Internal and Hospital Medicine Case Studies
Healthcare5 Case StudiesUpdated Sep 2026
Internal and hospital medicine sits where many risks meet: patients who are sicker than their admission label suggests, handovers across shifts, and protocols that must be applied under time pressure. The five cases below come from settings that look ordinary from the outside. Each one turns on a decision made with incomplete information, and each one shows how a team's habits, staffing and local rules shaped the outcome. Read them as working documents. Note what was known at each point, who knew it, and which step could have been done differently without heroics.
Case Study 1
Slowly worsening breathlessness on a general medical ward
Brightwater Regional Hospital, a 420-bed teaching hospital with a general internal medicine service and an outreach team that reviews deteriorating patients.
The situation
Walter, a 71-year-old retired electrician, was admitted to the general medicine ward with a flare of chronic lung disease and some new confusion noticed at home. By the second morning his oxygen requirement had crept upward across several checks, and his respiratory rate had drifted from the low twenties into the high twenties over the day. Priya, the day nurse, recorded every set accurately, but the readings sat in separate fields of the chart. Kwame, the ward registrar, reviewed him at midday, noted that he looked tired, and ordered a repeat chest film and blood tests for the afternoon. Nobody read the trend as a whole until the evening team began handover.
The challenge
The team had to decide whether a gradual change across several shifts justified escalation under the early warning score protocol, even though no single reading crossed the threshold alone. Each clinician had a reasonable explanation for the number in front of them. The real task was to read the direction of travel, prioritize investigations that could separate a chest cause from a metabolic or neurological one, and settle who owned the escalation call.
Analysis
Teams grounded in Internal and Hospital Medicine usually begin by separating what the patient needs from what the chart happens to show. Walter needed breathing support that was rising, a level of alertness that was slipping, and a plan that depended on results not yet back. The clinical evidence was a trajectory rather than a value. A respiratory rate in the high twenties can look tolerable at one check and alarming when it is the fourth check in a row. The early warning score had been calculated at every observation, yet the handover summary averaged it across the day, which flattened the slope the team most needed to see.
Coordination broke down at the point where information changed hands, not at any single bedside. The escalation protocol treats a rising score or a new change in alertness as a trigger for bedside review by the outreach team, and it allows the nurse in charge to make that call. In practice Priya held the score but had been taught to route concerns through the registrar, and the registrar had not been told the trend was building. Applying the protocol would have meant treating the trend as the trigger, documenting the reasoning in the notes, and asking for a second review rather than waiting for the next round.
- 1Serial observations recordedEach set captured separately across the day, in different chart fields.
- 2Score averaged in summaryThe handover view flattened the direction of change over several hours.
- 3Investigations orderedTests planned for the afternoon without reviewing the slope of the readings.
- 4Nurse lacks direct triggerConcern routed through the registrar, who had not seen the trend.
- 5Trend recognized in eveningThe evening team identified the pattern well after the first rise.
Actions taken
- Add a trend line to the ward observation summary so each handover shows the direction of the last several sets of readings, not only the latest score.
- State in the escalation protocol that the nurse in charge can request an outreach review directly when the trend criterion is met, without prior agreement from a doctor.
- Run a short case review at the next ward meeting in which staff rebuild the timeline of readings and decisions, asking what each person could see at the time.
- Audit a sample of escalations each month for documented reasoning and the time from the first rising trend to a review, and report results to the ward governance group.
Results
Within the quarter the trend line was added to the handover template, and the escalation protocol was revised to name the nurse in charge as able to request review directly. In the first audit cycle, documented reasoning appeared in nine of twelve escalations, compared with three of twelve in the earlier sample, and the average gap between the first rising trend and a documented review became much shorter. Walter was moved to a higher level of care for a period and later went home. The team agreed the case mattered more for the process than for the outcome.
Questions for discussion
- Which part of the escalation protocol would need to change for the nurse's observations to trigger a review without waiting for a doctor?
- How should a team weigh a gradual trend against a single reading that sits just inside the threshold?
- What would you need to see in a handover summary to judge whether a patient is deteriorating across a shift?
Case Study 2
A pending result left off the night handover list
Calder Park Hospital, a 300-bed community hospital whose medicine wards use a structured handover standard for every change of shift.
