Cardiothoracic and Vascular Surgery Case Studies
Healthcare5 Case StudiesUpdated Sep 2026
The cases below follow teams working in heart, lung, and blood vessel surgery, where a patient can move from stable to critical within a single shift. Organizations, staff, and figures are fictional, but the pressures are familiar. The aim is not to find one correct answer. It is to see how patient needs, clinical evidence, team coordination, and written protocols shape what people do next. Form a view on each escalation, handover, or allocation question before reading the analysis. Local protocols differ, so treat the actions as a framework to adapt rather than a template to copy.
Case Study 1
Subtle changes on a post-operative ward overnight
Penhallow Cardiothoracic Ward, a 28-bed post-operative unit attached to a regional referral hospital.
The situation
Imelda was four days past a valve operation when the night charge nurse, Tomasz, noticed her breathing rate creeping up across three observation rounds. Her pain scores were steady and she reported no new chest discomfort, so the change looked small on any single chart. The overnight registrar, Priyanka, reviewed the trend at the bedside and found that her oxygen requirement had risen slightly while her urine output had fallen across the shift. The ward used a standard early warning score, and Imelda's total had moved into a band that prompts review but does not automatically call the outreach team. Priyanka documented the trend, requested routine repeat investigations, and planned to reassess at the next round.
The challenge
By the time the day team arrived, the repeat results were back, but nobody had linked them to the overnight trend. The surgeon on call was in the operating room, and the outreach nurse assumed the ward had already escalated. The team had to decide whether the picture justified an earlier senior review, and how to pass on the uncertainty without overstating it.
Analysis
Teams grounded in Cardiothoracic and Vascular Surgery usually start by separating what the patient needs from what the chart happens to show. Imelda needed stable breathing, an accurate fluid balance, and a clear owner for any change. The evidence was a slope rather than a single abnormal value, which is exactly the kind of pattern a single-point score tends to smooth over. Three modest changes in the same direction carry more information than one isolated reading, and the registrar's bedside review captured context that the numbers could not. The first question should be whether the trend itself meets the escalation threshold in the local protocol, not whether any one reading does.
The protocol in this unit allowed a registrar to hold a trend for reassessment, but only if the reassessment time was recorded and the outreach team was told. Neither happened reliably here. The night charge nurse had raised a concern verbally, the registrar had documented a plan, and the outreach nurse had seen an earlier score without the trend. Each person held a partial picture. A handover template with a line for trajectory, and a rule that upward movement on two consecutive rounds prompts a direct call to outreach, would have turned three partial pictures into one decision. The lesson concerns where information was allowed to pool.
Actions taken
- Adopt a handover field that records the direction of each observation over the previous three rounds, not only the latest value.
- Set a rule that two consecutive upward movements trigger a direct call to the outreach team, with the call time written in the notes.
- Hold a weekly audit of overnight reviews where a score rose without escalation, recording reasons in a shared log for the following month.
- Run a short simulation for night staff using anonymized trends from the unit, practicing the call and the handover wording together.
Results
Within the first month of the revised template, outreach calls for trend-based concerns rose to a steady weekly count. The audit of overnight reviews found that most missed escalations had been handled by one person without shared sight of the chart. Two further near misses were recognized early, both on night shifts, and both were resolved on the ward. Staff said the trajectory line made handover shorter, not longer. The unit did not claim fewer serious events, because the period was too short and the numbers too small to show that. It kept the audit running and reported the trend at the monthly governance meeting.
Questions for discussion
- Which elements of the overnight picture were visible to each clinician, and which only became visible when combined?
- How should a unit weigh a protocol's escalation threshold against the judgment of a senior clinician who has examined the patient?
- What would a handover template need to contain before you trusted it to carry a trend across a shift change?
Case Study 2
A handover gap between recovery and the ward
Calderwood Heart and Lung Unit, a combined cardiothoracic ward and recovery area within a teaching hospital.
The situation
After an elective lung operation, Desmond was transferred from the recovery area to the ward at the end of an afternoon list. The recovery nurse, Aaliyah, gave a verbal handover to the ward nurse, Lorenzo, but the written summary used a template from another ward, one with a field for outstanding tasks that the receiving ward did not use. Aaliyah assumed the bedside sheet would carry the pending items, and Lorenzo assumed the anesthesia notes would hold the outstanding plan. Neither was wrong about the document, yet the outstanding items appeared in neither place. The gap surfaced only when the evening doctor asked why a planned review had not been booked.
