Hematology and Oncology Case Studies
Healthcare5 Case StudiesUpdated Sep 2026
Blood and cancer services run on timing. A result that waits a day, a handover that drops one line, or a referral left in a queue can change the shape of a patient's week. The five cases below come from fictional services. Each follows a team that receives incomplete or shifting information, weighs it against local protocols, and makes a decision others must act on. For every case, ask what the patient needs, what the record shows, how the team coordinates, and which protocols apply. The diagrams summarize each team's reasoning so you can compare it with your own.
Case Study 1
Escalating a referral while symptoms change
Eskerfield Community Cancer Unit, a 22-bed outpatient and day service serving a mixed rural and suburban catchment.
The situation
Dario is 64 and retired from land surveying. His general practitioner referred him after several weeks of tiredness, heavy night sweats, and a swelling in the neck that came and went. The referral was triaged as routine and waited three weeks for a slot. When he attended a nurse-led review, Keisha, the specialist nurse, found that his weight had dropped over the same period and that a blood count sent by the practice showed two values outside the normal range. Neither had been acted on. The triage note listed his symptoms but did not mention the results. Dario had telephoned twice to say the sweats were soaking through his bedding.
The challenge
Keisha had to choose between three routes: an urgent same-week review, a direct call to the consultant on duty, or returning the referral to routine with a request for more information. Her unit's protocol for suspected blood cancers sets a review window and names the clinician who must see the patient, but it assumes the triage note is complete. Here it was not. The consultant's clinic was booked for the next four weeks, and the practice had not confirmed that anyone had read the abnormal results.
Analysis
Teams grounded in Hematology and Oncology usually begin with the patient rather than the protocol. For Dario the needs were straightforward: an honest account of what was being checked, a named person to call, and a review that did not depend on him telephoning again. The clinical evidence was incomplete but it was not ambiguous. Symptoms that persist over weeks, together with weight loss and abnormal values on a routine count, form a pattern that referral protocols exist to catch. The first task was to stop reading the referral as one document and to treat the sequence as the finding.
Coordination failed at the point where information changed hands. The abnormal count reached the unit as an attachment, and the triage note summarized the symptoms without the result. Keisha escalated by telephoning the duty consultant, which the local protocol permits when clinical details change during a waiting period. She then asked the practice to confirm in writing that the result had been reviewed and to name the clinician holding it, so that ownership was no longer assumed. The protocol's escalation criteria are short and sound. They assume that whoever triages a referral reads the whole record, and that assumption is the weak point to address.
- 1Referral receivedTriage note lists symptoms but not summarized results.
- 2Symptoms reassessedWeight loss and persistent sweats added to the pattern.
- 3Consultant calledTelephone review replaces waiting for clinic slot.
- 4Owner confirmedPractice confirms receipt and names result holder.
- 5Review datedPatient receives named contact and written appointment date.
Actions taken
- Re-triage the referral the same day with the consultant on the telephone, recording which symptoms and results were weighed and why the chosen pathway was selected.
- Ask the referring practice to confirm in writing that the abnormal blood count has been reviewed, and to name the clinician who holds that result.
- Give Dario a named contact and a written date for his next review, so that he does not have to telephone to learn whether his referral is moving.
- Add open-referral age and results not yet summarized to the weekly triage huddle, with a named owner for any referral that has passed its protocol window.
Results
Dario was seen by the consultant in the same week, and the investigations the team had agreed were arranged before that week ended. The referral audit then looked at every suspected cancer referral open beyond its protocol window. At the start of the quarter there were 17 such referrals, and 9 had abnormal results attached but not summarized. After the huddle change, the count fell to 3 by the end of the next quarter. The team did not call the change a success until two more referrals had passed the new check without a telephone escalation.
Questions for discussion
- Which features of the Dario referral should have triggered escalation at first triage, and what would you need to see in the record to be confident in that decision?
- When a referral changes clinically while it waits, who should own the decision to re-triage, and how should a protocol handle disagreement between the referring practice and the receiving unit?
- How would you tell whether the huddle check is changing decisions rather than adding another step that staff complete without thinking?
Case Study 2
A night review missing from handover
Marlow Ridge Hematology Center, a 30-bed inpatient ward with an attached day unit for patients on active treatment.
