A missed imaging result rarely begins with one dramatic mistake. More often, it starts with a referral placed during a busy visit, an unclear owner, an inbox that is already overloaded, and a patient who assumes silence means everything is fine. This medical error prevention case study examines how an outpatient clinic addressed those small breakdowns before they became patient harm, complaints, or liability exposure.
The case is a composite based on common outpatient practice workflows. Its value is not in copying another clinic’s exact process. It is in showing how leaders can identify high-risk handoffs, assign ownership, and measure whether safer work is actually happening.
The medical error prevention case study: a clinic with hidden gaps
A six-provider internal medicine clinic had grown quickly over three years. Its physicians were clinically experienced, its front desk team was well regarded by patients, and its electronic health record was fully implemented. Yet the practice manager kept seeing troubling signals: delayed test-result communications, medication discrepancies after hospital discharge, referral notes without specialist reports, and patient calls asking whether anyone had reviewed an abnormal result.
No single event had caused severe harm. That made the problem easier to postpone. Staff members could usually explain each individual delay: a covering physician was unfamiliar with the patient, a fax had not arrived, a task had been routed to the wrong pool, or the patient had not returned a call.
When the leadership team reviewed 60 recent charts involving referrals, diagnostics, and medication changes, it found that 22 had at least one incomplete follow-up step. In 11 charts, the responsible person was not obvious from the record. The operational problem was clear: the clinic had processes, but not dependable closed-loop processes.
Start with the workflow, not the individual
The clinic avoided framing the review as a search for who had been careless. That matters. If staff believe that reporting a near miss will lead to blame, they will report less, document defensively, and work around the system in silence.
Instead, the leadership team asked a practical question: where can a patient leave the workflow without a named person noticing? They mapped the patient journey from the moment an order or referral was created through result review, patient communication, treatment decision, and documentation.
This exercise exposed several weak points. Medical assistants could enter referral requests, but no one consistently checked whether the consultation occurred. Test results arrived through multiple channels, including the EHR, scanned documents, and external portals. Nurses were expected to call patients with routine findings, but the definition of routine varied by physician. Medication reconciliation after discharge was sometimes completed by a nurse and sometimes deferred to the next physician visit.
The lesson for practice leaders is straightforward: a policy is not a workflow. A workflow must state what happens next, who owns it, how completion is visible, and what occurs when the expected response does not arrive.
Six changes that reduced follow-up risk
The clinic introduced six connected changes over a 90-day period. None required a new enterprise software platform. Each required clearer decisions and disciplined follow-through.
1. Assign one accountable owner for every open loop
Every diagnostic order, referral, and post-discharge medication issue received a designated owner. The ordering clinician remained clinically accountable, while a nurse or medical assistant could own the operational follow-up task.
This distinction prevented an unhelpful assumption that someone else was checking. If the assigned staff member was absent, the EHR task queue was reassigned daily by a team lead. The clinic also created a rule that tasks could not be marked complete merely because a voicemail was left. Completion required documented patient contact, a documented escalation attempt, or a clinician-approved next step.
2. Create a result-triage standard physicians will actually use
Before the change, each physician handled results differently. Standardization did not mean removing clinical judgment. It meant defining a common minimum for communication and escalation.
Results were categorized as critical, actionable, or routine. Critical results required same-day clinician review and direct patient contact or documented emergency escalation. Actionable results required a documented plan and patient notification within two business days. Routine results could be released through the patient portal with a standard explanatory message, provided the patient had agreed to portal communication and there was no clinical reason to call.
The trade-off was real. Physicians initially worried that standardized categories would create more messages and reduce flexibility. In practice, the standard reduced repeated questions from staff because they no longer had to guess which results needed immediate attention.
3. Build a referral tracker that measures completion, not placement
A referral order is only the beginning of a care transition. The clinic’s previous report showed how many referrals had been created, but it could not show whether a specialist appointment occurred or whether the consultation note had returned.
The new tracker included the referral date, urgency, destination, scheduled appointment date, consultation status, report receipt, and next action. Staff reviewed open urgent referrals twice each week and routine referrals weekly. Patients who could not schedule due to insurance, transportation, cost, or specialist availability were flagged for clinician review rather than simply labeled “noncompliant.”
That last detail improved both safety and patient communication. An incomplete referral is sometimes a patient-access problem, not a patient-behavior problem.
4. Standardize medication reconciliation at high-risk moments
The clinic focused medication reconciliation on transitions that created the most confusion: hospital discharge, emergency department visits, new specialist treatment plans, and the first visit with a new primary care provider.
Staff used one medication comparison format: what the patient was taking before the transition, what changed, why it changed, and what the patient should do now. Uncertain items were escalated to the clinician instead of being guessed at or left for a future visit.
The team also changed how it asked patients about medications. Rather than asking, “Are you taking everything as prescribed?” staff asked patients to describe how they took each medication during a typical day. This exposed duplicate drugs, discontinued prescriptions still being used, and medications that patients had stopped because of cost or side effects.
5. Give patients a clear role in the safety process
Patient-centered care is not asking patients to manage a clinic’s administrative work. It is giving them clear expectations and a simple way to raise concerns.
At checkout, patients received plain-language instructions about what should happen next: when results were expected, how referrals would be scheduled, and when to call if they had not heard from the clinic. For high-risk tests, staff used teach-back, asking patients to explain the next step in their own words.
This approach reduced ambiguity without shifting responsibility away from the practice. The clinic still monitored open orders and referrals. Patients simply had better information to recognize when something did not seem right.
6. Review near misses in short, regular huddles
Once a week, the clinic held a 15-minute safety huddle. The agenda was limited to recent near misses, overdue items, and one workflow improvement. The goal was learning, not a lengthy committee meeting.
Examples included a pathology result routed to a former staff member, a referral sent to an out-of-network specialist, and a discharge summary scanned without being assigned for review. Each issue produced one question: what system condition allowed this to happen, and what small change will prevent recurrence?
The practice manager recorded actions, owners, and due dates. This was essential. Safety discussions that produce no operational changes can become another meeting staff learn to endure.
What the clinic measured after implementation
After 90 days, the clinic did not claim that errors had disappeared. It measured whether the new system was more reliable. The percentage of open actionable test results beyond two business days fell from 18% to 4%. Completed referral documentation improved from 63% to 86%. The number of medication discrepancies identified during post-discharge follow-up increased initially, which leadership viewed as a positive sign that the process was finding problems earlier.
The most useful metric was not a single percentage. It was the number of open loops without a documented owner. That figure fell sharply because the clinic had made ownership visible.
For smaller practices, the same model can begin with a simple daily report and a designated staff member. For larger groups, EHR configuration, centralized referral teams, and dashboard reporting may be appropriate. The scale changes, but the underlying controls do not: clear ownership, defined timeframes, documented escalation, and regular review.
Common mistakes to avoid
The first mistake is treating training as the full solution. Training matters, but people cannot remember their way out of an unclear, high-volume workflow. Build cues and accountability into the daily process.
The second is measuring activity instead of completion. Counting orders placed, calls attempted, or referrals sent can create a false sense of control. Measure whether the patient received the needed follow-up and whether the record shows the clinical decision.
The third is adding alerts without removing noise. If every issue is urgent, staff will begin to ignore alerts. Reserve high-priority flags for genuinely time-sensitive situations and design escalation paths that can be used during a busy day.
A safer practice is built one closed loop at a time. Ask your team which result, referral, medication change, or message would be easiest to lose today. Then make it impossible for that item to remain invisible.
