Mental Health, Psychiatry and Behavioral Support Case Studies
Healthcare5 Case StudiesUpdated Sep 2026
These five cases follow mental health and behavioral support teams through decisions that rarely arrive with clear answers. The services are fictional, but the pressures are familiar: symptoms that drift in ways easy to misattribute, handoffs that drop a critical detail, demand that outruns staff, outcome data that refuses to line up, and consent questions that sit between a person's wishes and a team's duty of care. Form your own view of the evidence before turning to the analysis. Local protocols matter throughout, so note where a team relies on them and where it must use judgment.
Case Study 1
A changing picture during a depressive episode
Larkspur Community Mental Health Service, a publicly funded team of 14 clinicians covering a mixed urban and rural area of about 38,000 residents.
The situation
Thomas is a 58-year-old retired bus driver who has been under the service's care for a depressive episode for about four months. At his review in the third week of the current block, his primary nurse, Priya, recorded that he had asked the same question twice, looked unsteady when he stood, and seemed less sure of the date than he had been a month earlier. His sister, who attends most reviews, reported that he had stopped eating breakfast. The team had reasonably read the earlier fatigue as part of his mood. Priya's notes show she was unsure whether the new picture belonged to the depression, to a physical problem, or to both.
The challenge
The service has no physician on site. Its escalation protocol sets three tiers: a same-day discussion with the consulting psychiatrist, a same-day medical review through the general practice link, and emergency transfer for acute change in consciousness. Priya's instinct pointed to the middle tier, but the consulting psychiatrist was on leave and the covering colleague had not met Thomas. The next scheduled review was nine days away, so waiting for the usual route would have meant waiting for that date.
Analysis
The key question for the team was not whether Thomas was depressed, but whether the new features fit the mood episode. Confusion, unsteadiness, and reduced eating in a man of his age are changes that a physician needs to assess, and a mood explanation does not settle that. The evidence was thin but specific: repeated questions, a sister who could place the change in the weeks before the review, and a nurse who had recorded the same pattern twice. Patient needs covered safety at home, nutrition, and an honest account of what the team did not know. Priya's notes were strong enough to justify escalation without waiting for certainty.
Teams grounded in Mental Health, Psychiatry and Behavioral Support tend to read a protocol as a set of thresholds that someone must own, not a list of permissions. The Larkspur protocol placed a same-day medical review at the second tier for any new cognitive or mobility change in clients over fifty-five, and Priya had the authority to request it without the psychiatrist. The gap was coordination. Nobody owned the decision while the psychiatrist was away, and the covering colleague had not yet read Priya's notes. A sound handover would have named who decides, by what time, and what happens if no answer comes.
- Third week of blockNurse records repeated questions, unsteady stance, and reduced breakfast.
- Same afternoonSister confirms the changes began weeks before the review.
- Two days laterMedical review identifies a physical cause for the changes.
- Original review dateOriginal plan would have waited for this appointment.
Actions taken
- Priya calls the general practice link the same day to request a medical review, sends her written summary of the changes with the sister's account, and records the time of the request.
- The covering psychiatrist reads the file before the end of the day, confirms the escalation tier in writing, and names the clinician who will check the outcome of the medical review.
- The team adds a change prompt to the caseload dashboard, so that any new cognitive or mobility change in a client over fifty-five creates a review task within one working day.
- The service lead audits escalation records each quarter, counting how often a decision waited for one named clinician to return, and reports the count to the governance meeting.
Results
The medical review took place two days after the escalation and identified a physical cause for the cognitive and gait changes that had not been considered. The mood plan was revised to account for it, and Thomas's sister received a written list of changes to report between reviews. He stayed under the service for several more months. The first quarterly audit found that four of eleven similar changes had waited more than a week for a named decision owner, which confirmed the case for the dashboard prompt.
Questions for discussion
- Which parts of Priya's escalation were within her authority under the protocol, and which needed the psychiatrist's agreement?
- How should a mixed team share responsibility for judging whether a change belongs to a mood episode or to something else?
- What would a useful handover have needed to contain for the covering psychiatrist to act on the same afternoon?
Case Study 2
A detail left in the notes and not the handover
Cedar Hollow Inpatient Unit, a 22-bed adult psychiatric unit attached to a general hospital, with about 1,140 admissions over the past twelve months.
The situation
Marcus is 34 and was admitted to the unit for a short crisis stay after a period of escalating distress at home. His observation level was set at the routine tier at admission, and the day team reviewed it twice without change. On his fourth day, his nurse, Alisha, documented that he had asked where the stairwell exits were, had given a personal item to a visitor, and had said he was tired of being a burden. The entry was made at 16:10. The shift ended at 19:00, and the verbal handover to the night team covered his sleep and his leave requests. The written summary sheet, completed from the previous day's template, still listed the observation level as unchanged.
