Endocrinology and Diabetes Management Case Studies
Healthcare5 Case StudiesUpdated Sep 2026
Endocrinology and diabetes management relies heavily on administrative work: referral queues, enrollment registries, education scheduling, records that arrive in the wrong format, and handoffs between primary care, specialty clinics, and nurse educators. The five cases come from outpatient and regional settings where the plan itself is rarely the problem. Each one begins with a team that is busy rather than failing, and each shows how unclear roles, loosely defined service levels, unexamined workflow data, and unspoken stakeholder expectations wear down delivery targets. The analysis keeps the same frame throughout so the cases can be compared.
Case Study 1
Missed referral targets in a shared endocrine scheduling team
Larkspur Valley Diabetes and Endocrine Center, a hospital-affiliated outpatient group with nine clinicians and a central intake office
The situation
Larkspur Valley's intake office handles every new referral to the endocrine clinics, roughly 2,400 requests in a typical quarter. Three coordinators sort them, verify insurance details, and book first visits, while two nurse navigators handle the patients who need glucose monitoring enrollment and diabetes education. For most of the year the office met its target of booking new patients within ten business days. Over the last two quarters that share dropped steadily, and the medical director noticed that the backlog was not spread evenly. Some referrals sat in an unassigned queue for weeks while coordinators worked on others they considered urgent. Nobody could say who owned a referral once the navigator team took it on.
The challenge
Ruth, the operations manager, had two problems to untangle at once. Targets were slipping with no change in headcount, and nobody could name who carried a stalled referral. Coordinators blamed navigators for slow callbacks. Navigators blamed incomplete files that took hours to correct. Hassan, the medical director, wanted waits reduced without new hires, and the referring primary care groups wanted one contact they could call about a stuck referral.
Analysis
Teams grounded in Endocrinology and Diabetes Management start by separating symptoms from causes, and this office had two symptoms that looked like one problem. The long unassigned queue was concentrated in referrals with missing insurance fields. Slow callbacks were a handoff problem: coordinators marked files complete, and navigators then found verification gaps. Roles were written before the navigator program existed, so each group read the same task differently. Service levels covered booking but not the stages before and after it, so the handoff itself was never measured. Once the team traced each referral, the accountability gap showed up as a set of unowned moments rather than a personality clash.
The accountability plan had to fit the people involved rather than an ideal org chart. Ruth assigned a single owner to each referral from the moment intake accepted it, and that owner stayed the same through booking. Navigators kept clinical education responsibilities but took on the verification step, which removed the rework loop. Service levels were written for every stage with a named owner, so a stalled referral could be traced. Stakeholder expectations changed: referring primary care groups received a status number they could quote, not a promise of speed the office could not keep. Delivery targets were reported weekly by stage, which showed where time was lost.
Actions taken
- Assign one named owner to every referral at intake, recorded in the scheduling system, and make that owner visible to the referring office on each status call.
- Move insurance and completeness checks to the front of intake so navigators receive verified files, and track verification rework as its own measure for two months.
- Write service levels for each stage, with a time limit and a named owner, and publish them in the team charter so every role can see where it begins and ends.
- Hold a fifteen-minute weekly review of stalled referrals by stage, with Ruth chairing and each owner reporting a single next step rather than an explanation.
Results
Within the first full quarter after the changes, the share of new patients booked within ten business days rose back to its earlier level, and the unassigned queue fell to a handful of referrals. Verification rework dropped enough that navigators reported spending their mornings on education rather than repairing files. Primary care groups began calling the status line instead of escalating by email, which the office counted as a sign that the single contact was working. The weekly review was shortened after the second month because most stalled referrals now had a clear owner before the meeting started.
Questions for discussion
- Which of the five roles in this office had the clearest written expectations before the changes, and how did that gap shape the early conflict?
- How would you decide whether a single named owner per referral creates a bottleneck for senior staff who are pulled into exceptions?
- What measure would you add to the weekly review to catch a stage that is quietly absorbing time from the others?
Case Study 2
Rebuilding a partner hospital's trust after repeated complaints
Tallow Creek Endocrine Services, a contracted outpatient program that runs diabetes education for three partner hospitals
The situation
Tallow Creek runs a contracted diabetes education program for three partner hospitals, and about 650 patients a quarter enter it after discharge or clinic referral. The program has a documented service level: an initial education appointment within fifteen business days of referral, and a reply to partner inquiries within two working days. For most of the year it met both. Then Priya, chief of medicine at the largest partner hospital, began sending written complaints that her discharge planners were learning about the program's status only after patients had called her unit. She also said appointments were moved without notice, and that her team could not tell whether a referral had been received.
