Gastroenterology and Hepatology Case Studies
Healthcare5 Case StudiesUpdated Sep 2026
Gastroenterology and hepatology teams manage conditions that can shift within a single shift. A patient who looked stable at the morning review may be worsening by evening, and the signals that matter are spread across lab results, nursing notes, referral letters, and calls from family. The five cases below cover a delayed escalation, a handover gap, a capacity squeeze, an uneven outcome pattern, and a consent question outside standard pathways. Each is told by the people who had to act, with the protocols they used. Read them for the decisions and coordination, not for a single right answer.
Case Study 1
A slow slide in a patient with known cirrhosis
Marlowe Basin Hospital, a 310-bed teaching hospital with a hepatology clinic and an acute medical unit.
The situation
Ruth Ahmadi, 61, has had cirrhosis monitored in the hepatology clinic for four years. Over three weeks her clinic nurse, Carlos, logged a gradual rise in abdominal girth, poorer appetite, and two missed weight checks. The clinic plan was routine bloods and review in six weeks. In the second week she phoned the triage line to say she was confused in the evenings and her daughter had noticed yellowing of the eyes. The triage nurse, Deepa, booked a same-week telephone review rather than an acute assessment, since the note recorded no fever. That review was taken by a registrar who had not met Ruth and had only the clinic summary to work from.
The challenge
The telephone review confirmed what the summary suggested, but the registrar had no recent scan, no trend line for her blood results, and no sense of how fast the change was moving. Deepa had recorded the call as urgent, while the clinic letter still carried a routine label. The team had to decide whether the picture justified same-day assessment under local protocol, and who owned that decision when clinic, triage, and the on-call team each held part of the record.
Analysis
Ruth's needs in this case were clearer than the paperwork suggested. A person with cirrhosis who develops new confusion, abdominal distension, and jaundice is showing signs that the liver is failing to clear toxins and that infection or bleeding may be driving the change. The evidence that mattered was the trend, not any one value. Teams grounded in Gastroenterology and Hepatology usually start by asking what changed since the last full review, then check whether the plan still reflects that change. Here the plan had been written before the confusion started, so it no longer described her.
The local escalation protocol for hepatology patients sets a same-day in-person assessment when a patient with known liver disease reports new confusion, jaundice, or a rapid change in abdominal size. It also says the clinic lead must be told when a telephone review finds any of these. Neither happened cleanly. The triage nurse acted within her remit by booking a telephone review, but the review was not framed as a protocol check, and the registrar could not see the trend. The gap was not one person's error. It was the absence of a shared point where the clinic record, the triage log, and the review outcome met.
Actions taken
- Add a standing line to the triage protocol that any new confusion or jaundice in a patient with liver disease triggers a same-day in-person review, with no telephone-only option.
- Require the telephone review template to show the last three sets of results and the date of the most recent imaging, so the reviewer sees the trend before speaking to the patient.
- Copy the clinic lead on every triage contact that concerns a patient already on the hepatology list, and make the lead responsible for confirming the plan within one working day.
- Run a monthly audit of telephone reviews for listed patients, comparing the escalation decision with the protocol criteria, and report the agreement rate to the service governance meeting.
Results
Over the following quarter the clinic recorded 46 telephone reviews for patients on the cirrhosis list, and the audit found that 9 of them had met the escalation criteria without an in-person assessment being booked. The template change and the copied clinic lead reduced that to 2 afterward. Ruth was seen in person on the day of her call. The in-person review found a picture that needed inpatient care, and she was admitted under the hepatology team. The team read the case as a protocol that was not visible at the point of the phone call.
Questions for discussion
- Which signals in this case should have been visible to the registrar before the telephone review began, and how should the clinic summary have been read differently?
- How should a service decide whether a triage nurse's urgency rating or a clinic letter's routine label carries more weight when the two disagree?
- What would a protocol need to say for a telephone reviewer to feel confident that they, rather than the clinic lead, own the escalation decision?
Case Study 2
A handover gap across two shifts on a medical ward
Corvane Health Partners, a network of three hospitals sharing one electronic record and a common handover standard.
