HINF 520 Week 5 Reporting Mechanisms and Data Exchange Example

One Diabetes Measure, Five Audiences and the Outside Results That Never Arrived: Reporting Mechanisms and Data Exchange for a Health System's Glycemic Control Rate

[Student Name]

University of Phoenix

HINF/520: Data Management and Design in Health Administration

Week 5 Assignment

[Instructor Name]

[Date]

The health system, its reports, rates and exchange partners are composites written for a model paper; research findings come from the sources listed.

In one month, the composite health system reported the share of its patients with diabetes whose most recent glycemic test was above 9% four different ways: 27% on the executive dashboard, 31% in its federal quality submission, 24% to a commercial payer and 29% in a clinic performance report. A board member, the same one who had asked how many patients had diabetes, asked which number was right. This paper explains how the system's reporting mechanisms work, why they disagreed and how data exchange with outside providers affects every one of them.

A Measure Where Lower Is Better

The measure counts patients with diabetes whose most recent test result, usually an A1c, is above 9%, or who had no test during the period. Lower is better. It appears in federal programs, in payer contracts and in national quality reporting, but each audience defines its population and time period somewhat differently.

Audience One: Clinic Care Gap Lists

The clinics receive a daily list of their patients with a most recent A1c above 9% or no test in six months, drawn from the data mart. Its purpose is action, not scoring, so it includes every patient seen in the past two years and updates as soon as a new result arrives.

Audience Two: The Executive Dashboard

The executive dashboard shows the measure monthly by hospital region and clinic, using the system's own validated definition and a rolling 12-month period, with trend lines and control limits.

Audience Three: Federal Quality Reporting

The federal submission follows the measure's official specification: a calendar year, age limits of 18 to 75, exclusions for patients in hospice or with advanced illness and frailty and data submitted in a standard electronic quality reporting file format generated by the record. Because the specification counts patients with no test in the year as failing, and the system's own definition used a longer lookback, the federal rate was higher.

Audience Four: A Commercial Payer

The payer calculates the measure from claims and supplemental data the system sends. Its population includes only the payer's members, who are younger and less likely to have poorly controlled diabetes than the system's patients overall, so the rate was lower.

Audience Five: A Public Report

A regional health collaborative publishes clinic-level results using combined data from several payers, updated annually with an 18-month lag, so its figure described an earlier period.

Why Validation Matters

Electronic reports are only as good as the data and logic behind them. When researchers checked measures produced automatically by the record against nurses' manual reviews of the same charts, they found that sensitivity ranged from 46% to 98% across measures, and that electronic reporting overestimated cholesterol control among patients with diabetes, 57% compared with 37% on chart review, while underestimating other measures (Kern et al., 2013). The system now reviews a random sample of 50 charts before each federal submission, comparing the record's automated result with a manual review. Its first review found that results from a point-of-care device used in two clinics were recorded in a field the measure logic did not read. An electronic measure can be precise to a decimal point and still be wrong about the patients it counts.

The Missing Outside Results

All five reports shared one gap: A1c results from outside laboratories and clinicians. About 18% of the system's patients also receive care elsewhere, and results from those visits often arrive as scanned documents or not at all. A patient tested at an outside clinic in October appears on the system's list as having no test.

How Much Data Moves Between Organizations

The gap is not unique. National survey data showed that in 2015 only 29.7% of U.S. hospitals did all four things that interoperability requires, locating outside records, sending their own, receiving others' and bringing what they received into their own records, up from 24.5% a year earlier, and only 18.7% reported that clinicians often used information from outside providers (Holmgren et al., 2017). Integration, bringing outside data into the record as usable data rather than documents, showed the least progress.

Standards for Exchange

Several standards carry data between organizations. Laboratory results travel in HL7 version 2 messages. Clinical summaries travel as structured clinical documents, which can include coded results but often arrive as text. Quality submissions use standard electronic reporting files. Newer exchange uses FHIR, and a platform built on it allows third-party apps to run with different vendors' records using standard authorization, so a tool built once can be used across systems (Mandel et al., 2016).

The Exchange Plan

The system's plan has four parts. It will request structured laboratory results from the regional health information exchange through a standard interface rather than documents, starting with A1c and kidney function tests. It will ask its two largest partner clinic groups to send results through the same route. It will reconcile results from clinical documents into structured fields when they arrive. And it will send its own results to payers electronically as supplemental data, reducing requests for chart pulls.

Privacy in Reporting and Exchange

Every report and exchange must respect privacy rules. Clinic lists contain names and are available only to care teams. Executive dashboards show only aggregate rates, with cells under 11 patients suppressed so individuals cannot be identified in small clinics. Payer submissions go only for the payer's own members under the business relationship allowed by federal privacy rules, and exchange through the health information exchange follows its participation agreement and patients' opt-out choices. The system's privacy officer reviewed each report type before the new plan took effect.

Timeliness

Reports also differ in how current they are. The clinic lists update daily, the dashboard monthly, the federal submission annually and the public report after 18 months. Clinicians need daily data to act; executives need stable monthly trends; regulators need complete annual data. Matching timeliness to purpose avoids both stale action lists and noisy executive reports.

One Definition, Many Views

To reduce confusion, the system will label every report with its definition, population, period and data source and publish a short guide explaining why the numbers differ. Internally, it will use one standard definition for all management reports, matching the federal specification where possible.

Measuring Improvement

The system will track the share of A1c results from outside sources received as structured data, the difference between automated and chart-review results in each validation sample and the timeliness of reports.

Conclusion

One diabetes measure reported to five audiences produced five numbers, each defensible for its purpose. Differences in definitions, populations and periods explained most of the gap, while validation revealed a logic error and exchange data showed that outside results were missing from every report. Clear labeling, validation before submission and structured exchange with partners will make each number more accurate and the differences easier to explain.

References

Holmgren, A. J., Patel, V., & Adler-Milstein, J. (2017). Progress in interoperability: Measuring US hospitals' engagement in sharing patient data. Health Affairs, 36(10), 1820-1827. https://doi.org/10.1377/hlthaff.2017.0546

Kern, L. M., Malhotra, S., Barrón, Y., Quaresimo, J., Dhopeshwarkar, R., Pichardo, M., Edwards, A. M., & Kaushal, R. (2013). Accuracy of electronically reported "meaningful use" clinical quality measures: A cross-sectional study. Annals of Internal Medicine, 158(2), 77-83. https://doi.org/10.7326/0003-4819-158-2-201301150-00001

Mandel, J. C., Kreda, D. A., Mandl, K. D., Kohane, I. S., & Ramoni, R. B. (2016). SMART on FHIR: A standards-based, interoperable apps platform for electronic health records. Journal of the American Medical Informatics Association, 23(5), 899-908. https://doi.org/10.1093/jamia/ocv189