The challenge
Amara, the program director, had to decide quickly whether the complaints reflected a fixable process gap or a relationship that was already damaged. The contract allowed a formal review in the next quarter, and Priya had copied the network's executive team on her last two messages. Amara needed to respond within the week without promising changes her coordinators could not sustain.
Analysis
The first task was separating what Priya expected from what the contract actually required. The written service level measured appointment timing, but her complaints were about notification, which was not in the agreement at all. Tallow Creek's data showed that appointments were usually rescheduled in the same week, yet the partner hospital received no message about it. Referral receipt was confirmed in the scheduling system but never sent back to the referring unit. Once Amara laid the complaints beside the contract terms, most of them traced to two gaps: undefined communication duties and a status feed that existed only inside Tallow Creek's own system.
Trust came back through specificity, not reassurance. Amara drafted a one-page service agreement addendum that defined who told the partner what, and when. Notification would go to the discharge planner's shared inbox within one working day of any scheduling change, and every referral would receive a receipt confirmation within two working days. A named coordinator became the single contact for the partner, with a backup listed for absences. Targets were reported monthly to Priya's team, including the share of referrals acknowledged on time, so she could see performance rather than rely on a summary.
Actions taken
- Call Priya within two days to acknowledge the complaints in writing, separate the contract questions from the communication failures, and name a point of contact for the coming week.
- Start sending a receipt confirmation for every referral within two working days, using the existing scheduling data, so the partner sees acknowledgment without waiting for a status call.
- Agree a written addendum with the partner that sets one-working-day notice for any schedule change and names a single coordinator with a listed backup.
- Send a monthly scorecard covering acknowledgment timing, notice timing, and appointment dates, and review it with Priya's team in person at each quarter end.
Results
Priya accepted the addendum at the next steering meeting, though she asked for a second review. Acknowledgment timing stayed inside the two-day target for every month of the first quarter, and the partner's discharge planners stopped asking for status updates by phone. Notice of schedule changes reached the shared inbox within one working day in nearly all cases, with the exceptions documented on the scorecard. The formal review was cancelled. The partner's confidence rating on notification rose from three to seven across the next two surveys, and the program kept the scorecard as a standing item.
Questions for discussion
- Where in your own organization would a missing notification duty be most likely to hide inside a contract that looks complete on paper?
- How much weight should a single senior complaint carry when deciding which service gaps to fix first?
- What would you need to see in the first scorecard to conclude that trust had actually been rebuilt rather than merely reported?
Case Study 3
Pre-visit packet rework in a multisite endocrine practice
Rowan Hill Endocrinology Group, a nine-site outpatient practice with twenty-two clinicians and a shared records team
The situation
Rowan Hill's new-patient pre-visit packet is assembled from four sources: the referring office's fax or portal upload, the patient's insurance card, prior laboratory summaries from outside systems, and an intake questionnaire completed on a tablet in the waiting area. The packet is supposed to be ready five business days before the first visit. In the last stretch, about one visit in four started with a missing record, and the clinician spent the first part of the appointment chasing documents instead of reviewing the plan. Rework was constant: medical assistants re-requested records that had arrived but been filed under the wrong patient, and front desk staff rescheduled visits that had been confirmed too early.
The challenge
Ibrahim, the practice administrator, had been asked by the physician partners to fix the workflow without moving the visit schedule. The team had tried a checklist, which people stopped using within a month. Nobody could say where the delay started, since each step looked finished from its own side. Medical assistants believed the referral office was slow. The referral office believed packets were rejected for reasons it could not see.
Analysis
Mapping the workflow turned up seven handoffs where work changed hands, and four of them had no shared status field. The biggest loss sat between document receipt and indexing, where faxed pages waited in a general inbox for an average of three business days. Some were filed under the wrong patient because the referral form and the registration record used different spellings of the same name. Packets were marked complete when the questionnaire was signed, not when the outside records arrived, so the front desk confirmed visits too early. Root causes were structural: ownership was unclear at two handoffs, and the completion rule measured activity rather than readiness.