The challenge
The unit's handover protocol uses a structured tool with a verbal segment for any change in risk status during the previous four hours. The entry at 16:10 fell inside that window, but it was not read aloud. The night nurse, Dylan, worked from the summary sheet and from the verbal handover, and neither mentioned the change. At 21:30 an assistant found Marcus at the stairwell door with his coat on. No harm occurred. The team had to explain why a documented change had not reached the person responsible for observation.
Analysis
The evidence in this case was not a single symptom but a cluster. A question about exits, a gift of a personal item, and a statement about being a burden are each ambiguous alone. Together, and placed against an admission for escalating distress, they describe a change in risk that the observation plan should have reflected. Marcus's needs went beyond safety: he was a man in distress who had said something important about how he felt, and the unit owed him a team that could hear it and respond in proportion. Information about his change belonged in the team's shared plan, not in the notes of a single nurse.
The handover protocol did what it was designed to do. It asked for a verbal segment on risk changes and a written summary, and both existed. The failure was in the join between them. The summary template carried forward yesterday's observation level because nobody was prompted to update it, and the verbal handover was built around sleep and leave, the usual topics. A protocol that depends on a nurse remembering to say the one unusual thing will fail on the shift when the unusual thing matters. Safeguards that could have caught it included a charge nurse sign-off on observation level changes and a rule that any documented risk change triggers a bedside review at the next round.
Actions taken
- Add a required field to the handover summary that pulls the observation level from the live chart, so the written sheet cannot repeat yesterday's entry without a new sign-off.
- Build a verbal handover prompt that asks the outgoing nurse to read aloud any risk entry made in the previous four hours, with the time and the author named.
- Require charge nurse sign-off within thirty minutes of any documented risk change, and a bedside check at the next round until the observation level is reconfirmed.
- Audit forty randomly chosen handovers each month for whether every risk change in the prior four hours was read aloud, aiming for ninety-five percent or higher, and report results weekly.
Results
Within six weeks the unit had introduced the verbal prompt and the linked summary field. The first monthly audit of forty handovers found that twenty-eight of thirty-one risk changes were read aloud, a rate of about ninety percent, short of the target but a clear improvement on the earlier rate of about sixty percent. Charge nurse sign-off was used for every risk change logged during that period. Dylan and Alisha reviewed the case together and wrote the changes into the unit's handover guide. No further incidents of this type were recorded in the following quarter.
Questions for discussion
- Which single element of the handover protocol would have been most effective at preventing this gap, and why?
- How should a unit balance a verbal handover that is short enough to be reliable against one that captures every unusual detail?
- What would you expect to see in the summary template that distinguishes a carried-forward entry from a freshly confirmed one?
Case Study 3
Matching therapy capacity to referral demand
Riverbend Psychological Therapies Team, a stepped care service with nine therapists and a referral list of about 2,300 people.
The situation
The Riverbend team offers short-term psychological therapy for adults with anxiety and low mood. In the most recent count, the waiting list held about 2,300 referrals, and the team had nine therapists in post, two of them part time, with one permanent vacancy and a further therapist on long leave. Average waiting time had grown steadily, and the duty clinician, Samuel, was spending much of each week telephoning people to explain delays. Among the referrals, 212 carried a risk flag from the intake screen, indicating recent self-harm, significant functional decline, or a request for urgent contact. Those flagged cases were not being seen faster than the rest because the list was ordered by arrival date.
The challenge
Leadership proposed three responses. The first was to keep the arrival order and add a second therapist shift, which the budget could not cover. The second was to triage by risk and move flagged cases to the front, which would leave ordinary referrals waiting even longer. The third was to offer group and brief sessions to people with lower complexity. Staff were divided, and the team had to choose a combination that protected the flagged cases without writing off everyone else.
Analysis
Arrival date is a poor proxy for need, and the intake data made that visible. The 212 flagged referrals were a small share of the list but carried most of the duty clinician's calls. The evidence on the flagged group was specific: several had waited longer than the team's own standard for a first contact after a risk screen, and two had been contacted only after a family member phoned. Patient needs differed within the list. Someone with mild, stable symptoms and a clear preference for a self-directed approach has different needs from someone whose functioning has fallen away over the past few weeks. Triage that treated them identically would protect neither.