The challenge
The ward had two working days before the planned discharge review, and the consultant wanted a structured account of what had and had not been handed over. The team needed to find which safeguards had failed, whether any risk had reached the bedside, and how to change the process without adding paperwork that staff would ignore. Their findings would also go to the governance committee, so the account had to be accurate rather than tidy.
Analysis
Start with what the patient needed in the first evening after transfer: a clear record of observations, a named clinician responsible for review, and a list of tasks deferred on purpose. Desmond's chart showed observations were recorded on time, so monitoring itself was not the failure. The gap sat in the deferred tasks. In handovers, what goes missing is usually the pending list, because a verbal exchange that skips it feels complete to both parties. The evidence available to the team was the document trail, the list of tasks due that evening, and the time stamps on each entry. Read together, they showed where the chain broke.
The local handover standard required a structured verbal exchange and a written summary on the unit's own template, with a read-back of outstanding tasks. Both steps existed, but the two wards had drifted into different versions of the form, and nobody had checked whether the template matched the receiving team's workflow. The safeguard that should have caught the gap was the read-back, and it was performed as a formality rather than as a check. A protocol is only as strong as the points where two people must agree on the same words. This one relied on each nurse assuming the other held the list.
Actions taken
- Reconcile the two ward templates so that the pending tasks field appears in the same position on both forms, and retire the older version.
- Require the sending nurse to read the pending list aloud and the receiving nurse to repeat each item back, with both initials recorded against each line.
- Add a measurable target: every transfer from recovery to ward should have zero unlisted pending tasks in a weekly sample of twenty handovers.
- Hold a joint walkthrough with both wards each quarter, tracing one real transfer from the recovery desk to the bedside and noting every hand point.
Results
The revised template went live in the second week of the following month. The first weekly sample of twenty transfers found two handovers with an unlisted pending item, both corrected on the same shift. The read-back became a short spoken exchange rather than a tick box, and staff said the wording was easier to remember than the old form. The unit tracked the zero-unlisted target for the rest of the quarter and recorded one miss in the third sample. No harm was traced to the original gap, and the revised process stayed in place while the sample grew.
Questions for discussion
- Which of the safeguards in this case depended on memory, and which depended on a document that could be checked later?
- How would you tell whether a read-back step has become a formality, and what evidence would change your view?
- Who should own a shared template when two wards use the same process but report to different managers?
Case Study 3
Choosing between urgent referrals and scheduled lists
Ferncliff Thoracic Service, a surgical program that shares a cardiac imaging suite with two other specialties.
The situation
In the third week of a busy quarter, the operating list was full, two urgent referrals arrived from a partner hospital, and the shared cardiac imaging suite had only one free session each afternoon. The service coordinator, Rhiannon, had to decide how to use that time and the staff who supported it. Three surgeons had scheduled cases, operating room staffing was fixed, and the ward of eighteen beds had seven patients awaiting step-down care. The urgent referrals were not immediately life-threatening, but each letter described a deterioration risk. Rhiannon had to balance the rota, the imaging queue, and a waiting list already longer than the service target.
The challenge
The options were not simple. Moving a scheduled case would disrupt a patient who had prepared for weeks, while leaving the urgent referrals in the queue risked a worse outcome for them. The decision also had to be defensible to the surgeons, the imaging department, and the families who would ask why one list had changed, and by the ward, which would absorb the consequences of any delay.
Analysis
The starting point is a clear view of need, not of queue position. Each urgent referral carried documented symptoms that had changed since the letter was written, and the referring clinicians could be reached for a short conversation. The scheduled patients had prepared for their dates and were fit for them, so moving a case would not be neutral for anyone. The useful questions are concrete: what would the imaging result change, how much does a delay alter the next step, and does the letter describe a trajectory or a single snapshot? A test that does not change management should not take a slot ahead of one that does.