The situation
Agnes, 71, has spent four days on the ward after a planned admission linked to her treatment schedule. At 3 in the morning the night nurse, Oluwaseun, recorded a temperature change and called the on-call registrar, who asked for a repeat set of observations and a senior review before the morning round. Oluwaseun wrote the plan on the whiteboard and in the nursing notes. The verbal handover to the day team at 7 covered Agnes's mobility and her diet but not the pending review. The day nurse, Marcus, found the observations unchanged at mid-morning and no review logged.
The challenge
Marcus had to decide whether the observations he found, which looked stable, justified waiting for the senior review that had been requested overnight, or whether the missing review was itself the problem. The ward's handover template had a field for pending actions, but it had been left blank. Nobody on either shift had done anything wrong, yet the request had moved from one person to another without a clear owner. The team had to act before the afternoon, when the registrar's list would be full.
Analysis
A handover is a transfer of responsibility, and the patient's needs are only met if the responsibility arrives intact. Agnes needed a monitored response to a change she had already shown, and a senior decision that would settle whether further action was needed. The clinical evidence in the notes was clear about the request and unclear about its status. Experienced clinicians would read the gap as a missing task rather than a missing fact, because the observations were already recorded. That distinction shapes the fix. The system has to carry the open action forward, not just the data.
Safeguards existed but none of them caught this. The structured handover format asks for pending actions, yet the template allowed the field to be skipped without prompting. The verbal handover followed the order of the bed list rather than the order of risk, so the most time sensitive patient was discussed after the mobility and diet items. The escalation policy gives the registrar ownership of the review, but the day team owns the patient once the shift starts. A better design gives every open action an owner's name and a time, and makes the day nurse confirm each one aloud at the start of shift.
- Early night shiftTemperature change recorded and registrar called for review.
- Overnight, after the callPlan written on whiteboard and notes, not the handover template.
- Start of day shiftVerbal handover covers mobility and diet but omits pending review.
- Mid-morningDay nurse finds observations unchanged and no review logged.
- Before afternoon roundReview completed and decision recorded against a named owner.
Actions taken
- Add a required pending-actions line to the handover template, with an owner name and a time for each open item, and block sign-off while it is blank.
- Start every handover by reading open actions first, ordered by clinical risk rather than bed number, and have the receiving nurse repeat each one back.
- Audit a sample of twenty handovers each week for open actions with no owner or time, and report the proportion to the ward governance meeting.
- Set a target that unowned open reviews appear at handover on no more than one shift in ten, and revisit it after the first full quarter of audit data.
Results
Agnes was reviewed before the afternoon round, and the decision was recorded against a named owner. The first audit cycle examined 60 handovers and found pending reviews missing from 14 of them, with no owner recorded in 9. After the template change, the no-owner count fell to 2 across the next 60 handovers. The numbers are small enough that one difficult shift could move them, so the audit continues. Staff reported that the read-back added about four minutes per handover, which the governance group judged a reasonable cost.
Questions for discussion
- Where in your own handover routine would an open request be most likely to lose its owner, and what would you need to observe to find out?
- Should the escalation policy define ownership at the moment a senior review is requested, and what would change if the requesting nurse remained responsible until the review was documented?
- How would you weigh four extra minutes per handover against the risk that a single unowned action goes unnoticed for a full shift?
Case Study 3
Choosing where scarce chairs go during peak weeks
Brennan Court Oncology Day Unit, a 14-chair treatment service with a nurse-led telephone clinic that shares staff with the ward.
The situation
The day unit had 14 chairs and a planned schedule that filled every one of them. In the third week of a busy quarter, referrals from two new prescribing clinics added enough activity that the schedule overran by roughly a third each day. Two pharmacist verification slots were also lost to sickness. Among the waiting patients were Hassan, who needed a scheduled session that could not move without affecting his plan, and Lucía, whose symptoms since her last visit had prompted a request for an earlier review. The unit manager, Teodora, had until the end of the week to publish the revised timetable.
The challenge
The question was not simply how many chairs to fill. Teodora had to decide what could move, what could be done by telephone or in a nurse clinic without weakening care, and what should be protected even if the timetable slipped. Moving a scheduled treatment session was ruled out. Moving a symptom review was not, but only if the review could happen within the window the referring clinician had set. Any option also had to be explained to patients without implying that their need was less.
