Skip to main content

DIME PROJECT

AI-Enabled Care Navigation Framework
for Health Systems

AI Care Nav icon

Where navigation breaks down & where AI can help, responsibly

AI Care Navigation
UNIFIED PATIENT JOURNEY

AI strategy begins with the patient

The Unified Patient Journey below shows where care navigation breaks down. Map your patients to the moments in the Unified Patient Journey. Pick a population or service line, and identify which of the 14 moments cause the most delay, harm or cost.

Within each care moment, open the matching AI Opportunity Cards to view:

  • An AI-enabled response to a specific patient barrier
  • How to get started
  • What to measure
  • What risks to manage
CROSS-CUTTING FORCES

Five forces that shape all phases in the patient journey

The journey moments show where patients get stuck. These five forces help explain why. Each one cuts across all four phases and shows up again and again in the moments above. Any navigation solution, AI-enabled or not, needs to account for them to work in the real world.

Patient trust & equity

Every moment depends on whether patients can recognize a need, reach and afford care, and trust the system enough to engage. Those abilities aren’t evenly shared. 65.8% of US adults have low trust that their health system will use AI responsibly, lower still after discrimination in care.

WHERE THIS SHOWS UP
  • Literacy: over 75 million US adults have limited health literacy.
  • Language: adverse events more often caused harm for patients with limited English proficiency (49.1% vs. 29.5%)

“The static fields in our EHR’s patient portal today offer some translation that’s decent, but the free text is still really only English. And so what we hear from families, and what our clinical teams see, is basically families take what’s in the patient portal, run it through Google Translate, copy and paste it back into the patient portal. That’s unfortunately the current workflow today, probably for many families navigating our overly complex systems of care.”
– Eric Maurer, Chief Innovation & Strategy Officer, Community University Health Care Center (CUHCC)

Payment & cost burden

Cost shapes nearly every care decision a patient makes: whether to seek care, where to go, and whether to fill a prescription or stretch it. Patients delay visits, skip refills, or ration medication because they can’t afford them, and these choices put their health at risk. They usually happen out of the care team’s view. From the outside, a patient who can’t afford treatment can look like one who isn’t following instructions. The response, whether more counseling, a dose change, or a “noncompliant” label, then misses the real barrier.

WHERE THIS SHOWS UP
  • Delayed care: in 2024, 38.6% of uninsured and 17.0% of privately insured adults delayed or skipped care or medication because of cost.
  • Hidden non-adherence: 8.2% of adults on prescriptions didn’t take them as prescribed to save money, rising to 22.9% among the uninsured.
  • Authorization delays: 94% of physicians say prior authorization delays necessary care.
The care administrator challenge

Patients rarely navigate alone. 63 million Americans, nearly 1 in 4 adults, are family caregivers who book appointments and coordinate care. Yet systems are built as if the patient were the only person in the room.

WHERE THIS SHOWS UP
  • Unrecognized roles: caregivers often use the patient’s login; 32%–50% of health systems offer no caregiver accounts. 
  • Consent and confidentiality: over half of adolescent portal accounts studied were accessed by guardians, failing to account for HIPAA and confidentiality laws.

“One of the families we’ve worked with, she’s taking care of three individuals [with] complex conditions across five different pharmacies, 10 different specialists, and she gets text messages from the pharmacy and the doctor: ‘Your appointment [is] scheduled,’ ‘Your prescription is ready to pick up.’ But because of PHI, she has no idea who it’s about, so her inbox is flooded with all these automations saying, ‘What do I do?'”
– Shantenu Agarwal, CEO & Co-Founder, Carey

  • Unprepared caregivers: over half of caregivers perform complex medical tasks, yet just over 20% have formal training.

The settings and people involved in healthcare aren’t always particularly clean:

“Particularly when you’re working with complex patients, you’re often trying to coordinate not only within your institution but across communities. So when we actually do care mapping on our kids, parents identify 33 on average distinct entities that are involved in their child’s well-being, of which only 25% are under medicine.” 
– David Bergman, Pediatrician and Professor Emeritus, Stanford University

Read two vignettes from Stanford Medicine Pediatrics for examples of how these challenges uniquely show up in child and adolescent scenarios:

VIGNETTES | STANFORD MEDICINE PEDIATRICS

Jones Family

Jared and Sarah work full time and have 2 kids – Jacob (4) and Itzel (6). Jacob is in pre-school and Itzel is in 1st grade. Jacob has asthma. Itzel is struggling with learning to read. They have scheduled back-to-back well checks for the kids and are in the room together.

CHALLENGE

AI CARE NAVIGATION RESPONSE

Both parents need access to both kids
Relationship-aware permissions across the family
Histories for the siblings overlap but aren’t identical
AI can propose what is shared and flag what is child-specific
Developmental screening varies based on age/location
Risk assessment and screening between visits and available to multiple users
Care spans clinic, home, school, daycare, etc.
Agentic coordination across multiple entities (school, speech, other clinicians)
Adhering to treatment plans
Gamifying care for kids to help with treatment plans and encourage adherence
KEY THEMES

01 The family is the unit of care

02 Care happens outside the clinic

Jessica

Jessica is 16 with type 1 diabetes and wears a continuous glucose monitor. Her mom works extra hours, but the two have struggled with food insecurity. Jessica has been resistant to monitoring her glucose and taking her insulin because she just wants to be normal and play volleyball.

