DDHA 8503 Week 8 Intervention Design Example

Hold Visit Slots for the Messages That Need a Visit, Then Watch What the System Does Back: An Intervention Design for a Primary Care Inbox Loop

Student Name

Walden University

DDHA 8503: Strategic Management and Systems Thinking

Instructor Name

Month Day, Year

Hold Visit Slots for the Messages That Need a Visit, Then Watch What the System Does Back: An Intervention Design for a Primary Care Inbox Loop

The Condition Being Addressed

Earlier weeks traced a reinforcing loop at Brightwater Medical Group's fourteen primary care clinics. A growing message backlog pushes clinicians into evening inbox work, which builds burnout, which over a year or more reduces the clinical hours clinicians offer, which reduces visit availability, which leads patients to send messages in place of visits, which grows the backlog. The condition this intervention addresses is the link between visit availability and messages sent in place of visits, where the loop turns an access problem into an inbox problem within weeks.

The Change Proposed

The intervention has three parts. First, each clinician's schedule will hold four 20-minute video visit slots per week, released only to patients whose message the triage nurse or clinician judges to need a visit, and bookable within five business days. Second, each full-time clinician will receive two hours a week of protected inbox time inside the scheduled day, counted as clinical time in productivity reports. Third, the clinic dashboard will drop its two-day reply percentage and will report instead the share of patients who send a second message on the same issue within seven days, a measure of whether the first reply resolved anything.

The scheduling and triage teams will implement the first part, clinic medical directors the second and the quality office the third. No staff are added; the visit slots and protected time come from existing schedules.

Why Act Here

The obvious place to intervene is reply speed, the variable the previous standard targeted and the one leaders see on the dashboard. That intervention failed once, as the Week 1 analysis showed, because quick replies generated follow-up messages. Meadows (2008) argued that the most effective points of intervention in a system are often not the most visible ones, and that changing the information a system uses to steer itself, or the rules that govern it, usually achieves more than adjusting a single flow. This design acts in two such places. The visit slots act on the structural link where scarce visits create messages. The new measure changes the information that steers clinic behavior, rewarding resolution rather than speed. Protected inbox time acts on the link from backlog to evening work, which matters for the loop's slow burnout link even though its effect will not appear for months.

How the System Will Respond

A change described without its consequences is half a design. Each affected group will respond, and the timing matters.

Clinicians will respond first, within the first month. Four visit slots and two inbox hours come out of routine visit time, which at 20-minute visits removes about ten routine slots per clinician per week, four of them converted into fast-lane video visits. Clinicians with long routine waits will feel this immediately and may resist, or may fill held slots with routine patients if the slots go unused.

Triage nurses will respond within weeks. They gain a way to resolve messages that need a visit, which should reduce their back-and-forth with patients. They may also route more messages to the fast lane than intended, since booking a visit is the safest answer when unsure.

Patients will respond over one to three months. Those who use the fast lane will learn that a message can lead to a prompt visit, and some will begin writing to obtain visits they could not book directly, turning the portal into a side door to the schedule.

Schedulers will respond as routine waits lengthen, receiving more calls from patients who cannot get a routine appointment.

Compensating Effects

Two compensating effects could restore the condition the intervention is meant to change. The first is the most important. The slots held for message-driven visits come from routine visit capacity, so routine visit availability falls. In the loop, falling visit availability drives messages sent in place of visits. The intervention could therefore reduce messages from patients who need a visit while increasing messages from patients who cannot get a routine one. Whether the net effect is favorable depends on whether the fast lane resolves enough messages to free more clinician time than it takes.

The second is success attracting demand. If the fast lane works, patients will use it more, and held slots will fill with patients who learned that messaging gets them seen faster. Repenning and Sterman (2001) describe how improvement efforts can be undermined when the time they free up is immediately absorbed by new demand, so that people see no lasting benefit and abandon the change. Murray and Berwick (2003), writing about advanced access, note that reducing delay requires working down backlog and shaping demand together, not only adding same-day capacity.

Sequencing

The order of the three parts is a deliberate choice. The measure change comes first, in month one, because it costs nothing and because without it the dashboard would still reward quick replies that close messages without resolving them. Protected inbox time comes second, in month two, because clinicians need to see that the group is taking work off them before they are asked to give up routine visits. The fast-lane slots come third, starting in three clinics in month three and spreading to the others only if the monitoring signals below look favorable.

Monitoring

The design names signals that should move before the outcome does. Repeat messages within seven days should fall within six weeks if replies are resolving more. The share of fast-lane slots that go unused or are filled by routine patients will show within a month whether the slots are being protected. The third-next-available routine visit will show within two months whether routine access is paying for the fast lane. Evening inbox minutes from the record system logs should fall within three months of protected time starting. The outcome itself, clinical hours offered, will not move for a year or more; if the early signals are favorable, the design should be given that year rather than judged on the first quarter.

References

Meadows, D. H. (2008). Thinking in systems: A primer (D. Wright, Ed.). Chelsea Green.

Murray, M., & Berwick, D. M. (2003). Advanced access: Reducing waiting and delays in primary care. JAMA, 289(8), 1035-1040. https://doi.org/10.1001/jama.289.8.1035

Repenning, N. P., & Sterman, J. D. (2001). Nobody ever gets credit for fixing problems that never happened: Creating and sustaining process improvement. California Management Review, 43(4), 64-88. https://doi.org/10.2307/41166101