Analysis
Protecting care quality under pressure starts with sorting needs rather than slots. A treatment session carries its own safety checks and its own timing, and it is planned around the patient's wider care. A symptom review is different: its value depends on how quickly it happens relative to what the patient has reported. The clinical evidence Teodora had was limited to the referral notes, which described Lucía's symptoms but not how fast they were changing. Teams that plan around demand tend to ask for that information before choosing a route, because a telephone contact that arrives late leaves the same gap as a missed chair.
The unit's capacity policy ranks patients by clinical need and by the timing rules in each treatment plan, and it allows a nurse-led telephone review to stand in for a visit only when the referring clinician agrees. Either route needs that agreement, and in a busy week the agreement can take longer than the slot it is meant to free. A practical option is to pre-agree which review types can move, so the decision is made once for a category rather than again for each patient. Scheduling, pharmacy verification, and nursing rosters all need to be redrawn together, or the freed chair stays empty while the telephone queue grows.
Actions taken
- Sort every waiting request into one of three groups: protected treatment, symptom review with a time limit, and routine follow-up, and record the group on the booking.
- Move routine follow-up to the nurse-led telephone clinic for the affected week only, with the referring clinician's agreement recorded in the booking notes.
- Use the spare verification slot for the afternoon session, and cross-train a second staff member on verification so one absence does not remove a whole session.
- Track waiting time for each group weekly, and set a rule that no symptom review waits beyond its agreed window without a clinician being told.
Results
Hassan kept his scheduled session without a change to his plan. Lucía had a clinician telephone review on the day she asked, followed by an in-person visit within the window her referring clinician had set. Across the affected week, 38 routine follow-ups moved to the telephone clinic, and the unit recorded no cancelled treatment sessions. Three symptom reviews ran past their agreed window, and each was seen by a clinician the following day. Teodora said the grouping took longer to explain to the team than to use.
Questions for discussion
- When the day unit cannot seat everyone who needs a chair, what evidence should decide which symptom reviews move to telephone contact, and who should sign off that decision?
- What are the risks of pre-agreeing review types by category, and how would you test whether a category was drawn too broadly?
- How would you balance the cost of cross-training staff against the cost of a session that depends on one person's roster?
Case Study 4
Explaining uneven outcomes across three clinics
Stratton Hill Cancer Service, three linked clinics that share one referral pathway and one clinical record system.
The situation
In the previous quarter the service's quality group noticed that the time from referral to first clinical review varied widely across its three clinics. The northern site also documented fewer care plans in the shared record. The group pulled anonymized records for 412 patients referred in the same period, keeping only age band, referral route, and clinic site alongside a study number. The median wait to first review at the northern clinic was roughly double that of the other two sites. Unplanned admissions in the first months after referral were also higher there, although the group was careful about how far that difference could be read.
The challenge
The group had to explain the variation without leaping to a verdict about the northern team. Several explanations were plausible: a different case mix, a referral form that captured less information, a different documentation habit, or a shortage of clinic slots that pushed first reviews later. The data could not yet separate these. The group's sponsor, Ines, the clinical lead, wanted a proposal she could take to the three clinic leads within the month, and she wanted to know what would count as improvement before anyone changed anything.
Analysis
Outcome indicators are blunt on their own, so the first task is to ask what each number is standing in for. Time to first review reflects booking capacity, triage practice, and whether a patient's symptoms are recognized early. Documented care plans reflect habit as much as activity. A plan can exist in a clinician's head and still be missing from the record. The patient needs behind those numbers are continuity, a clear point of contact, and timely review. Any improvement work should therefore follow the patient through the process, not reward a clinic for moving its own figure.
The likely process causes are narrower than the headline numbers suggest. The northern clinic shares its nurse coordinator with a second service, so referrals wait for a shared inbox to be cleared before they are booked. Its referral form also lacks the symptom field that the other two clinics use, which means triage starts with a phone call more often. The regional referral protocol sets a review window but says nothing about who records the plan. Teams that can see this chain usually find that the fix is a shared template and a named coordinator owner, not more clinician time.
- Map the journey Time each handoff from referral receipt to first clinical review.
- Change one step Add symptom field and a shared care plan template at one site.
- Measure indicators Track time to review, plan completion, and unplanned admissions.
- Adjust and spread Decide on wider rollout only after the pilot period is reviewed.
Actions taken
- Map the referral journey at the northern clinic with the coordinator present, timing each handoff from receipt to booking to first review and noting where the record is updated.