The situation
Miriam, a 58-year-old school administrator, was admitted to a medicine ward at Calder Park Hospital for observation after a fall at home. Late in the evening shift a radiology report arrived in the results system describing an incidental finding that needed a follow-up plan. The night doctor, Tomasz, had already written his handover list and sent it at the usual time, so the report was not on it. The day team, led by Ines, a junior doctor, started with the list and did not open the results inbox for that patient until later in the morning. The report stayed unread until the afternoon.
The challenge
Separate the information that existed, the information that was passed on, and the information that was acted on, because the gap sat between the second and third. The question for the team was not whether a handover should include everything, since no list can. It was whether a result that changes the plan for a patient has a reliable route into the handover, and whether the day team had a habit of checking for it.
Analysis
Patient needs in this case were specific. Miriam was under observation for a fall, so the team's core questions were about her mobility, her cognition and whether anything had been missed that would change her plan. The radiology finding was clinical evidence with a clear implication: it called for a documented decision about follow-up, either a planned test, a referral, or a reasoned decision not to act. That decision had to be owned by someone. Under the structured handover standard, the outgoing doctor is accountable for listing pending results and their owner, and the receiving doctor is accountable for confirming them at the start of the shift.
Team coordination here depended on a single list and on a habit of checking the results inbox, and the list was the weaker safeguard. Handover lists are written from memory and from the notes already in the chart, so a result that arrives after the writer has finished is structurally at risk. The results acknowledgement policy said that a clinician must mark a result as seen within a set interval, but the policy did not say who was responsible for results that arrived during handover. The useful point is that this is an ownership gap, not a question of individual diligence, and the fix belongs in the process design.
- Late evening shiftRadiology report arrives with a finding that needs a follow-up decision.
- Night handover sentThe written list omits the report because it arrived after drafting.
- Morning review beginsDay team works from the list and does not open the results inbox.
- Afternoon discoveryA nurse notices the unread result while checking pending items.
- Same-day follow-up arrangedPlan agreed and documented once the finding is reviewed by the team.
Actions taken
- Add a required field to the handover template for results received during the shift, with the name of the clinician who will review each one.
- Require the receiving doctor to open every pending result for the handed-over patients at the start of the shift and record the outcome before the first ward round.
- Hold a joint handover review with the night and day teams to walk through the last ten handovers and find where pending items were lost or delayed.
- Track the share of handovers with an unacknowledged result at the start of the day shift each month, and set a target the ward can review at its governance meeting.
Results
After the review, the handover template gained a results section, and the ward agreed that the receiving doctor would confirm every pending item before the first round. In the next month's audit, the share of handovers with an unacknowledged result at the start of the day shift fell from roughly one in five to under one in twenty, based on the ward's own tracking. Miriam's follow-up was arranged the same day, and she was discharged after a planned review. The team did not treat the change as finished. It agreed to re-audit after the next busy period.
Questions for discussion
- Who on your ward currently owns a result that arrives after the handover list has been written, and how would a new doctor find out?
- What would make a handover list reliable enough that the receiving team would not need to check the results inbox separately?
- How would you tell whether a safeguard is working when no incident has occurred?
Case Study 3
Monitored beds and admissions in a surge week
Harlow Creek Hospital, a 260-bed regional hospital with one telemetry unit, a shared bedside ultrasound machine and a bed management office.
The situation
Over one busy stretch, the medicine service received more admissions than it could place. On a single afternoon, fourteen patients were waiting in the emergency department for a medicine bed, and only nine beds on the wards had continuous cardiac monitoring. The bed manager, Yusuf, could move patients to a step-down area, but that space was staffed for lower acuity and already carried a full list. Two bedside ultrasound scans that the ward teams had requested were stuck behind one shared machine. Staff were tired, and the nurse manager, Beatriz, warned that another night at full census would make the monitoring rounds unsafe.
The challenge
The team had to allocate monitored beds, step-down space and the scanner among patients with different needs, without a clear rule for who moved first. Triage had been done by feel, and the same patient could be placed differently by two different people. The challenge was to choose an approach that protected the sickest patients, kept the monitoring rounds safe, and could be explained to the patients who waited.