The situation
Mihai Popescu, 54, was admitted to a 28-bed medical ward with a suspected bleed from the upper gut and was reviewed by the admitting team during the afternoon. The day nurse, Janelle, recorded a falling hemoglobin on the chart and a plan for repeat bloods in the evening, but the verbal handover at shift change named only the admitting diagnosis. The night nurse, Tomas, took over with a full list of twelve patients and read the notes for the three patients he judged most unstable. Mihai was not among them. The repeat bloods were requested, but the result was filed to the day team's inbox, which nobody on nights checked until the morning.
The challenge
By the time the falling result was noticed, the patient had been on the ward for hours without the escalation point the admitting plan had set. Nobody had acted carelessly. The handover tool gave the outgoing nurse a free-text box, and the verbal summary followed habit rather than the written plan. The team now had to establish what had been missed, check on the patient immediately, and work out which other handovers that week had the same weakness.
Analysis
The patient's need in this case was simple to state and easy to lose in transfer: a falling blood count in someone with a suspected bleed needs a named person checking the next result and a clear threshold for calling the senior doctor. The evidence was already in the chart, which is where the handover should have started. A trend, a time for the next sample, and a stated action are the elements that turn a result into a decision. Without them, the night nurse received a number with no owner and no deadline.
The hospital's handover standard asks for a structured summary covering the working diagnosis, the pending results with their owners, and the actions that must happen before the next review. Here the structure existed but was used loosely. The verbal handover followed the diagnosis list, the pending result was filed to a team inbox rather than a named clinician, and the night team had no way to see which results were still outstanding. The protocol had the right questions. The coordination around it did not make those questions unavoidable, and the weakest link was the moment when the day nurse left and nobody yet owned the sample.
- Name the result Write the test, due time, and the clinician who will act.
- Assign an owner A named clinician, not a team inbox, receives every pending result.
- Read it aloud The receiving nurse repeats the outstanding item and its deadline back.
- Check the loop Audit whether each pending result was acted on before the next review.
Actions taken
- Replace the free-text handover field with required fields for pending results, each with a named owner, a due time, and the action expected if the value crosses an agreed threshold.
- Route every lab result for an admitted patient to the responsible clinician's task list rather than a shared team inbox, and block sign-off until the result has an acknowledging name attached.
- Make the outgoing nurse and the incoming nurse review the written plan together at the bedside for any patient with an open result, and document the time both have confirmed the plan.
- Track one measure monthly: the share of sampled handovers where every pending result shows an owner, a due time, and a recorded action. Set the target at 95 percent.
Results
Mihai's repeat result was acted on within the hour once the night team saw it, and the senior doctor reviewed him before the morning round. Over the following two months the audit covered 180 sampled handovers and found that 41 pending results had no named owner at transfer. After the required fields and bedside review began, that figure fell to 9 in the second month's sample. The same audit found 3 other patients whose results had sat unread for more than an hour, which prompted the inbox change across all three hospitals.
Questions for discussion
- Which part of a handover should carry the most weight when a result is pending: the diagnosis, the plan, or the named owner, and why?
- How would a ward measure whether a required field is being completed with meaningful content rather than a placeholder entry?
- What would you want the outgoing nurse to say first at a bedside handover for a patient whose result is still outstanding?
Case Study 3
Ranking a diagnostic waitlist under staffing strain
Harlow Point Outpatient Center, a community clinic running two diagnostic lists each week with a shared nursing pool.
The situation
Harlow Point Outpatient Center runs two diagnostic lists each week for patients referred with unexplained changes in bowel habit, abnormal liver tests, and anemia. Over the last quarter monthly referrals rose from 210 to 340, while two nurses left and the nursing pool fell from six to four. Waiting time for a list slot grew from five weeks to nine, and two referring practices began calling the clinic lead directly to ask for urgent slots. Some of those calls described symptoms that the referral form had not captured. The lead, Ngozi, had to decide how to use the four remaining nurses across the next four weeks.
The challenge
Every option carried a cost. Moving urgent cases up would protect the patients most at risk but push routine patients further back, some of whom were also showing warning signs in their referral letters. Cutting the list to fit four nurses would reduce the number of patients seen but might leave clinics running with unfilled slots. The team needed a method that could be explained to referrers and defended at the governance meeting, and that did not depend on whichever call came in loudest.