- Add the symptom field to the northern referral form and make the three clinics use one template for the care plan, so that documentation stops depending on local habit.
- Assign a named coordinator owner for the northern inbox, with a daily clearance target and a report of referrals older than two working days.
- Run the changes at the northern clinic for one quarter, tracking time to first review, care plan completion, and unplanned admissions, then decide whether to spread them.
Results
The northern clinic's referral journey was timed over 30 referrals, and the largest delay was the shared inbox, not the clinic's own diary. After the pilot, the median time to first review fell by about a third, and the proportion of care plans in the shared record rose from well under half to most patients. Unplanned admissions were too few in one quarter to show a reliable change. The other two clinics were not changed during the pilot, which gave the group a comparison, although small numbers meant the comparison was indicative only.
Questions for discussion
- Which of the three indicators would you trust least as a measure of clinical quality, and what additional information would make it more useful?
- How should a quality group decide whether a small pilot result is strong enough to spread, given that the numbers in one clinic may be too small to support firm conclusions?
- What risks arise when a shared coordinator serves two services, and how would you reduce them without adding a new role?
Case Study 5
Sharing results with a daughter without formal authority
Corrin Lake Cancer Network, three treatment sites and a shared palliative support team coordinating care across a regional area.
The situation
At 58, Ruth has been treated across two sites in the network over the past year. She has asked that her results be discussed with her, not with anyone else, but she also brings her daughter, Bethany, to most appointments and has said she wants Bethany to be involved in decisions. Bethany holds a power of attorney document that has been signed but not registered, and she has begun telephoning the team to ask for updates. In the third week, Bethany asked for the latest results and said Ruth had been too tired to take them in. The lead clinician, Ambrose, paused the call and asked the nurse coordinator, Yuki, to check the record before replying.
The challenge
Ambrose had to weigh Ruth's stated wishes against Bethany's role, the unregistered document, and the team's duty to keep Ruth informed in a way she could absorb. Sharing results with Bethany without Ruth's agreement would breach confidentiality. Refusing all contact with Bethany could also cut Ruth off from help she may want. The next scheduled decision conversation was in four days, and Ruth had not yet been asked which family members she wanted present.
Analysis
Ruth's wishes are the starting point, and they are clear enough to act on. She has the right to decide what is shared, with whom, and when, provided she can understand and retain the information. The clinical evidence in the record does not say whether she can do that during a difficult week, so the team cannot assume either way. Teams that handle consent questions separate three matters. They ask what Ruth has told them directly, what she can take in on a given day, and what Bethany is entitled to hear on her own account. Those matters have different answers, and running them together is where errors start.
The network's confidentiality protocol allows information to be shared with a family member only on the patient's agreement or under a registered authority, and the document Bethany holds does not yet meet that test. The nurse coordinator, Yuki, can confirm this with the information governance lead without disclosing anything. Escalation should be proportionate. The decision conversation can address who Ruth wants present, and if Bethany's requests continue, the clinical ethics committee can advise on keeping communication open without breaching Ruth's wishes. The palliative support team has a role in preparing the conversation, not in deciding who receives results.
Actions taken
- Ask Ruth directly, at a time she chooses and with Yuki present, which people she wants to receive updates and what she wants kept from them.
- Tell Bethany, in a short call that discloses nothing, that the team cannot share results until the authority is confirmed, and offer a named contact for general questions.
- Ask the information governance lead to confirm in writing whether the unregistered document can be relied on, and what registration would require.
- Record Ruth's stated preferences in the shared record at the start of each decision conversation, and review them with her at the end so that her wishes are visible to every team member.
Results
Ruth chose to have Bethany present for the next decision conversation and asked that results be given to both of them together, which she recorded herself. The governance lead confirmed that the unregistered document could support sharing only after registration, and the family was given the registration steps in writing. Bethany's calls fell from several a week to one in the following fortnight. The team noted that the change came from Ruth's direct statement, not from the escalation, and that the ethics committee was not needed. Ambrose said the record now showed whose wishes applied to which decision.
Questions for discussion
- When a patient has stated preferences but may not be able to act on them consistently, what evidence would you need before changing how information is shared?
- How would you explain to a family member that the team cannot share information yet, without implying that the family member is being distrusted?
- What should a record show about a patient's wishes so that a new team member can apply them correctly on a day the patient is not at her best?
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.