Ibrahim compared three options against the same criteria: effect on waiting time, staff effort, and risk to the visit schedule. A stricter checklist was cheapest but had already failed, since it relied on memory at the end of a busy day. Moving intake to a dedicated records coordinator would fix ownership but added a salaried role and a new handoff. A rule change, where a packet counts as ready only when every outside record is indexed and matched, cost little and attacked the completion gap directly, though it would delay some confirmations. The team chose the rule change first, with a records coordinator considered again if the backlog returned.
- 1Referral receivedFax or portal upload lands in a general inbox with no owner.
- 2Records indexedPages wait an average of three business days before being matched.
- 3Questionnaire signedPacket marked complete here, before outside records are confirmed.
- 4Front desk confirmationVisit confirmed early, then rescheduled when records are still missing.
- 5Clinician reviewVisit starts with document chasing, cutting into time for planning.
Actions taken
- Redefine packet readiness so a packet counts as complete only when every outside record is indexed to the correct patient, not when the questionnaire is signed.
- Set a same-day indexing standard for faxed and uploaded pages, with an end-of-day queue check owned by a named medical assistant on each site.
- Match referral spelling against registration at the point of receipt, using date of birth as a second identifier, so misfiled pages are caught before the packet is built.
- Delay front desk confirmation until the packet passes readiness, and report the share of visits held back so the schedule impact stays visible to physicians.
Results
After the rule change had run for two full cycles, the average time from referral receipt to a ready packet fell from about eight business days to under five. Misfiled pages dropped to a few per month, mostly from referrals that arrived with unusual name formats. Early confirmations fell, and same-week reschedules dropped. The cost was real: about one visit in ten was now held for records, which the physicians accepted but asked to watch. The team kept the records coordinator idea on a shelf, to be reopened only if the same-day indexing standard slipped for two weeks running.
Questions for discussion
- Which handoff in your own intake process has the weakest owner, and what evidence would you collect to confirm it?
- When a physician partner accepts a schedule cost in exchange for fewer reschedules, how should that trade be reviewed later?
- Why might a checklist fail in a busy practice even when the people using it agree with its purpose?
Case Study 4
Rolling out a mandatory education enrollment field
Cedar Mile Diabetes Partners, a regional group of six clinics with a telehealth team and a quality committee
The situation
Cedar Mile introduced a required structured field for diabetes education enrollment, along with a new referral portal that replaced email and shared mailboxes across six clinics. The policy was set by the group's quality committee after a review found that enrollment status was recorded inconsistently, with some patients listed as enrolled in notes but absent from the education registry. Clinicians were told the change would take about two minutes per visit. Early on, nurse educators reported entering the same information twice, several physicians stopped completing the new field, and coordinators raised concerns that the portal would make them responsible for work they did not control. Adoption among clinicians stood near half of visits in the second week.
The challenge
Kenya, the clinical operations lead, needed the policy adopted without reopening the decision. Leadership wanted the field used in every visit by the end of the quarter. Clinicians wanted proof that the time cost was real and that the data would not be used against them. Nurse educators wanted the duplicate entry removed before they would sign off. Kenya needed a communication sequence that addressed each group's concern in the right order, without treating resistance as a communication failure alone.
Analysis
Each group's resistance pointed to a different gap in the original rollout. Clinicians objected to the time cost, but the time study had been run only in the portal's test environment, which ran faster than live use. Nurse educators were not consulted during design, so the duplicate entry was an oversight. Coordinators feared accountability for referral completeness they could not fully control, and the policy had assigned that accountability without changing their queues. Leadership expected uptake to follow from a memo, which assumed the people affected understood why the data mattered. Stakeholder expectations had never been written down, so each group filled the gap with its own reading.
The adoption sequence started with the parts that cost the least trust. Nurse educators were brought into a two-week working group to remove the duplicate entry, and the registry was set to pull the new field automatically. Physicians then received a short briefing on what the registry would and would not be used for, with that commitment written into the policy text. Coordinators were given a named escalation contact and a service level for resolving incomplete referrals in the portal, so responsibility matched authority. Uptake was measured weekly as the share of eligible visits with the field completed, broken out by clinic, so the clinics needing help stood out.
- Week oneWorking group removes duplicate entry and maps registry pull.
- Week twoPhysician briefing on registry use, with terms written into policy.
- Second month startCoordinators receive escalation contact and portal resolution service level.
- Mid second monthWeekly uptake report by clinic shows where support is needed.
- End of quarterField completion target reviewed against the original memo expectation.
