HI499 Unit 4 Source Synthesis Example

Four Findings and One Gap: A Unit 4 Professional Source Synthesis on Matching Thresholds, Standardized Addresses, Patient-Entered Data and the Cost of Duplicate Records

[Student Name]

Purdue University Global

HI499: Bachelor's Capstone in Health Information Management

Unit 4 Professional Source Synthesis

[Instructor Name]

[Date]

Callowfield Health is composite; every source cited is real, and the search described was run for this sample.

How the Sources Were Gathered and Judged

I searched PubMed, CINAHL and the AHIMA body of knowledge for work published since 2010 on patient matching, duplicate records and master patient index quality, then read the reference lists of the strongest items. Each finding below follows the same sequence: the claim, the sources that support it, how strong that support is, any disagreement, and what the finding means for Callowfield Health's self-scheduling tool. Strength is judged openly. A professional review states what practitioners agree on, not what has been tested. A study that measures matching accuracy before and after a change is stronger evidence on a narrower question. A survey-based estimate is weaker than its headline suggests.

Finding 1: Records in an Uncertain Middle Band Belong in Human Review

Claim: matching engines that score candidate pairs should send pairs in an uncertain middle range to a trained person rather than deciding them automatically in either direction. Support: reviews of patient identification techniques describe probabilistic matching as a scoring process with two thresholds, above which records are treated as the same person and below which they are treated as different, with the band between them reserved for manual review (Riplinger et al., 2020). Research on index data discrepancies supports the same structure from the other side: because names, birth dates and addresses differ between true matches for ordinary reasons, a strict automatic rule produces duplicates, while a loose one produces overlays (Just et al., 2016). Strength: this is professional consensus supported by descriptive research, not an experiment. No source tested what happens when the middle band is sent to automatic record creation, which is what Callowfield's tool does. Implication: the tool's setting conflicts with the most consistent advice in the field. If the root cause analysis confirms that most duplicates scored in the uncertain band, the fix has strong professional backing even without a trial. Every source that describes matching thresholds treats the uncertain band as a question for a person, and none treats it as permission to create a new record.

Finding 2: Standardizing Fields Improves Matching

Claim: putting names, addresses and other demographic fields into a standard form before matching improves accuracy. Support: the strongest study is an evaluation that applied standardization and validation rules to the demographic fields of real patient records and measured matching accuracy before and after; its authors reported that putting surname and address entries into a standard form raised accuracy (Grannis et al., 2019). The federal Project US@ effort produced a unified specification for formatting addresses in health records, which reflects the same reasoning at a national scale. Strength: moderate and direct. The evaluation measured accuracy on real data, but in data sets that had already been collected, not in a live registration screen. Disagreement: none found on the direction of the effect; sources differ only on which fields matter most. Implication: an address standardization service in the self-scheduling tool, checking a patient's address against a postal standard as it is typed, is the second change with published support.

Finding 3: Patient-Entered Data Is Thinly Studied

Claim: data that patients enter themselves may differ in quality from data that staff enter, but the evidence is thin. Support: research on identifier matching shows that duplicates and overlays arise when identifying fields are missing, shared or entered inconsistently, and that these errors carry safety risks (McCoy et al., 2013). That work, however, studied records created mostly by staff. I found no study that isolated records created by patients through online forms and compared their error rates with staff-created records. Strength: weak for this question. The section is shorter than the others, and it says so. Implication: the capstone cannot borrow a published rate for patient-entered data. Its own baseline, 4.09 percent for the tool against 0.66 percent at clinic desks, is the best evidence available for this setting, and the final paper should offer it as what Callowfield found in its own records, not as a rate other systems should expect. A second point follows. If patients make more entry errors than staff, the remedy may lie partly in the screen itself, such as clearer field labels, a prompt to enter a former surname and a check on birth dates that fall on the booking date, which is a common slip when a patient types today's date by mistake.

Finding 4: Cost Estimates Vary Widely, for Identifiable Reasons

Claim: duplicate records cost organizations money, but published estimates differ by large margins. Support: a national policy report on patient matching describes match failures as a source of repeated tests, delayed care and administrative rework, and calls for better standards across the industry (The Pew Charitable Trusts, 2018). Estimates of the cost per duplicate in industry surveys range widely because they count different things: some include only the labor of researching and merging a pair, while others add denied claims, repeated imaging and the downstream effects of a missed allergy. Strength: weak to moderate. Most figures come from surveys of organizations rather than measured costs. Implication: the capstone's cost plan should build its own estimate from local, checkable inputs, such as minutes per merge and claims held per month, and should name what it leaves out rather than borrow a national figure.

The Gap: No Evaluation of Self-Scheduling as a Registration Channel

I searched for evaluations treating online self-scheduling, patient portals or kiosks as their own registration channel, using combinations of "self-scheduling," "online scheduling," "patient portal," "registration," "duplicate record" and "patient matching." The search found descriptions of self-scheduling's benefits for access and studies of portal adoption, but no study measuring the duplicate rate of records those tools create. This gap matters for two reasons. It means the capstone's local data cannot be compared with a published benchmark, and it means a short report of Callowfield's results, after the change, could add something useful to the field.

What the Synthesis Means for the Capstone

Taken together, the findings point in one direction. The tool sends uncertain matches to record creation, which contradicts the field's consistent advice. It accepts addresses as typed, without the standardization that has measured support. Patient-entered data is the least studied part of the problem, so local evidence must carry more weight. And cost should be built from local inputs. The root cause analysis in Unit 5 will test whether the uncertain band is where Callowfield's duplicates actually come from, and the constraint review in Unit 6 will check what the rules allow before any change is recommended.

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

Just, B. H., Marc, D., Munns, M., & Sandefer, R. (2016). Why patient matching is a challenge: Research on master patient index (MPI) data discrepancies in key identifying fields. Perspectives in Health Information Management, 13(Spring), 1e.

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

Riplinger, L., Piera-Jiménez, J., & Dooling, J. P. (2020). Patient identification techniques: Approaches, implications, and findings. Yearbook of Medical Informatics, 29(1), 81-86. https://doi.org/10.1055/s-0040-1701984

The Pew Charitable Trusts. (2018). Enhanced patient matching is critical to achieving full promise of digital health records.