Analysis
Patient needs drove the sequence more than bed availability did. Some patients waiting in the emergency department needed continuous monitoring because of their observations or a recent change in rhythm, while others needed it only briefly. The clinical evidence for each patient sat in the observation charts and the latest review notes, not in the admitting diagnosis. A triage tool that sorted by diagnosis would have sent a stable patient with a serious label to the monitored bed and left a deteriorating patient with a milder label waiting. The monitored bed protocol relied on observation thresholds and the treating team's judgment, and both had to be applied at once.
Team coordination was the weak point. The bed manager, the nurse manager, the emergency physician and the ward consultants each held part of the picture, and the decisions were made in a series of phone calls rather than at one table. The local escalation protocol names the bed manager as the owner of allocation decisions, yet the monitoring rounds depend on nurses who had no say in which patient arrived. Careful allocation here means agreeing the criteria in advance, recording why each patient was placed where they were, and giving the nurse manager a clear point at which to say the unit cannot safely take another patient.
Actions taken
- Agree a written triage rule for monitored beds that ranks patients by current observations and the treating team's assessment, not by diagnosis, and review it with both nursing and medical leads.
- Set a daily allocation huddle at a fixed time, with the bed manager, nurse manager and an emergency representative, so that placement decisions are made together and recorded.
- Give the nurse manager a defined right to pause new monitored admissions when staffing falls below the agreed ratio, and log each pause with the reason and the time.
- Move non-urgent scan requests to a scheduled ward clinic so the shared machine stays free for urgent cases, and track how many urgent scans wait beyond the agreed interval each week.
Results
The service agreed the triage rule and the daily huddle within the first week of the review. Over the following stretch, the number of monitored admissions placed without a documented reason fell to nearly none, and the nurse manager used the pause right twice, both times on nights when staffing had dropped below the agreed level. Emergency department waiting times for monitored beds shortened in the middle of the week. Harlow Creek did not add capacity during this period. The change was in how the existing beds were allocated and how clearly the limits were recorded.
Questions for discussion
- How would you decide which monitored bed should go to a patient whose observations are stable but whose diagnosis is high risk?
- What evidence would convince a nurse manager that a staffing level is no longer safe for monitored admissions?
- Where should a bed allocation decision be recorded so that the next shift can understand why a patient was placed where they were?
Case Study 4
Uneven outcomes across two medicine teams under review
Ridgeline Community Hospital, a 180-bed hospital whose general medicine service is split into two teams that share a single admission pathway.
The situation
Over a recent period, the medicine service at Ridgeline Community Hospital reported uneven results across its two teams. Anonymized indicators showed that patients on one team were more often reviewed by a senior doctor within the first day, while patients on the other team were more often transferred unexpectedly to a higher level of care in the first two days. Team A and Team B had the same admission pathway on paper, but their documentation habits differed. The quality lead, Sanjay, had a dataset of several hundred admissions, a set of process measures, and a clinical governance group that wanted a recommendation before the next review.
The challenge
The difference could be real, a function of case mix, staffing or local habits, or it could reflect how the indicators were recorded. The challenge was to identify the process causes that were most likely to matter, design a change that the two teams would actually use, and set up a monitoring cycle that would show whether the change worked without making individual clinicians feel singled out.
Analysis
Patient needs are the real test of these indicators, since they only matter if they track what happens to people. The first indicator asks whether a senior doctor saw the patient early enough to change the plan, which is a reasonable proxy for diagnostic momentum. The second asks whether a patient needed a higher level of care sooner than expected, which is a signal that the early assessment missed something. Clinical evidence from the case notes suggested that the unexpected transfers clustered in patients admitted at the end of the day, whose initial review had been brief and whose trend data had not yet been recorded in the chart.
Likely process causes sit mostly in the handoff between the admitting team and the daytime review, and in how the admission pathway is applied. The pathway asks for a documented early review and a trend check, but it does not say who completes the trend check when admissions happen late. Team B's notes often recorded the admission plan without a written observation trend. Team A reviewed earlier, but its trend checks were often copied from the admission note rather than updated. Both point to the same gap. A useful change would fix the timing and ownership of the trend check, not ask either team to work harder.
- Define the measure Agree indicators, data source and what counts as real change.
- Test one change Run the trend field and late-admission rule in one block of admissions per team.
- Review the data Compare the pilot run chart with the baseline, adjusting for case mix where possible.
- Adapt or standardize Keep what worked, refine the rest, then extend to the whole service.
