DNP880 Module 7 assignment: final DNP project paper: implementation, results, discussion and sustainment, a full sample

From 5 Percent to 88 Percent: Results of a Food Insecurity Screening and Same-Visit Referral Project for Adults With Type 2 Diabetes

Student Name

Doctor of Nursing Practice Program, Aspen University

DNP880: DNP Project

Instructor Name

Month Day, Year

From 5 Percent to 88 Percent: Results of a Food Insecurity Screening and Same-Visit Referral Project for Adults With Type 2 Diabetes

This paper reports the implementation and results of a 16-week quality improvement project in two clinics of Maple Street, a composite federally qualified health center, in which medical assistants asked adults with type 2 diabetes two questions about food insecurity at each visit and community health workers connected patients who screened positive with the regional food bank. The university's review board determined the project to be quality improvement. The report follows the SQUIRE 2.0 guidelines, describing context, the changes as they evolved, results, and limits (Ogrinc et al., 2016). All data presented are composite.

Implementation as It Happened

The plan called for four PDSA cycles, and all four were run, but not exactly as written. In Cycle 1, the medical assistant screened 20 of 21 eligible patients, and rooming time rose by a median of 40 seconds; the only change adopted was to move the questions after the depression screen, where patients were already answering personal questions. In Cycle 2, only 64 percent of positive screens met the community health worker before leaving, because the worker was often with another patient. The team adopted the planned contingency, a phone handoff within 48 hours, and moved the worker's desk next to the rooming area on the two busiest mornings. In Cycle 3, transportation emerged as the leading barrier to reaching the food bank, and the food bank agreed to deliver produce boxes to the clinic on the second Thursday of each month. Cycle 4 spread the workflow to the second clinic in week 12. The final workflow differs from the one proposed in three places, and each change came from a cycle's data rather than from a meeting.

Results: Screening

Screening documentation in the baseline sample was 5 percent at the weekly median. During implementation, 874 of 1,055 eligible visits, or 82.8 percent, included completed screening items or a documented decline; declines accounted for 3 percent of screened visits. At the first clinic, all 13 weekly points after the start of screening fell above the baseline median, meeting the shift rule for nonrandom change (Perla et al., 2011), and the median over the final four weeks was 89 percent. At the second clinic, the median over its final four weeks was 86 percent, above the 85 percent target, but with only five weeks of data it could not produce the six consecutive points the shift rule requires. The target of a shift at both clinics was therefore not formally met at the second clinic, a consequence of the timeline rather than of the data.

Results: Referral and Food Received

Of screened visits, 192, or 22.0 percent, were positive, representing 164 unique patients at their first positive screen. This proportion is close to the 19.1 percent found among adults with diabetes sampled from academic, community health center, and specialty clinics in Massachusetts (Berkowitz et al., 2015). Of the 164 patients, 131 met a community health worker the same day and 11 more were reached by phone within 48 hours, so that 142 were referred. By the two-week call or the food bank's confirmation list, 86 of the 142, or 60.6 percent, had received food, meeting the 60 percent target. Among the 56 who had not, 24 could not be reached, 21 reported transportation problems, 6 cited distribution hours, and 5 chose not to go. After clinic delivery of produce boxes began in week 10, transportation was cited less often at the first clinic, although the numbers are too small to separate this from other changes.

Results: Visit Length and A1c

Median visit length rose by 1.5 minutes in the first three weeks at each clinic and settled at 0.8 minutes above baseline, below the two-minute threshold set as a balancing concern. Among the 164 patients who screened positive, 97 had A1c values both before and during the project. Their mean A1c fell from 9.1 to 8.9 percent, a change that was not statistically significant (paired t test, p = .09). Because follow-up was short, no comparison group existed, and 67 patients had no second value, this is reported as an observation, not an effect. By language, screening rates for patients preferring Spanish, 84 percent, were similar to those preferring English, 83 percent.

Discussion

The project showed that a two-item screen can be added to rooming at a busy community health center with little effect on visit length, and that pairing it with a community health worker and a food bank partner connected most patients who screened positive with food within two weeks. The largest loss in the cascade was not at screening but after the visit, where 17 percent of referred patients could not be reached. That finding is consistent with the evidence synthesis, which judged the link from screening to receiving help less certain than the accuracy of the screen itself, which rests on a large validation study (Hager et al., 2010). The produce box delivery addressed transportation, and a text message reminder before the follow-up call is the next change the team plans to test.

The project's contribution is modest and practical: it adds results from a setting underrepresented in the literature and reports patient comments that were mostly positive, with several patients saying no one had asked before. Its limitations include the uncontrolled design, a short spread period at the second clinic, reliance on patient report for referral completion, and the composite nature of this report.

Sustainment

At week 16, ownership of the workflow passed to the nurse manager. Screening rate is now on the monthly quality dashboard, the referral form remains in the record, the food bank delivery continues under its grant, and a 15-minute module on the screen has been added to medical assistant orientation. The quality committee voted to extend screening to the center's three remaining clinics in the next quarter, using the refined workflow.

Conclusion

Routine food insecurity screening with same-visit referral raised documented screening at two clinics from 5 percent to above 85 percent, connected 60.6 percent of referred patients with food within two weeks, and added less than a minute to visits. A1c change was small and uncertain. The workflow has been adopted for spread, and the next cycles will focus on the patients who were referred but never reached.

References

Berkowitz, S. A., Meigs, J. B., DeWalt, D., Seligman, H. K., Barnard, L. S., Bright, O.-J. M., Schow, M., Atlas, S. J., & Wexler, D. J. (2015). Material need insecurities, control of diabetes mellitus, and use of health care resources: Results of the Measuring Economic Insecurity in Diabetes study. JAMA Internal Medicine, 175(2), 257-265. https://doi.org/10.1001/jamainternmed.2014.6888

Hager, E. R., Quigg, A. M., Black, M. M., Coleman, S. M., Heeren, T., Rose-Jacobs, R., Cook, J. T., de Cuba, S. A. E., Casey, P. H., Chilton, M., Cutts, D. B., Meyers, A. F., & Frank, D. A. (2010). Development and validity of a 2-item screen to identify families at risk for food insecurity. Pediatrics, 126(1), e26-e32. https://doi.org/10.1542/peds.2009-3146

Ogrinc, G., Davies, L., Goodman, D., Batalden, P., Davidoff, F., & Stevens, D. (2016). SQUIRE 2.0 (Standards for QUality Improvement Reporting Excellence): Revised publication guidelines from a detailed consensus process. BMJ Quality & Safety, 25(12), 986-992. https://doi.org/10.1136/bmjqs-2015-004411

Perla, R. J., Provost, L. P., & Murray, S. K. (2011). The run chart: A simple analytical tool for learning from variation in healthcare processes. BMJ Quality & Safety, 20(1), 46-51. https://doi.org/10.1136/bmjqs.2009.037895