HIM 540 Module 8 Data Quality Short Paper Example

The Data Element Everything Depends On: Data Quality and Patient Identity at Tidewater Crossing Health

[Student Name]

Southern New Hampshire University

HIM 540: Health Information Governance

Module Eight Short Paper

[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.

The Data Element Everything Depends On: Data Quality and Patient Identity at Tidewater Crossing Health

Every report Tidewater Crossing Health produces assumes that one patient means one person. Readmission rates, quality measures, population health lists and research all count patients, and each count is wrong if one person has two records or two people share one. When the four hospitals merged their master patient indexes, a sample review suggested that about 7% of records were duplicates, and two overlays last year placed one patient's results in another patient's chart. This paper uses patient identity to show how data quality should be judged and governed.

How Data Quality Is Judged

Kahn et al. (2016) proposed a harmonized framework that groups data quality into three categories. Conformance asks whether data follow the expected formats, value sets and relationships. Completeness asks whether expected data are present. Plausibility is about believability: a value should fit the rest of the record and ordinary reality. Each can be checked by verification against internal rules or validation against an external standard. Applied to patient identity, conformance means a date of birth is a valid date, completeness means required identifiers such as sex, date of birth and address are recorded and plausibility means a newborn's record does not show a date of birth decades earlier.

Why Identity Errors Spread

Identity errors are unlike other data errors because they corrupt the unit being counted. A duplicate record splits a patient's history, so a clinician may miss an allergy or a prior result, and a readmission may be counted as a new admission. An overlay mixes two people's histories, which can lead to treatment based on someone else's data. McCoy et al. (2013) examined identifiers in large record systems and found that many patients shared the same name and date of birth with other patients, which makes mismatches likely unless additional identifiers are used and data are captured carefully. As the system grows, the chance of such coincidences grows too.

How Duplicates Arise

A review of 300 confirmed duplicate pairs at Tidewater Crossing identified the main causes. Most arose at registration: a nickname entered instead of a legal name, transposed digits in a date of birth, a missing middle name or an address typed differently. Others came from the merger itself, when the two platforms' indexes were combined without a full match, and from emergency registrations under temporary names that were never reconciled. Table 1 summarizes the findings.

Table 1. Causes of 300 Confirmed Duplicate Record Pairs

Note. Review by the health information data integrity team; composite data.

Matching Methods

Systems decide whether two records belong to the same person with matching algorithms. Deterministic matching requires exact agreement on chosen fields, which is simple but misses records with small differences. Probabilistic matching weighs agreement and disagreement across many fields and produces a score, so it finds more true matches but requires thresholds and human review of uncertain pairs. Joffe et al. (2014) compared deterministic and probabilistic methods for selecting record pairs for manual review and found that the approach chosen affected how many true duplicates reviewers would find and how much review work was needed. Tidewater Crossing will use probabilistic matching with two thresholds: automatic linking above the upper threshold, human review between them and no link below the lower one.

Standardizing Identity Data

Better data in means better matches out. Grannis et al. (2019) found that standardizing how names and addresses are formatted before matching improved matching accuracy, particularly standardizing addresses to a postal format. Registration at all four hospitals will therefore use address verification against postal data, required fields for legal name and prior names, a separate field for preferred name, photo capture for returning patients where they consent and a search of the index by several fields before any new record is created. Emergency registrations under temporary names will be reconciled by health information staff within 24 hours.

The Human Side of Identity

Registration staff create most duplicates, but they work under pressure. At the emergency department front desk, a registrar may have seconds to register an unconscious trauma patient or a family arriving at midnight. Blaming staff will not fix that. The plan gives registrars better tools, such as a search screen that suggests possible matches as they type and a clear rule for when to create a temporary record, and it measures duplicate creation by site and shift so that training reaches the places where errors occur. Registrars with low error rates will help teach others, and monthly feedback will show each team how its work affects patient safety.

Patients as Partners

Patients can help keep their identity data accurate. The patient portal will display demographic information and allow patients to request corrections, and registration staff will ask returning patients to confirm key details rather than reading them aloud, which reduces errors and protects privacy at busy desks.

A Governed Cleanup

The data governance committee's patient identity steward will lead the cleanup. Pairs above the upper threshold will be merged by trained health information staff with a second review for any pair involving a clinical conflict, such as different blood types. Suspected overlays will be investigated immediately, with affected clinicians and patients notified and records corrected. Every merge will be logged. The steward reports progress monthly.

Measures

The program will track the estimated duplicate rate from quarterly samples, with a target below 2% within eighteen months; the rate at which newly created records later prove to duplicate an existing patient, with a target below 1%; the number of overlays, with a target of zero; and the time to reconcile emergency registrations. Conformance and completeness of identity fields at registration will be reported by hospital.

Conclusion

Data quality frameworks give Tidewater Crossing a common language, and patient identity shows why that language matters: an error in who the patient is corrupts everything counted about them. Understanding how duplicates arise, matching with probabilistic methods and human review, standardizing identity data at registration and governing the cleanup with clear measures would make the system's most important data element trustworthy.

References

Grannis, S. J., Xu, H., Vest, J. R., Kasthurirathne, S., Bo, N., Moscovitch, B., Torkzadeh, R., & Rising, J. (2019). Evaluating the effect of data standardization and validation on patient matching accuracy. Journal of the American Medical Informatics Association, 26(5), 447-456. https://doi.org/10.1093/jamia/ocy191

Joffe, E., Byrne, M. J., Reeder, P., Herskovic, J. R., Johnson, C. W., McCoy, A. B., Sittig, D. F., & Bernstam, E. V. (2014). A benchmark comparison of deterministic and probabilistic methods for defining manual review datasets in duplicate records reconciliation. Journal of the American Medical Informatics Association, 21(1), 97-104. https://doi.org/10.1136/amiajnl-2013-001744

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

McCoy, A. B., Wright, A., Kahn, M. G., Shapiro, J. S., Bernstam, E. V., & Sittig, D. F. (2013). Matching identifiers in electronic health records: Implications for duplicate records and patient safety. BMJ Quality & Safety, 22(3), 219-224. https://doi.org/10.1136/bmjqs-2012-001419