Analysis
The patients on these lists differed more than their referral categories suggested. Some had weight loss or blood results that pointed toward a condition needing investigation within days, while others had long-standing symptoms with stable readings and little risk of harm from a short delay. The evidence that separated them was in the referral letter and the most recent results, not in the length of time they had waited. Ranking by waiting time alone would treat a stable patient and a deteriorating one as equal. The referral form had been designed to capture symptoms, but it rarely recorded trend, which is where most of the urgency sat.
The clinic's access policy set a priority tier for urgent referrals, but it was written for a larger team and assumed a clinician would review each referral within two working days. With four nurses that assumption no longer held. Sorting was happening, but the criteria were applied differently by each nurse, and the calls from referrers went to whoever answered the phone. Coordination had become a question of who held the list on a given day. A practical answer lies in reallocating tasks: one person owns triage each morning, and the policy's tiers are applied from a single checklist rather than memory.
Actions taken
- Adopt one triage checklist built from the access policy tiers, printed on the referral form and mirrored in the scheduling system, so every nurse applies the same questions.
- Assign one nurse each morning as the triage owner, with a protected two-hour block, so referrals are sorted once and calls are routed to that person rather than to whoever answers.
- Ask referring practices to flag suspected weight loss, bleeding, or new jaundice on the form, and treat any flagged referral as a same-week review until a clinician has read it.
- Report the share of urgent referrals seen within the target window each month to the governance meeting, alongside the number of routine patients whose wait has passed the set limit.
Results
Within the quarter the morning triage owner was in place on every clinic day, and the referral form was revised to include the three flags. The median wait for urgent referrals fell from four working days to one. Routine waits did not fall as far, stabilizing at seven weeks, which the team accepted because the clinic could not add nursing hours without cutting another service. Two patients who would previously have waited in the routine queue were moved up after flagged referrals were read the same week. The lead reported these figures without claiming the problem was solved.
Questions for discussion
- Which option in this case protects the most patients at risk, and which protects overall fairness across the list, and can both be measured?
- How should a clinic weigh a referrer's direct call against a written triage tier when the two point in different directions?
- What would it take for a triage owner's decisions to be reviewed without adding a meeting that consumes the same nursing hours?
Case Study 4
Uneven follow-up outcomes across a liver service
Greyfern Community Hospital, a 240-bed district hospital whose liver service covers a main clinic and two satellite sites.
The situation
The liver service at Greyfern Community Hospital sees about 1,200 patients a year across its main clinic and two satellite sites. Over the last reporting cycle the service's own indicators showed that patients seen at one satellite site waited longer for a follow-up review and were more often lost to follow-up. The gap was not large in any single month, but it was steady across the year. Clinic leads had noticed it informally, and a quality nurse, Yusuf, was asked to pull anonymized records for the last 12 months. He removed names and dates of birth before sharing the extract with the review group.
The challenge
The review group needed to decide whether the difference reflected patient mix, which was real and partly outside the service's control, or process, which the service could change. The indicators were not yet comparable enough to settle that. Some records lacked a recorded reason for a missed follow-up, and the two satellite sites used different booking codes. Staff were also wary that the review would turn into a search for blame.
Analysis
Patients with chronic liver conditions need reliable follow-up because their risk can change between visits without a clear symptom. The indicators the service tracked were the proportion seen within the target interval, the proportion with a documented plan after each review, and the proportion of results acted on within the agreed window. Each measure is only as good as the data behind it. Before concluding that the satellite site performed worse, the group had to confirm that a missed review was truly missed and not transferred to another clinic that had not updated the record. Evidence here depends as much on good records as on good tests.
The service's governance framework requires a monthly review of follow-up indicators, with each site reporting against the same definitions. In practice the satellite site used a different booking code for transferred patients, so transfers counted as missed reviews, and the main clinic had no field for documenting a patient's reason for not attending. The problem sat between sites rather than within either one. Each team followed its own process faithfully. The protocol assumed that definitions were shared, and they were not. Fixing this would need a meeting of both leads, a short agreed dictionary of terms, and one owner for the data definitions.
Actions taken
- Agree one set of definitions for transfers, non-attendance, and missed reviews, and publish it to both sites before the next monthly indicator run.
- Add a required field to the main clinic's record for the reason a patient did not attend, with a short fixed list plus a free-text option for anything unusual.
- Assign a single data owner to maintain the definitions, reconcile transfers each week, and flag any record where the site and the clinic disagree about the outcome.