Actions taken
- Form a two-week working group with nurse educators to remove the duplicate entry and confirm that the registry pulls the new field without manual transfer.
- Publish a written statement of what the registry will and will not be used for, and attach it to the policy so clinicians can read the commitment before they comply.
- Give coordinators a named escalation contact and a four-day service level for incomplete referrals in the portal, so accountability matches the authority they actually hold.
- Report weekly uptake by clinic, not as a group average, and hold a short call with any clinic below the agreed threshold before the next report.
Results
Uptake reached the agreed threshold in five of six clinics by the end of the quarter. The sixth clinic still ran a local referral inbox outside the portal, and moving it took a separate month. Nurse educators reported that the duplicate entry was gone and that registry gaps had fallen. Physicians still raised the time cost, so the group measured it directly and found an average of a little under a minute, far below the original estimate. The memo's uptake target was revised to reflect the sixth clinic's transition, not to lower the bar.
Questions for discussion
- Which concern in your own rollout would you have treated as resistance, and which would you have recognized as a design gap?
- How should a leader respond when a physician group wants a target lowered after the data shows the task takes less time than they claimed?
- What would make a weekly uptake report by clinic misleading, and how would you guard against that?
Case Study 5
Choosing among three initiatives with thin local evidence
Northfield Metabolic Care Network, a regional group of six clinics and one telehealth team, with a discretionary budget of 480,000 dollars
The situation
Northfield's operations council had 480,000 dollars in discretionary funds for the coming year and three initiatives competing for it. The first was an expansion of telehealth follow-up visits, which the network had piloted in one clinic. The second was an outreach program for adults recently diagnosed, modeled on a program a neighboring group had described at a regional meeting. The third was a referral triage redesign that would route new referrals by urgency and insurance status, with the aim of shortening waits. Each initiative had a sponsor, and each sponsor had a set of numbers that made the case. None of those numbers were measured the same way.
The challenge
Simone, the chief operating officer, was expected to bring a recommendation to the council within four weeks. Each sponsor argued from the evidence closest to hand, and no two initiatives could be compared on the same measure. The council wanted a choice that would not lock up the whole budget before anyone knew what worked, and it wanted the reasoning written down so the next cycle could check it.
Analysis
Simone began by agreeing on criteria before scoring anything, since the sponsors had been ranking initiatives by what they cared about most. The council settled on four: expected reach across the network, confidence in the evidence available locally, cost of being wrong, and time to a first readout. Confidence was the hard one. Telehealth had six months of pilot data from one clinic, enough to see that scheduling improved but not enough to estimate effects across sites. Outreach had only a neighboring group's report, which described a different population. Triage had clean internal workflow data but no evidence yet on whether shorter waits would reduce missed visits.
The staged decision followed from the risk profile rather than from anyone's preference. Triage was the cheapest to test and the most reversible, since it changed internal routing, so it received the first tranche with a defined readout in the second quarter. Telehealth expansion had the most local evidence and the widest reach, but its staffing commitments were hard to unwind, so only one additional clinic was funded at first. Outreach was deferred, not rejected, with a small planning grant to collect local baseline data. The council agreed on a rule: if the triage readout showed no effect on missed visits, the next tranche would move to telehealth.
Actions taken
- Agree the four criteria and their weights in writing before any initiative is re-scored, and record which sponsor proposed each change to the weights.
- Fund triage routing first with a fixed readout date and a missed-visit threshold set in advance, so the council judges the result against a number it agreed to.
- Limit telehealth expansion to one additional clinic and attach staffing conditions, so the costliest commitment stays reversible until local effects are measured.
- Give the outreach proposal a small planning grant to collect baseline data from the local population, and bring it back with a comparison rather than a neighboring report.
Results
The council approved the staged plan with the triage readout set as the gate for the second tranche. Triage routing went live in the second quarter, and the readout later showed a reduction in missed visits that fell just short of the agreed threshold, so the rule moved the next tranche toward telehealth with staffing conditions attached. The telehealth clinic reported that its added capacity was used but that scheduling conflicts at the sites took longer to fix than expected. Outreach received its planning grant and returned with local baseline figures that changed the target population. The reasoning document became the template for the next cycle's proposals.
Questions for discussion
- What criteria would you add to the four used here, and what would each one cost in time to measure?
- How should a council respond when a readout lands just short of its threshold, and who should argue for a second look?
- Which of the three initiatives would you have been most tempted to fund first, and what evidence would have changed your mind?
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.