HI501 Unit 8 Patient Engagement Analysis Example

612 Cuffs Lent, 208 Still Sending at Week Eight: A Unit 8 Patient Engagement Analysis That Counts Actions, Not Logins, and Reads the Portal Screen as Patients Do

[Student Name]

Purdue University Global

HI501: Health Informatics

Unit 8 Patient Engagement Analysis

[Instructor Name]

[Date]

Harlow Ridge Health, its program and every figure are composite; the studies cited are real.

The Program and the Question

Eighteen months ago, Harlow Ridge's primary care offices began lending wireless cuffs to adults whose last two office blood pressures were at or above 140 over 90. The cuff pairs with the patient's phone, readings flow to the portal, and the portal shows each patient a trend line of their own readings. Care teams review flagged readings weekly. Home monitoring with this kind of support can lower blood pressure, especially when readings lead to changes in treatment (Tucker et al., 2017). The question for this analysis is whether the returned data led patients to act, and if not, where the chain broke.

Definitions

Engagement here means an action a patient took because of returned data: sending readings, sending a message about them, booking a visit, or having a dose changed after contact with the care team. Portal logins and page views are reported but kept separate, because a login says a patient arrived, not that anything happened. Researchers studying digital health tools have argued for this distinction, treating engagement as both the extent of use and what users experience and do as a result (Perski et al., 2017).

How the Logs Were Read

The informatics analyst extracted six months of device, portal and messaging logs for every patient lent a cuff, linked by a study code rather than a record number. Each log event was assigned to one funnel stage. A patient counted as still sending at week 8 if at least two readings arrived during that week. Trend views were counted from the portal's page log, and messages were matched to readings when they were sent within three days of a reading and mentioned blood pressure, a number or the cuff. Two reviewers sorted the messages into content categories separately and settled disagreements together. Dose changes were taken from medication orders written within fourteen days after a message or a flagged reading review. Preferred language came from registration. These rules are stated here so that the same analysis could be repeated next quarter and the figures compared.

Results: The Funnel

Of 612 patients lent a cuff over six months:

1. Activated the cuff and paired it with a phone: 402, or 66 percent.

2. Sent a first reading: 371.

3. Still sending readings at week 8: 208, meaning 44 percent of those who started had stopped.

4. Viewed their trend at least once a month in months 3 through 6: 166.

5. Sent at least one portal message about readings: 109.

6. Had a medication dose changed after a reading review or message: 47.

Portal logins averaged 5.2 per month among active users, a figure that looks healthy and says little. The steepest drop comes between the first reading and week 8, when nearly half of patients who started stopped sending. Most patients who stopped did so after weeks of sending readings that drew no visible response, which suggests the display asked them for data without ever telling them what it meant.

Results: What Patients Asked

The 109 patients who messaged sent 240 messages about their readings. Sorted by content, 61 percent asked some version of whether a number was acceptable, such as "Is 148 over 92 okay?" or "Should I worry about this?" Another 18 percent concerned the cuff or the phone connection, 12 percent asked about medication side effects, and 9 percent were other questions. The dominant question, whether a number is good or bad, is one the portal could answer for every reading and does not.

Results: Who Never Saw the Data

Activation differed sharply by preferred language. Among 512 English-preferring patients, 71 percent activated the cuff; among 100 Spanish-preferring patients, 38 percent did. The portal has translated pages, but the cuff's pairing instructions and the welcome message that explains the program were sent in English only. Earlier research found portal enrollment and use lower among Black and Latino patients, even among those with access, which shows that a portal can widen gaps it was meant to close (Goel et al., 2011). Here the gap began before any data were returned: many Spanish-preferring patients never paired the cuff.

Discussion: Reading the Screen as a Patient Does

Opening the portal as a patient, I found a line chart of systolic and diastolic readings over time, the date of each reading and nothing else. Three things are missing.

1. A target. The chart has no band showing the patient's goal, so a reading of 148 looks the same as one of 128 unless the patient already knows their target.

2. A sign that anyone looked. Care teams review flagged readings weekly, but the portal never says so. A patient who sends readings for six weeks sees no evidence that a person has seen them, which is the most plausible explanation for the week 8 drop.

3. A next step. When a reading is high, the screen does not say whether to recheck, call or wait for the team, so patients send a message asking, and many simply stop.

None of these additions requires new data; the target is already in the problem list, the review is already logged and the instructions already exist in the clinic's hypertension protocol. The display returns data to the patient but not information. It asks for engagement while offering nothing to engage with.

Recommendations

1. Add a shaded target band to every patient's chart and a plain-language label on each reading: at goal, above goal or very high.

2. Show a dated line, "Reviewed by your care team," each time a nurse or physician reviews the patient's readings.

3. Add a short instruction for very high readings, approved by the clinical lead, telling the patient what to do.

4. Send pairing instructions and the welcome message in Spanish to Spanish-preferring patients, and offer a phone call to help pair the cuff.

5. Report the funnel monthly by preferred language, with week 8 retention and dose changes as the main measures.

Limits

The logs show what patients did, not why, and no patients were interviewed for this analysis. Dose changes may have followed office visits unrelated to the readings. And six months is short for hypertension, where control often takes several medication adjustments.

References

Goel, M. S., Brown, T. L., Williams, A., Hasnain-Wynia, R., Thompson, J. A., & Baker, D. W. (2011). Disparities in enrollment and use of an electronic patient portal. Journal of General Internal Medicine, 26(10), 1112-1116. https://doi.org/10.1007/s11606-011-1728-3

Perski, O., Blandford, A., West, R., & Michie, S. (2017). Conceptualising engagement with digital behaviour change interventions: A systematic review using principles from critical interpretive synthesis. Translational Behavioral Medicine, 7(2), 254-267. https://doi.org/10.1007/s13142-016-0453-1

Tucker, K. L., Sheppard, J. P., Stevens, R., Bosworth, H. B., Bove, A., Bray, E. P., Earle, K., George, J., Godwin, M., Green, B. B., Hebert, P., Hobbs, F. D. R., Kantola, I., Kerry, S. M., Leiva, A., Magid, D. J., Mant, J., Margolis, K. L., McKinstry, B., ... McManus, R. J. (2017). Self-monitoring of blood pressure in hypertension: A systematic review and individual patient data meta-analysis. PLOS Medicine, 14(9), Article e1002389. https://doi.org/10.1371/journal.pmed.1002389