DNP880 Module 5 assignment: data collection plan: measures, definitions, sources and quality checks, a full sample

Numerator, Denominator, Source, Owner: The Data Collection Plan for a Food Insecurity Screening and Referral Project

Student Name

Doctor of Nursing Practice Program, Aspen University

DNP880: DNP Project

Instructor Name

Month Day, Year

Numerator, Denominator, Source, Owner: The Data Collection Plan for a Food Insecurity Screening and Referral Project

A project is only as credible as the data it collects. This plan specifies each measure for the project at Maple Street, its exact definition, where the data come from, who collects them and how often, and how their accuracy will be checked. It follows the project's PICOT question, in which screening rate and referral completion are the primary outcomes, A1c is exploratory, and visit length is a balancing measure.

A Family of Measures

Improvement teams are advised to track a small set of measures together: outcome measures that show whether the system is producing better results, process measures that show whether the planned changes are being carried out, and balancing measures that watch for problems the changes might create elsewhere (Langley et al., 2009). The same logic echoes the distinction Donabedian drew among structure, process, and outcome as the three kinds of information from which the quality of care can be inferred, with process and outcome linked by evidence that one leads to the other (Donabedian, 1988). For this project, screening and referral are processes that the evidence synthesis connected to food access, and food access is the intermediate outcome the project can reasonably influence in 16 weeks.

Measures and Operational Definitions

Table 1 defines each measure. All rates are calculated weekly by clinic.

Decisions Built Into the Definitions

Several choices in Table 1 deserve explanation. The two items will be scored as in their validation study, with an answer of often true or sometimes true to either item counted as positive (Hager et al., 2010), so that the local positive rate can be compared with published figures. A documented decline counts as a screened visit because the aim is to offer the questions, not to compel answers; declines will also be tracked separately so that a high rate would prompt a review of how the questions are being asked. Patients seen more than once are counted at each visit for the screening rate, since each visit is an opportunity, but only once for referral completion, at their first positive screen, to avoid inflating the result. Referral completion relies on patient report at the follow-up call, cross-checked against the food bank's list for patients who agreed to share their information, because neither source is complete alone. A measure that cannot be counted the same way twice by two different people is not yet a measure.

Baseline Data

Baseline screening will be calculated from the three months before implementation using the same definitions, with any documentation of food access in the social history counted as screening. Because no structured field existed before the project, baseline will be measured by the student's review of a random sample of 150 eligible visits, 75 from each clinic, drawn by the analyst. The earlier review of 100 visits, in which food access was documented in 6, will be reported as a preliminary estimate but not used as the formal baseline, because it was drawn from all clinics rather than the two in the project.

Who Collects What, and When

Medical assistants enter screening results in the rooming template as part of care; no separate data form is needed. The community health worker completes a brief follow-up call form with four fields: date of call, whether the patient was reached, whether food was received, and the main barrier if not. The informatics analyst runs a weekly report every Monday morning containing the de-identified fields defined in the ethics submission. The student downloads the report the same day, calculates the weekly rates, and updates the run charts before the Tuesday team meeting. The food bank liaison sends a confirmation list every other Friday. A1c values are extracted once at baseline and once at week 16.

Checking Data Quality

Electronic record data can look complete while being wrong. Three checks will be used. First, in weeks 2 and 8 of implementation, the student will review 30 randomly selected visits per clinic, comparing the template entries to the clinician's note and asking medical assistants about any mismatch; agreement below 90 percent will trigger retraining. Second, the student will compare the weekly report's count of eligible visits against the scheduling system's count of diabetes visits, investigating differences greater than 5 percent. Third, missing follow-up call forms will be flagged weekly to the community health worker so that calls are not lost. All checks and their results will be logged and reported in the final paper.

Patient Comments

Numbers will not explain why patients decline screening or do not reach the food bank. The community health worker will record, in a free-text field on the follow-up form, any comment the patient volunteers about the experience of being asked or of using the food bank, without names. The student will group these comments by theme monthly. This gives the project a small but direct view of how patients experience the change, which the evidence synthesis identified as a gap.

Conclusion

The plan defines six measures across process, outcome, and balancing categories, with numerators, denominators, and sources; fixes the rules for declines, repeat visits, and referral confirmation; sets a baseline by sample review; assigns every data stream an owner and a schedule; and checks the accuracy of record data at set points. The analysis of these data is described in the evaluation plan.

References

Donabedian, A. (1988). The quality of care: How can it be assessed? JAMA, 260(12), 1743-1748. https://doi.org/10.1001/jama.1988.03410120089033

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

Langley, G. J., Moen, R. D., Nolan, K. M., Nolan, T. W., Norman, C. L., & Provost, L. P. (2009). The improvement guide: A practical approach to enhancing organizational performance (2nd ed.). Jossey-Bass.