The 13-Drug Discharge With No Follow-Up
Mr. R is a 68-year-old admitted 4 days ago with acute decompensated heart failure (EF 30%, eGFR 38, A1c 9.1%). His discharge medication list now has 13 drugs reconciled from three prescribers (cardiology, nephrology, endocrinology); his pre-admission list had 9. He lives alone, his daughter is 90 minutes away, and he has no car. No primary care appointment is scheduled, cardiology follow-up is 6 weeks out, and the SGLT2 inhibitor is pending prior authorization with a $240 out-of-pocket estimate. He has been readmitted twice in 5 months, both times within 10 days of discharge. The hospitalist says, "He's medically stable, he can go tomorrow."
1. You have 48 hours before discharge. Identify the three coordination actions you would prioritize and justify the ordering.
Read the worked answer
Reasoning frame: rank by risk × modifiability × time sensitivity, not by how easy the task is.
- Close the follow-up gap - PCP or transitional care visit within 7 days. Highest yield because it is both high-risk and fully modifiable before tomorrow. Two admissions within 10 days of discharge mean his failure point is the first week at home, not the hospital stay. Cardiology at 6 weeks is the wrong time constant for the risk window.
- Full medication reconciliation with explicit deprescribing review, not just a corrected list. Thirteen drugs across three prescribers predicts duplication and interaction; the discharge summary must reflect one reconciled list with a stated reason for every change, or the PCP inherits ambiguity.
- Resolve access before the order is written - prior authorization status and the $240 cost. An unaffordable prescription is a non-prescription; a guideline-directed drug that the patient abandons at the pharmacy produces no outcome benefit and adds a failure point.
What I would not prioritize: education volume. More teaching hours on top of an unresolved access and follow-up problem does not change the trajectory; it feels productive and is not.
Explicit trade-off: if only one action survives the 48 hours, it is the 7-day appointment - it is the only one that creates a clinical safety net after the hospital's monitoring ends.
2. Sort these actions by who should own each, and describe how you would handle a disagreement with the hospitalist about delaying discharge to sort out the prior authorization.
Read the worked answer
Ownership by competence, not convenience:
- 7-day appointment with transport plan - CNL/care coordinator owns it, because it requires cross-system scheduling and a transport solution (bus route needs two transfers; ask whether the daughter can do the first visit).
- Reconciliation and deprescribing - pharmacist leads, hospitalist retains prescribing authority, bedside nurse validates against what the patient actually takes at home.
- Prior authorization and cost - social work plus pharmacy; CNL sets the deadline.
The disagreement: I would not frame it as "delay discharge." I would bring a specific, time-boxed question: "The apixaban-adjacent item on his list is unaffordable at $240. Do you want him discharged on an alternative today, or should we hold until tomorrow noon for the prior auth?" That converts a values conflict into a clinical decision the prescriber owns.
If the hospitalist still discharges tomorrow: that is within their authority. My obligation shifts to harm reduction - document the unresolved access issue in the discharge summary, call the receiving pharmacy, and set a 48-hour follow-up call so a failed fill is caught before it becomes an ED visit. The failure mode I am guarding against is silence: a discharge summary that reads as if the plan worked.
3. Which single outcome measure would best tell you whether this coordination effort worked, and what balancing measure would you track alongside it?
Read the worked answer
Outcome measure: 30-day all-cause readmission - but interpreted as a rate with its denominator stated (readmissions per 100 discharges for this heart-failure cohort), and trended monthly on a run chart. I choose it because it is the outcome his pattern predicts and it is attributable to the transition, not to inpatient care.
Why not "discharge summary sent" or "follow-up scheduled": those are process measures. They tell you the work happened; they do not tell you it helped. Use them as leading indicators, never as the outcome.
Balancing measure: at least one measure that would catch harm from my own intervention. Two candidates:
- Time-to-discharge / length of stay - my coordination could quietly extend stays for everyone, which is a real cost.
- ED treat-and-release visits - readmission alone undercounts failure; a patient who bounces to the ED and goes home is a failed transition that never appears in the readmission numerator.
Caveat to state out loud: a single patient cannot demonstrate anything. If I want to claim the change worked, I need the cohort rate over several months and a comparison - otherwise I am telling a story about one good outcome and calling it evidence.