- Run a six-week cycle: change one process at the satellite site, compare its follow-up rate with the main clinic's on the same definitions, and review the result at the governance meeting before extending it.
Results
After the definitions were agreed and the transfer codes reconciled, the satellite site's measured gap narrowed by roughly half within the next two cycles. About a third of the apparent difference had been recording artifact, not missed care. The remaining gap was real and traced to appointment availability on one afternoon each week, which the service changed by moving a clinic session. The reason field was used for 85 percent of non-attendances within two months, and the group now reviews both sites on the same page each month.
Questions for discussion
- Which of the contributing factors would you test first, and what evidence would convince the governance group that it had been resolved?
- How should a service adjust its indicators when part of a gap comes from how records are coded rather than how care is delivered?
- What risks come with a six-week process change at one site, and how would you watch for harm to patients during the trial?
Case Study 5
A family request that conflicts with the patient's wishes
Lowther Creek Health, a rural health system with one teaching hospital and a regional outreach team for hepatology and palliative care.
The situation
Arturo Vance, 72, has advanced liver disease and has been clear with his hepatology team that he wants to be told the facts of his condition directly, but he has also asked that decisions about tests be made with his wife, Lila, present. Over the past month Lila has phoned the clinic several times asking the team to hold back news from him about a worsening result. The consultant, Ingrid, and the social worker, Malik, met Arturo alone in the second week. Arturo said he did not want his wife making decisions about what he was told, and asked the team to keep his wishes in the record.
The challenge
The team now held two legitimate concerns. Arturo had capacity on the assessment recorded, and his instruction about information was clear. Lila was not acting out of malice. She was frightened and believed that a worse picture would make him give up. The team had to respond to her worry without overriding him, decide what could be shared with her and with his consent, and agree a single voice for future calls before the next results were due.
Analysis
Arturo's needs were both clinical and relational. Advanced liver disease can change quickly, and he needed information that let him make decisions about his care, including what to expect from his next results. The evidence in the record supported that his instruction was informed and recent. A recorded capacity assessment, the consultant's notes, and his own words in the meeting all pointed one way. Under the consent framework, the team owed him the information he asked for and had to be cautious about sharing it with anyone else without his agreement. Those duties held even though her distress was real.
The service's communication policy asks teams to record a patient's named contacts and preferences, and to agree in advance how information is shared with family. Arturo's file did not include a named contact. Lila had become the default contact because she answered the phone each time, a pattern the policy warns against. The social worker and the consultant agreed that the next step was a structured family meeting only if Arturo consented, and that the ethics support service should be consulted if the disagreement continued. Escalation here was not a sign that the team had failed. It was the protocol doing its job.
- 1Confirm the patient's instructionCheck the recorded capacity assessment and his stated wishes about information.
- 2Meet the patient aloneGive him space to restate preferences without the family present.
- 3Hear the family concernAcknowledge Lila's fear and explain what can and cannot be shared.
- 4Consult ethics supportEscalate if the disagreement continues or the patient's capacity is questioned.
- 5Record and reviewDocument each decision, the reasons, and the agreed contact for future calls.
Actions taken
- Add a named-contact and information-sharing section to the hepatology intake form, with the patient able to name one person, several people, or no one, and to limit what each may hear.
- Set a standing rule that the first call from a family member about a patient's results goes to the named clinician, who checks the patient's sharing preferences before any information is given.
- If the patient agrees, hold a planning conversation with the family member within the week, and offer her a separate session with the social worker for her own concerns.
- Bring the case to the ethics support service when the disagreement persists, and record in the file who was consulted, what was advised, and how the advice was applied.
Results
Arturo agreed to a planning conversation in which Lila was present for part of the discussion, and the social worker met her separately afterward. He confirmed that she could receive general updates but that specific results would be discussed with him first. The team recorded a named contact and a sharing plan in the file. Lila's calls dropped in the following weeks, and she reported feeling more included. The ethics service reviewed the file and agreed that the team had acted within its duties. The process was documented so the next team member could follow it.
Questions for discussion
- Where should the line sit between a family member's legitimate worry and a patient's right to control what he is told, and who should draw it?
- What features of the intake process would have made it clear, before the first call, who could receive results on Arturo's behalf?
- How would the team know that a patient's stated preferences were still current several months into a long illness?
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.