HLTH 5643 Module 5 Data Governance Plan Example

When Is a Bed Clean? Owners, Definitions and a Tie-Breaker for the Data in a Hospital's New Bed Management System

Student Name

American College of Education

HLTH5643: Information Systems Management for Healthcare Administrators

Module 5 Assignment

Instructor Name

October 2, 2028

The Disputes That Made Governance Necessary

Six weeks after the bed management system switched on across the invented regional hospital that anchors this course, three disagreements reached the project lead in the same week. Charge nurses on the surgical units complained that beds marked clean by environmental services were not ready for patients because equipment such as suction setups and pumps had not been returned. Environmental services answered that cleaning was their job and equipment was not. The emergency department's weekly boarding report showed a median of 3.1 hours from admission decision to departure from the department, while the throughput report the chief operating officer presented to the board showed 2.4 hours for the same weeks. And a unit manager asked for a report of each housekeeper's average cleaning time, which the implementation team had told environmental services staff would never be produced.

None of these is a software problem. Each is a question about what the data mean, who decides and how the data may be used, which is the work of data governance. A new system does not create disagreements about data; it makes old disagreements visible and gives them numbers.

Roles: Owners, Stewards and Custodians

The plan assigns roles at three levels. Data owners are executives accountable for a domain of data and for decisions about its use. The chief nursing officer owns bed and unit status data, and the chief operating officer owns throughput measures, including boarding time and time to bed assignment. Data stewards are managers responsible day to day for the accuracy and definitions of specific data elements: the environmental services manager for cleaning statuses and timestamps, the director of patient flow for assignment and pending statuses, the patient access manager for admission, discharge and transfer events and the emergency department nurse manager for decision-to-admit timestamps. The data custodian is the information systems department, which maintains the system, the interface and the reports, but does not decide what the data mean.

Separating stewards from the custodian matters. During the first six weeks, the analysts who built the reports had been settling definitions themselves, each choosing a reasonable meaning, which is how two reports came to show different boarding times. Rosenbaum (2010), writing about data governance and stewardship in health care, frames stewardship as accountability for how data are managed, accessed and used, which is a role distinct from technical maintenance of the systems that hold them.

Definitions, Written Down

The stewards met twice to write a data dictionary of 16 terms, beginning with the contested ones. The most important was clean. The agreed definition separates two statuses: cleaned, set by environmental services when cleaning is complete, and ready, set when the room also has its standard equipment in place. Ready is set automatically when the unit's equipment checklist, completed by the charge nurse or a nursing assistant on the mobile device, is marked complete for that room, and supervisors assign only beds that are ready. The median gap between cleaned and ready will be reported monthly, which makes the equipment problem visible without blaming either department for it.

The boarding time dispute came from different start and stop points. The emergency department's report started at the physician's decision to admit, recorded in the record, and stopped when the patient left the department. The throughput report started at the bed request entered in the new system, often 30 to 40 minutes later, and ended when a bed was assigned. Both are legitimate measures of different things. The dictionary now names them separately: boarding time, from decision to admit to departure from the emergency department, and placement time, measured from the moment a bed is requested until one is allocated, and specifies which appears in each report. Provost and Murray (2011) argue that an operational definition should state exactly how a characteristic is measured so that different people produce the same number, and that disputes about data are often disputes about definitions in disguise.

Checking Data Quality

Definitions mean little if the data do not follow them. The plan uses the framework proposed by Kahn et al. (2016), which groups data quality checks into three categories: conformance, whether values follow the required formats and allowed values; completeness, whether expected data are present; and plausibility, whether values are believable. A monthly quality report from the custodian applies checks in each category. For conformance, it lists status values outside the dictionary's allowed list, such as a free-text status left by a configuration error. For completeness, it reports the share of discharged beds that received a cleaned timestamp, with a target of at least 98 percent. For plausibility, it flags cleanings recorded in under ten minutes, which usually mean a status set in advance or by mistake, and rooms marked ready before they were marked cleaned.

Each steward receives the part of the report for their elements and must explain any check that falls below its threshold at the next monthly meeting. The report is not a performance measure for staff; it is a measure of whether the data can be trusted for the decisions that depend on them.

Who Settles Disagreements

The plan names a path and a time limit for every disagreement about data. A steward who disagrees with another steward about a definition, a report or a data correction raises it at the weekly patient flow stewards' meeting, which lasts 30 minutes and is chaired by the director of patient flow. If the stewards cannot agree within two meetings, the issue goes to the Patient Flow Data Council, which meets monthly and includes both data owners, the stewards and the information systems director. If the council cannot reach consensus, the chief operating officer decides for throughput measures and the chief nursing officer for status data, and if the question spans both domains, the chief operating officer makes the final decision after consulting the chief nursing officer. Every decision is recorded in the data dictionary with its date and reasoning.

The time limit is as important as the path. Without one, contested definitions stay unresolved and each department keeps its own version, which is exactly how the two boarding reports developed. The plan allows no more than 45 days from the first raising of a dispute to a recorded decision. A tie-breaker that everyone knows about is used less often than one nobody has named.

Rules for Use

Governance also decides how data may be used. During implementation, environmental services staff were told that bedside status updates served dispatch planning and would never become a stopwatch on any one person, and that promise was part of how the team reduced resistance. The request for each housekeeper's average cleaning time tested it. The council has written the promise into policy: cleaning data may be reported by unit, shift and room type, but reports identifying individual employees may be produced only for a specific, documented concern about an individual raised through human resources, approved by the environmental services director and never as a routine ranking. The unit manager's request was declined, with an offer of a report by shift that addressed the manager's underlying question about evening delays.

The same principle applies to charge nurses' use of the admission pause. Reports show how often each unit uses the feature and for how long, which helps supervisors plan, but not which charge nurse placed each hold. Access to the audit log that does identify users is limited to information security and privacy staff investigating a specific concern.

Conclusion

The bed management system made the hospital's disagreements about bed data visible within weeks. The governance plan assigns owners and stewards by name and position, keeps the custodian out of decisions about meaning, writes contested definitions down, checks data quality monthly in three categories and names who settles a dispute and how quickly. It also turns a promise to front-line staff into a written rule about use. With these in place, the next disagreement about what a number means should reach a recorded decision in weeks rather than producing another competing report.

References

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 18. https://doi.org/10.13063/2327-9214.1244

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

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