The service's stepped care protocol and triage policy both require a risk-flagged referral to be reviewed within two working days by a clinician, and to receive a named therapy contact within a set window. The protocol says nothing about the order of ordinary referrals, which is where the team had room to decide. Coordination mattered too. The weekly allocation meeting, attended by the clinical lead, the duty clinician, and the group therapy coordinator, had been treating capacity as a single pool, and nobody was tracking the flagged cases separately. Protecting quality meant agreeing which options were safe for which people, and writing that down so that a locum therapist could apply it.
- Triage Rank referrals by risk and functional need, not arrival date, using the agreed tool.
- Schedule Offer group and brief sessions where suitable, reserving individual slots for higher need.
- Allocate Shift a fixed share of therapist hours toward the waiting list each week.
- Monitor Track waiting times and deterioration, and reopen triage rules when either drifts.
Actions taken
- Create a separate flagged queue, reviewed by the duty clinician every working day, with each referral assigned a named therapist and a first contact date recorded within two working days.
- Set a weekly capacity split in advance, with a fixed share of therapist hours for flagged cases and a fixed share for group and brief sessions, and review the split at each allocation meeting.
- Write a one-page eligibility guide for group and brief sessions, listing the presentations that are suitable and the risk flags that exclude a referral, agreed with the clinical lead.
- Track median waiting time for ordinary and flagged referrals separately each month, and reopen the triage rules if either group's waiting time rises for two consecutive months.
Results
The flagged queue was running within the two-day window by the end of the second month, and the number of flagged referrals waiting beyond that window fell from 38 to 6. Ordinary referrals waited longer on average, and the median wait for them rose by about three weeks before the group sessions began to absorb some of the load. Samuel's telephone calls to explain delays dropped. Two therapists raised concerns about the group format for people with more complex histories, and the eligibility guide was revised to exclude them. The team reported the split and its outcomes to the governance meeting at the end of the quarter.
Questions for discussion
- What evidence would you need before accepting that arrival order is an unacceptable basis for ordering the list?
- How should a service decide which group formats are safe for people with complex histories, and who should hold that decision?
- When capacity is fixed, what would make you reopen a triage rule rather than add staff or sessions?
Case Study 4
Where follow-up and engagement diverged across teams
Pinecrest Outpatient Recovery Service, a clinic of 11 clinicians offering time-limited care for adults with mood and anxiety disorders, with about 1,860 active clients.
The situation
Pinecrest runs three clinician groups, each with its own caseload and scheduling habits, and reports a monthly set of anonymized indicators to the governance group. Over the last several reporting periods, the indicators had moved in different directions for different groups. Group A had the highest share of clients completing their planned sessions and the most consistent review timing. Group C had the lowest share of reviews completed within thirty days of the risk screen. Group B had strong satisfaction scores but a noticeable number of clients who disengaged after the first few sessions. The quality lead, Mei Ling, had the data but not yet a clear reason for the differences.
The challenge
The obvious explanation, that one group simply had harder cases, could not be tested from the summary indicators alone. The team had no risk-adjusted view, the indicators were calculated from different scheduling systems, and two of the groups had recently absorbed referrals from a primary care link with a different case mix. Monthly figures were small and noisy, so a single bad month could look like a trend. Clinicians were wary of being ranked, and the first conversation about the data had become a discussion about who was to blame.
Analysis
The indicators describe what happened to clients, and the clinical question is why. Early dropout could mean a poor match between need and treatment, or it could mean recovery, and the summary cannot distinguish the two. The risk screen suggested that Group C's lower review rate was concentrated in the first month of treatment, when clients were newly allocated and the review date was not yet in the calendar. Patient needs in each group therefore had to be read alongside engagement and risk status, not in place of them. A client can be satisfied with a contact and still not receive the care the plan intended, so satisfaction was treated as a weak signal.
The service's review protocol requires a risk review within thirty days of the initial screen, and a named clinician to hold the follow-up date. The requirements exist but live in different places. Group C booked follow-up dates in a paper diary that was reconciled weekly, while Group A used the shared electronic scheduler, and the handover from intake to treatment was a verbal exchange at the Monday meeting. When an allocation came through on a Friday, the follow-up date often never reached the clinician's calendar. The process cause is plausible, but it is not yet proven. The monitored cycle should test whether the timing gap accounts for the difference before the team changes anything for everyone.
Actions taken
- Pull the risk screen date and the first follow-up booking date for every new client over the next full reporting period, and record the gap in days in one shared sheet, with group labels only.
- Ask each group to move follow-up bookings into the shared scheduler on the day an allocation is made, and have the duty clinician check the calendar for any new client by Friday afternoon.
