal of Clinical & Translational Endocrinology, 8, 29-34. https://doi.org/10.1016/j.jcte.2017.03.003

Healy, S. J., Black, D., Harris, C., Lorenz, A., & Dungan, K. M. (2013). Inpatient diabetes education is associated with less frequent hospital readmission among patients with poor glycemic control. Diabetes Care, 36(10), 2960-2967. https://doi.org/10.2337/dc13-0108

Jayakody, A., Bryant, J., Carey, M., Hobden, B., Dodd, N., & Sanson-Fisher, R. (2016). Effectiveness of interventions utilising telephone follow up in reducing hospital readmission within 30 days for individuals with chronic disease: A systematic review. BMC Health Services Research, 16, 403. https://doi.org/10.1186/s12913-016-1650-9

Stedge, B., Xu, J., Kubes, J. N., Han, E., Russie, C., Woods, L., & O'Donnell, C. (2023). Meds to beds at hospital discharge improves medication adherence and readmission rates in select populations. Southern Medical Journal, 116(3), 247-254. https://doi.org/10.14423/SMJ.0000000000001532

Talevski, J., Wong Shee, A., Rasmussen, B., Kemp, G., & Beauchamp, A. (2020). Teach-back: A systematic review of implementation and impacts. PLOS ONE, 15(4), e0231350. https://doi.org/10.1371/journal.pone.0231350

Wexler, D. J., Beauharnais, C. C., Regan, S., Nathan, D. M., Cagliero, E., & Larkin, M. E. (2012). Impact of inpatient diabetes management, education, and improved discharge transition on glycemic control 12 months after discharge. Diabetes Research and Clinical Practice, 98(2), 249-256. https://doi.org/10.1016/j.diabres.2012.09.016