The local capacity policy set a priority order for shared diagnostic time, but it assumed urgent cases had already been triaged, and they had not. Rhiannon had to bring the surgeons, imaging lead, and ward matron into one conversation, and that was where most of the delay sat. The policy also required any change to a scheduled list to be recorded with a reason, so later reviews could see why a case moved. Coordination failed less through disagreement than through several people holding separate parts of the decision. A short daily huddle with a shared list of pending decisions would have stopped the same bottleneck from recurring each week.
- 1Triage the referralsConfirm each letter describes current symptoms and a clear next step.
- 2Check the slot valueAsk whether the imaging result would change the planned management.
- 3Consult the surgeonsAgree which scheduled case could move with the least disruption.
- 4Record the changeLog the reason for any moved case so later reviews can follow it.
- 5Notify the familiesExplain the change before the patient hears it from the ward.
Actions taken
- Introduce a twice-daily triage huddle with the surgeons, imaging lead, and bed manager, using a shared list of decisions still pending.
- Ask referrers to state the current symptoms and the specific decision the scan would inform, so urgency is judged from the same information each time.
- Hold a fixed reserve of one imaging session each week for urgent referrals that arrive after the list is set, and track how often it is used.
- Report waiting times for urgent and scheduled cases separately each month, so the service can see whether reserved capacity is protecting the right patients.
Results
The reserve session was used in four of the first six weeks, and each time the urgent referral was seen before any scheduled list was disturbed. Two scheduled cases did move, both with the patients' agreement, and the families received a call from the surgeon rather than a letter. The waiting time report showed urgent cases were reviewed sooner on average, while scheduled waits grew by a small margin the service judged acceptable. The huddle continued for the rest of the quarter and was shortened once the list settled. According to the coordinator, the bottleneck had been the conversation, not the scanner.
Questions for discussion
- What information would you need in a referral letter before you were willing to reorder a scheduled list?
- How should a service decide how much capacity to hold in reserve, and who should be able to override it?
- When a scheduled patient is moved, what does the team owe that patient in terms of explanation, and who should give it?
Case Study 4
Inconsistent results across two cardiac sites
Greywater Cardiothoracic Network, a group of three surgical sites sharing one clinical governance team.
The situation
The network's governance lead, Oluwaseun, received an anonymized indicator pack showing one site had a higher rate of unplanned returns to the operating room than the other two. A second site showed a longer average stay with no obvious explanation. The indicators covered 2,400 operations over the last twelve months, but the sites recorded some fields differently. Oluwaseun had no direct access to case notes and could only work with anonymized extracts, local incident logs, and the protocols each site used. Clinical leads disagreed about whether the variation was real, whether it reflected case mix, or whether it came from how events were coded.
The challenge
The network needed a plan that would not turn a data question into a blame exercise. The team had to decide which differences were likely process causes, which needed clinical review, and how to design a change that could be monitored without a large new data burden. Any proposal also had to survive scrutiny from clinicians who distrusted the numbers, and from a board that wanted a decision before the next review meeting.
Analysis
The indicators point to outcomes that matter to patients, but they do not explain those outcomes on their own. Returns to the operating room and extended stays bring discomfort, delay, and the strain of repeated admission. Before anyone assumes a process fault, the review should check whether case mix differed, whether risk was recorded consistently, and whether the sites counted events the same way. Anonymized data can show a pattern across hundreds of cases, but it cannot show why one patient deteriorated. The evidence that matters is the link between a documented decision and a documented result, and that link is thinner in the extract than the team would like.
Both sites followed the network's post-operative protocol, but it was written at summary level and each site had filled in the detail differently. One site used a daily checklist that named an owner for discharge planning, while the other relied on the surgeon's end-of-day note. Coding compounded this. One site sometimes recorded a return to the operating room as a planned staged procedure, which hid it from the unplanned indicator. Coordination across sites depends on shared definitions as much as shared intentions. Without a common data dictionary and a named clinical lead at each site, the network could not tell whether it had a practice gap or a recording gap.
- Week oneIndicator pack shows unplanned returns higher at one site.
- Weeks two to threeLocal leads compare protocols and check coding against incident logs.
- Week fiveJoint review confirms a checklist gap and a coding difference.
- Week eightRevised checklist and shared definitions go live at both sites.
- Each month afterMonitored indicators reported to the board with site-level run charts.
Actions taken
- Agree a shared data dictionary for unplanned return, planned staged procedure, and extended stay, and apply it retrospectively to the last twelve months.
