HIM 550 Module 9 Final Project Example

Final Project: A Data Management Plan for Patient Flow Data at Cypress Hollow Medical Center

[Student Name]

Southern New Hampshire University

HIM 550: Data Management and Data Quality

Final Project

[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.

Final Project: A Data Management Plan for Patient Flow Data at Cypress Hollow Medical Center

Executive Summary

Cypress Hollow Medical Center is about to spend a year trying to move more discharges into the morning. The measure leaders planned to use was built on the time a clerk closes an encounter, and the data behind it contained blank admission sources, overwritten unit history and unchecked nightly loads. This plan establishes how the hospital's patient flow data will be governed, defined, created, stored, protected, monitored and reported so that the program is judged on accurate numbers. It requires no new software. Its costs are staff time for stewardship and analytics work, estimated at about 0.6 full-time equivalents in the first year, and its benefits include a trustworthy measure, fewer wasted investigations of false signals and a model that other data domains can follow.

Scope and Goals

The plan covers the patient flow data domain: admission, transfer and discharge events, admission source and type, nursing unit, discharge disposition, departure time and emergency boarding times, as stored in registration, the clinical record, the bed management system and the data warehouse. Four goals guide it. Leaders should see a discharge timing measure that reflects when patients leave. Missing values in key fields should fall below 2%. Every field should have a definition and an owner. And any secondary use of these data should follow a documented review.

Evidence from the Project

The quality assessment of 9,840 discharges found the registration discharge time complete and valid but trailing actual departure by a median of 3.4 hours, admission source blank on 12% of records and unit codes broken by a reorganization. The analysis of 8,640 routine discharges showed a stable weekly before-noon rate near 37%, no shorter stays for morning departures and shorter emergency boarding on days with more morning discharges. Rajkomar et al. (2016) found a similar absence of any stay reduction, which supports framing the program around bed flow. These findings shape every section that follows.

Governance and Stewardship

The vice president for patient flow will serve as data owner for the domain, accountable for its definitions and for approving any change. Three departmental stewards, from patient access, nursing and health information, will each maintain the fields their departments create, and an analytics steward will maintain warehouse rules. Stewards will meet monthly; unresolved disputes go to the existing data governance committee. Rosenbaum (2010) argued that stewardship needs clear authority, open processes and accountability, and this structure provides each: the owner decides, the rules are published and stewards report quality results monthly.

Standards and the Data Dictionary

Every field in the domain will have a dictionary entry listing its business definition, source system and field, allowed values, owner, known limitations and approved uses. Admission source values will follow the categories the state discharge submission requires, so internal and external reporting match. The discharge timing measure will be defined as the share of routine discharges with a nursing departure time before noon, excluding deaths, transfers to other acute hospitals and departures against medical advice. Changes to any definition will be dated in the dictionary so past reports can be interpreted.

The Data Life Cycle

Table 1 sets out how the plan manages the data at each stage of their life.

Table 1. Management of Patient Flow Data Across the Life Cycle

Note. Prepared by the author from the project's quality assessment and recommendations.

Warehouse Design

The warehouse will keep the dimensional structure Kimball and Ross (2013) describe, with one change: dimensions that change over time, beginning with nursing units and later service lines, will keep dated history so every discharge stays linked to the unit that existed when it occurred. Each nightly load will compare rows extracted with rows loaded, record run time and failures and alert analytics before 6 a.m. if counts are incomplete. Reports will display a notice when a load has not finished.

Privacy and Secondary Use

Patient flow data are attractive for research, operations models and outside partners, so every request beyond routine reporting will pass a written review. Requests from outside the hospital will be met with de-identified data prepared by a recognized method or, where dates and geography are essential, a limited data set released only after a signed agreement. El Emam et al. (2011) showed that re-identification succeeds mainly against data that were not properly de-identified, so removing names alone will never count. Internal predictive models will require a stated purpose and a check of results across payer and neighborhood groups before use.

Quality Monitoring and Reporting

Kahn et al. (2016) describe routine checks of conformance, completeness and plausibility, and three will run weekly: the share of records without an admission source, the share without a departure time and any departure stamped earlier than its admission. Each will be plotted on a control chart and reviewed at the stewards' meeting, with a named steward responsible for investigating signals. The leadership dashboard will show the weekly before-noon rate as a run chart with its target, unit panels on a common zero-based axis, emergency boarding hours beneath it and a note giving the measure's definition and the week's missing departure count.

Timeline and Resources

The first month covers naming the owner and stewards, starting the load audit and publishing dictionary entries for the ten core fields. Months two and three bring the redesigned registration and discharge screens, the dated unit dimension and the new dashboard, with old and new measures shown side by side for one month. Months four through six introduce the secondary use review and weekly control charts. The remaining months extend the dictionary and dated dimensions to service lines. Resources total about 0.6 full-time equivalents in the first year: roughly 0.4 of analytics time for the load audit, dimension rebuild and dashboard, about 0.1 of patient access and nursing informatics time for screen changes and training and about 0.1 spread across the owner and stewards for meetings, dictionary work and signal reviews.

Evaluation

Success will be judged at six and twelve months on five measures: missing admission source and departure time each below 2%, every nightly load reconciled, all core fields defined and owned, every secondary use request reviewed and leaders' ability to read the dashboard correctly in a short user test. A brief report to the governance committee will present results and recommend whether to extend the model to other domains, such as surgical scheduling data.

Conclusion

The hospital's patient flow program depends on data that, until this project, no one owned and few had examined. This plan gives the data an owner, clear definitions, protection at each stage of their life and continuous monitoring, so that the discharge program and the decisions that follow it rest on numbers leaders can trust.

References

El Emam, K., Jonker, E., Arbuckle, L., & Malin, B. (2011). A systematic review of re-identification attacks on health data. PLoS ONE, 6(12), Article e28071. https://doi.org/10.1371/journal.pone.0028071

Kahn, M. G., Callahan, T. J., Barnard, J., Bauck, A. E., Brown, J., Davidson, B. N., Estiri, H., Goerg, C., Holve, E., Johnson, S. G., Liaw, S.-T., Hamilton-Lopez, M., Meeker, D., Ong, T. C., Ryan, P., Shang, N., Weiskopf, N. G., Weng, C., Zozus, M. N., & Schilling, L. (2016). A harmonized data quality assessment terminology and framework for the secondary use of electronic health record data. eGEMs, 4(1), Article 1244. https://doi.org/10.13063/2327-9214.1244

Kimball, R., & Ross, M. (2013). The data warehouse toolkit: The definitive guide to dimensional modeling (3rd ed.). Wiley.

Rajkomar, A., Valencia, V., Novelero, M., Mourad, M., & Auerbach, A. (2016). The association between discharge before noon and length of stay in medical and surgical patients. Journal of Hospital Medicine, 11(12), 859-861. https://doi.org/10.1002/jhm.2529

Rosenbaum, S. (2010). Data governance and stewardship: Designing data stewardship entities and advancing data access. Health Services Research, 45(5, Pt. 2), 1442-1455. https://doi.org/10.1111/j.1475-6773.2010.01140.x