Actions taken
- Add a required trend field to the admission pathway, completed by whoever holds the patient at the time of admission, with the name recorded beside it.
- Agree a single rule for late admissions so that the trend check is repeated at the first review on the day team's round, whichever team the patient is under.
- Pilot the changes for one block of admissions in each team, with a named observer checking a sample of notes against the pathway each week.
- Report the indicators as a single shared run chart, with anonymized team labels, and agree in advance what change would count as a real improvement.
Results
The pilot ran for one block of admissions in each team. The trend field was completed in most of the sampled notes by the end of the first month, up from a small share before, and the repeated trend check at the day round became routine for late admissions. The share of patients with an unplanned transfer in the first two days came down on the team that had carried the larger gap, though the sample was small. The next step was a longer monitoring period, with the run chart reviewed at each governance meeting.
Questions for discussion
- What would you need to see in the notes to be confident that the trend check reflects the patient at the time of review?
- How would you separate a real difference between teams from one caused by case mix or by how the indicators were recorded?
- Why might a change that works in a pilot block of admissions fail when it is applied to the whole service?
Case Study 5
A proxy, a doubtful capacity question and a daughter's request
Meridian Vale Hospital, a 350-bed acute hospital with a medicine unit for older adults and a clinical ethics advisory service.
The situation
Ruth, an 84-year-old former teacher, was admitted to the older adults medicine unit at Meridian Vale Hospital after repeated falls at home. Her team was considering further investigation that would take time to arrange and might change her treatment plan, and they needed a decision about whether to proceed. Her daughter, Anita, asked the team to share the results with her rather than with Ruth, saying her mother had become confused and would be distressed. Ruth had told a nurse that she wanted to know everything. The unit's records held an old document naming Anita as a contact, but nothing clearly giving her authority over medical decisions.
The challenge
The team had to decide whether Ruth could make this decision herself, how much of the information she was entitled to hear, and what Anita's request meant in law and in practice. Those questions were tangled. A capacity assessment, a confidentiality decision and a proxy question each needed an owner, and the bedside team could not settle all three on its own.
Analysis
Start with what Ruth needs, which is beyond the information itself. She needs a decision about investigation made in a way that reflects her values, and she needs to be treated as a person who can still take part. The clinical evidence points to fluctuating attention, so the capacity question is decision-specific and time-specific. A person can have capacity to decline a test and lack it for a complex financial choice. The assessment should ask whether Ruth can understand the information relevant to this decision, retain it long enough to weigh it, and communicate a choice. Her earlier wish to know everything is evidence that should carry real weight.
Team coordination and the applicable protocols decide what happens next. The hospital's consent policy says that a patient who has capacity for a decision is entitled to direct the information shared about it, and that a family member can be involved only with the patient's agreement. The same policy allows a clinical ethics advisory service to be called when capacity, confidentiality and family wishes pull in different directions. Escalation is not a failure here. Naming the ethics service early gives the bedside team a shared route, and it protects Anita's relationship with her mother, whose wishes still have to lead.
Actions taken
- Ask a second clinician, ideally a senior colleague, to complete a documented capacity assessment for this decision, with the specific information that will be discussed written into the note.
- Hold a private conversation with Ruth, without Anita present unless Ruth asks for her, to confirm what information she wants and who she wants involved.
- Refer the question of Anita's authority to the hospital's legal and ethics advisory route, and record the advice received alongside the clinical decision.
- Agree with the ward team a short, shared script for explaining what can and cannot be shared, so that every clinician answers Anita and Ruth in the same way.
Results
The senior colleague completed the capacity assessment within the same day, and concluded that Ruth could decide about the investigation but needed information given in short sessions with written summaries. In the private conversation, Ruth said she wanted to know the results herself and wanted Anita to help her afterwards, not to decide for her. The team shared results with Ruth directly and with Anita only at Ruth's request. The advisory route confirmed that no document gave Anita authority over this decision. The reasoning was recorded, and the ward kept the shared wording.
Questions for discussion
- What would change in your own assessment if Ruth had expressed a different wish about sharing results on an earlier admission?
- How should a team decide when a family member's concerns about distress justify limiting information, and when they do not?
- Which parts of the shared wording would you expect to vary between clinicians, and how would you check that they stay consistent?
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.