- Set a shared target that at least ninety percent of new clients have a risk review date in the calendar within two working days of allocation, and report the figure by group each month.
- Run a staggered test over three reporting cycles, switching the reminder on for one group at a time, and compare review timing and dropout against the previous three months.
Results
The first two reporting cycles showed the timing gap was real. In Group C, the share of reviews completed within thirty days rose from roughly half to just under four in five once the calendar reminder was switched on. Group A's figures barely moved, since its process had already been close to the target. Dropout in Group B fell slightly but not beyond the normal monthly variation, so the team did not yet count the change as an improvement. The reminder was extended to all groups at the end of the third cycle, and the service agreed to keep the indicator set unchanged for two further quarters.
Questions for discussion
- Which indicator in this service would you trust least as a measure of clinical improvement, and what would you measure instead?
- How would you separate a case-mix effect from a process effect when the two groups receive referrals from different sources?
- What should happen to a monitored change if the results are encouraging but the team cannot yet explain why they moved?
Case Study 5
A capacitous refusal and a worried daughter
Brookfield Assertive Community Team, a multidisciplinary team of 17 staff supporting about 240 adults with severe and enduring mental illness.
The situation
Ruth is 46 and has been supported by the team for several years after recurrent depressive and psychotic episodes, and has been stable recently. In recent contacts she told her key worker, Jonah, that she wanted to decline a referral for a step-up assessment and asked that her daughter, Keisha, not receive any information about it. Keisha called the team twice, worried about changes in sleep and a withdrawal from her usual routine. Jonah assessed Ruth and found she understood the options, could weigh them, and expressed a consistent view across three contacts. Keisha had raised safety concerns, and the team had a record of Ruth's previous episode in which she had been unwell for several weeks before anyone noticed.
The challenge
The team has to decide whether Ruth's refusal stands, how much of the conversation with Keisha can be shared, and whether the concerns her daughter raised justify overriding the refusal. Jonah wants to honor the decision while keeping the door open. The team lead, Obi, is uneasy about a three-week gap between the last contact and the next scheduled review. Two colleagues disagree about whether a family call is enough to trigger a wider risk discussion.
Analysis
Ruth's needs and her wishes are not the same thing, and the case turns on keeping both in view. She has been stable, has a history of episodes that were not recognized early, and has told the team what she wants. The clinical evidence is a capacity assessment that found her able to understand, retain, weigh, and communicate a decision, plus collateral from a daughter who is worried but has not seen the most recent contacts. That evidence supports respecting the refusal for now. It does not settle whether the change Keisha describes is an early warning, which would need a clinical review of sleep, routine, and mood at a face-to-face contact.
The team's consent and confidentiality policy separates two questions that are often merged. The first is whether Ruth has capacity for a specific decision, which Jonah answered on the record. The second is whether information can be shared with Keisha, which depends on Ruth's agreement or on a clear and recorded basis such as serious risk. The protocol asks for a written decision about each question, with the reason. Obi was not told of the second family call until two days later, and the two colleagues who disagreed had not met to discuss the risk screen. Escalation to the consultant psychiatrist was appropriate, but only once the team had agreed what it was asking.
Actions taken
- Jonah records Ruth's capacity decision in full, including what she understood, what she declined, and the specific information she agrees can be shared with Keisha, if any.
- Obi convenes a fifteen-minute case discussion within two working days, with Jonah, the colleague who raised the risk concern, and the duty clinician, to agree the risk screen outcome and the next contact date.
- Offer Ruth a planned contact, face to face, within ten days, focused on sleep, routine, and what she would want Keisha to know, so that her view is re-asked rather than assumed.
- Refer the case to the consultant psychiatrist with a written question on the refusal and on whether the family concern meets the threshold for wider sharing.
Results
Ruth attended the planned contact in person and said she still did not want the step-up assessment, but she agreed that Keisha could be told the team was checking in and given a number to call if she was worried. The consultant psychiatrist agreed that the family concern did not yet meet the threshold for wider sharing and asked for the next review to focus on sleep and routine. Keisha was told what could and could not be discussed, and she said the explanation helped although she disagreed with the outcome. The team recorded each decision, its basis, and its review date, and the case was reviewed at the next scheduled meeting.
Questions for discussion
- At what point, if any, would a capacitous refusal become a reason to share information with a family member, and who should make that call?
- How would you record the difference between a person's stated wishes and the team's concern so that a later reviewer can follow the reasoning?
- What would change your advice if Keisha's account had described a specific plan rather than changes in sleep and routine?
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.