- Replace the summary post-operative protocol with a site-neutral daily checklist that names an owner for each open decision, including discharge planning.
- Set a monitored target for checklist completion at each site and review a random sample of ten charts per site every month for accuracy.
- Assign a named clinical lead at each site to report monthly run charts to the network board, with a stated threshold that triggers a joint case review.
Results
After the coding was reconciled, the gap in unplanned returns narrowed considerably. About a third of the original difference came from planned staged procedures recorded as unplanned at one site. The remaining difference was smaller but real, and the checklist audit showed that the higher-return site completed discharge planning items less consistently. Two months into the monitored cycle, checklist completion rose at both sites, although the sample was small. The network decided to keep the monthly run charts and review the threshold at the end of the next quarter rather than declare the problem solved.
Questions for discussion
- How would you separate a true process difference between sites from a difference in how the same events are recorded?
- What is the minimum shared definition a network needs before its indicators can be compared fairly across sites?
- When a monitored cycle shows improvement on a small sample, how long should the team wait before calling the change sustained?
Case Study 5
A family disagreement over a difficult heart decision
Aldergate Heart Center, a specialist cardiac unit with a dedicated patient liaison service.
The situation
Ruth, a retired teacher in her seventies, was being considered for a complex heart operation after several months of clinic discussion. She had asked that her daughter, Keisha, not receive detailed information without her permission. Keisha had attended two appointments, however, and the notes listed her as next of kin. The multidisciplinary team agreed the operation carried meaningful risk and that continued medical management had its own limits. Ruth was willing to proceed but wanted more time, while Keisha pressed for a faster decision. A nurse specialist, Bronwen, had heard both positions in separate conversations and was unsure what had been agreed.
The challenge
The team needed to clarify what Ruth had consented to, what she wanted shared with her daughter, and whether the disagreement changed the timing of any decision. Confidentiality and autonomy pulled one way, while the family's distress and the time sensitivity of the case pulled another. They also had to avoid making Keisha feel excluded from a conversation the notes had invited her into.
Analysis
Ruth's first need was a decision she could make without pressure, which meant understanding the options in plain terms and having time to test them. The team had documented how the options differed in their risks and benefits over time, so the evidence existed. Confidentiality was a need too. Her instruction about information was clear, and the notes did not override it simply because a relative had attended. What the team still lacked was a fresh account from Ruth herself, given alone, confirming what she wanted disclosed, what she wanted deferred, and whether her view on timing had changed since the last clinic.
The consent policy distinguished between a patient's agreement to a procedure and her permission to share information, and it required the record to show both. Here the notes listed Keisha as next of kin, an entry made at admission for administrative purposes, and several staff had treated it as permission. Bronwen was right to pause. The correct step was to bring the disagreement to the consultant and the liaison service rather than settle it informally in a corridor. Escalation was not a sign of distrust. It was how the protocol expected ambiguity to be handled, with a named person accountable for the final conversation and for the written record.
Actions taken
- Arrange a private conversation with Ruth, led by the consultant with the nurse specialist present, to confirm her wishes on information sharing and timing.
- Update the record so that next-of-kin status and information-sharing permission are separate entries, with the date and the staff member who confirmed each.
- Offer Keisha a supported meeting with the liaison service that explains what the team can and cannot share, without disclosing Ruth's private views.
- Escalate to the clinical governance lead if the disagreement affects the timing of a decision or if permission remains unclear after one private conversation.
Results
Ruth met the consultant alone and confirmed that Keisha could receive general information about the operation and recovery, but not Ruth's own views or detailed risk discussions unless Ruth chose to share them. She asked for one more clinic visit before deciding on timing, and the team agreed. Separate permission fields were added to her record that week, and Keisha attended the liaison meeting, where she said the boundary was clearer than the earlier conversations had been. The documented limits were then used at the decision meeting, and the unit now applies the same two-field approach to families in complex cases.
Questions for discussion
- When a patient's record lists a relative as next of kin, what should staff assume about that person's right to information?
- How can a team respect a patient's wish for privacy while still letting a worried family member feel included?
- At what point does a disagreement between a patient and a relative require escalation rather than another conversation?
Learning Resources
Healthcare CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.