Post-Discharge Follow-Up: Why the 72-Hour Window Matters Most

Across study populations and care settings, the evidence points to the same concern: the first 72 hours after hospital discharge carry the greatest risk for readmission, medication error, and unmet follow-up need. Patients discharged on a Friday afternoon may be unable to reach their primary care provider until Monday morning with questions about discharge instructions, leaving them especially vulnerable. The inpatient-to-outpatient transition is where care coordination matters most, particularly when "someone should call this patient" does not become "someone actually did."

The hospital-side transition requires different tools. CTPathways focuses on the outpatient side: what happens when the patient gets home, and whether the primary care practice has a working protocol to close the loop within the first 72 hours.

What the primary care practice knows about a discharge

Information flow between hospitals and outpatient primary care practices varies widely. Some hospitals have reliable EHR-to-EHR discharge notifications, but many do not. Often, a practice learns of a hospitalization when the patient calls for a follow-up appointment, a paper discharge summary arrives by fax, or a provider happens to see an ADT (Admission, Discharge, Transfer) notification through a connected EHR.

When the admitting hospital and practice use different EHR networks, discharge notification may be delayed. A patient discharged on Wednesday may not be reported to the practice until Thursday or Friday, depending on the fax workflow. Within the 72-hour window, that delay is clinically significant.

The first coordination issue is informational: how and how quickly does the practice learn about a discharge? The second is actional: once informed, what process ensures outreach within 24 hours?

Building a 24-hour outreach protocol in a small practice

Literature on post-discharge follow-up identifies a care team phone call within 24 to 48 hours as the gold standard. For a small primary care practice managing 1,800 to 2,500 active patients, this is achievable with a clear protocol, reliable alert mechanism, and designated role. It is not achievable when the plan is "someone will call when they have time."

A workable protocol needs four elements: a trigger for reaching the practice, a responsible role, a standard script or check-in questions, and documentation in the patient record. Without all four, gaps will emerge during high-volume periods, staff absences, and outreach to patients who are hardest to reach.

For practices using athenahealth or eClinicalWorks, ADT workflows can route discharge alerts to a designated task queue. Without that capability, a patient or family member calling for an appointment often becomes the trigger, and the 24-hour window may be closed before the practice knows about the discharge. In that setting, the hospital discharge coordination team is the main information source, making an active working relationship worthwhile.

What a 72-hour check-in needs to cover

A 72-hour check-in is not a review of the entire hospital course. It addresses transition problems: medications not picked up, unclear discharge instructions, worsening symptoms, and unscheduled follow-up appointments. These are the four most common causes of early readmission that a care team contact can address.

A structured call covering those four areas takes roughly 5 to 8 minutes. A message asking the patient to answer "yes" or "no" about filling medications and scheduling follow-up takes about 30 seconds, giving the coordinator the key data in under a minute. No response is also informative: a patient who does not respond to outreach in the first 48 hours is at elevated risk and warrants a direct phone call.

Check-in questions should reflect the discharge diagnosis. After heart failure exacerbation, confirm weight monitoring, diuretic adherence, and a scheduled cardiology follow-up. After an elective surgical procedure, review wound care instructions and confirm the surgeon follow-up date. Templates organized by discharge category produce meaningfully better information than a generic protocol.

A concrete scenario: the 72-hour gap

Consider a patient with COPD admitted for an acute exacerbation and discharged on a Thursday. Instructions included a new inhaler regimen, a prednisone taper, and follow-up with his primary care practice within 5 days. The practice had no ADT notification configured for his insurer's hospital network and learned of the admission when he called the following Tuesday to schedule an appointment.

By Tuesday, he had returned to the emergency department. He had not filled one inhaler because he did not know which pharmacy had received the prescription. A family member picked up the prednisone and misread the taper instructions, leaving it incomplete. A Friday call could have resolved both issues.

This scenario is not unusual. Delayed discharge information, a medication access barrier, and a misread instruction together account for a large proportion of heart failure and COPD readmissions at small practices. The solution does not require sophisticated technology, only a protocol that identifies the gap early enough to close it.

Where information flow must improve

At the primary care level, post-discharge coordination is under-resourced relative to its clinical importance. Hospitals invest heavily in transition care nurses and discharge planning teams, while small independent practices receive little infrastructure support for managing the transition from their side. Patients return to the community with a discharge summary, medications, and the assumption that their primary care team is informed and prepared.

Every primary care practice does not need a hospital-grade transition care program. A reliable alert mechanism and a 5-minute check-in protocol for the 72-hour window can catch a material proportion of avoidable readmissions among complex patients. Building that protocol requires little compared with the patient outcome and downstream care utilization costs of not having one.