CHALLENGE

AI CARE NAVIGATION RESPONSE

Providers across multiple systems caring for her health

One navigator that allows for cross-system coordination: healthcare systems, schools, durable medical equipment (e.g. CGM, insulin pump)

Asserting independence at 16

Agentic AI can provide shared plan of care and tools to keep Jessica on target

Mom still makes medical decisions; Jessica may need confidential care

Permissions that shift with age and confidentiality laws

Social needs may impact glucose control

AI-assisted resource coordination; CGM data to help with real-time assessment

College and adult care ahead

Guided transition to adult specialists

AI-assisted CBT programs to help with behavioral needs through young adult phase

KEY THEMES

01 Autonomy is a dial, not a switch

02 Care must travel with the patient

Fragmentation of care settings

Complex patients often see several providers at once, none with the full picture. A typical Medicare beneficiary sees two primary care physicians and five specialists across four practices yearly.

WHERE THIS SHOWS UP
  • No shared plan: 69.3% of primary care physicians say they usually send referral information, but only 34.8% of specialists say they usually receive it.
  • No single owner: only 35% of a typical beneficiary’s visits are with their assigned physician.
  • Real consequences: fragmented chronic care is linked to more ED visits, testing and costs.
The interoperability gap

Data moves more freely than a decade ago, but workflows often don’t act on it. In 2023, 71% of US hospitals routinely had outside patient information available at the point of care, yet only 42% said clinicians often used it.

WHERE THIS SHOWS UP
  • Alerts without action: many hospitals must send admission, discharge and transfer notifications, but notification doesn’t guarantee follow-up.
  • Untracked referrals and results: 38.9% of referral attempts had no documented status and up to 62% of outpatient lab results lacked follow-up.
  • Uneven capability: Smaller, rural, and independent hospitals have fewer resources to invest in technology and the staff to run it. As a result, they’re less likely to reliably share and receive patient records with other providers. That gap compounds existing inequities for the communities that depend on them.
PATIENT PERSONAS

How the unified patient journey maps to real life

Each persona below is a fictional composite that shows how ordinary moments, like a missed follow-up, a hard handoff, or a coverage change, can shape someone’s health and where better navigation could change the outcome.

Danielle
Maternal & child

Minh
Older adult oncology

Marco
Young adult cardiovascular

Kayla
Pediatric & adolescent

MATERNAL & CHILD HEALTH

Danielle’s care journey

Danielle is 32, Black, pregnant for the first time, and works hourly shifts with no paid leave. Medicaid covers her through pregnancy and for 12 months after delivery. High blood pressure readings at urgent care get no follow-up before she’s pregnant, she waits five weeks for her first prenatal appointment, and she’s induced at 37 weeks for preeclampsia. She goes home with two discharge plans, hers and her newborn Jordan’s, and reads only his. Her home blood pressure climbs, as postpartum pressure tends to peak on days 3 to 7, until on day six she arrives at an emergency room that has no record of her delivery, and her depression is caught only later at Jordan’s pediatric visit. When her postpartum Medicaid ends at 12 months, she is still on blood pressure medication and carries a lifelong higher risk of heart disease.

Age 32, first pregnancy
Coverage Medicaid, through pregnancy and 12 months postpartum
Who helps Her mother, 20 minutes away. Her partner works nights.
What the chart tracks Blood pressure, mood, and baby Jordan

Hover over or tap a moment number to see Danielle’s experience:

Not yet recognizedBeing managedMissed moment
Danielle’s care journey Chart of Danielle’s blood pressure, mood, and her newborn Jordan’s health over time, from stable health to acute crisis. Colored bands show the four phases of the Unified Patient Journey. Numbered points match the journey’s 14 moments; each moment is described in the buttons that follow. PHASE 1Stability & drift PHASE 2Acute event & system entry PHASE 3Transition & handoff PHASE 4Active management Acute crisis Rising acuity Stable health Time (not to scale) Blood pressureMoodJordan (newborn)
  • WHY THIS JOURNEY MATTERS
  • In 2024, the maternal mortality rate for Black women was 44.8 per 100,000 live births, about three times the rate for White women (14.2) (NCHS, 2024 maternal mortality data).
  • Maternal mortality review committees judged 84% of pregnancy-related deaths preventable, and 53% occurred 7 to 365 days after delivery (CDC MMRC data, 2017–2019). That is the Transition & Handoff and Active Management window, where navigation is thinnest.
  • Medicaid finances about 4 in 10 U.S. births (KFF).

OLDER ADULT ONCOLOGY

Minh’s care journey

Minh is a 74-year-old Vietnamese-speaking retired machinist with type 2 diabetes and a long smoking history, and he has Medicare Advantage. His daughter Lan manages his care from two states away. He has qualified for lung cancer screening for years but was never offered it, and his 2.4 cm lung nodule turns up only on a CT scan after a fall. The referral stalls until Lan calls again and again, and his stage II diagnosis brings five specialists, prior authorizations, and discharge instructions Lan has to translate over the phone. During immunotherapy, a local emergency room doesn’t know his treatment history, and a steroid taper destabilizes his diabetes and overwhelms Lan, showing how language, distance, and scattered specialists turn a curable cancer into a coordination problem.

Age 74, retired machinist, former smoker
Coverage Medicare Advantage with Part D
Who helps His daughter Lan, two states away. His wife Hoa has limited mobility.
What the chart tracks Lung cancer, type 2 diabetes, and Lan’s capacity as caregiver

Hover over or tap a moment number to see Minh’s experience:

Not yet recognizedBeing managedMissed moment
Minh’s care journey Chart of Minh’s lung cancer, his type 2 diabetes, and his daughter Lan’s caregiver capacity over time, from stable health to acute crisis. Colored bands show the four phases of the Unified Patient Journey. Numbered points match the journey’s 14 moments; each moment is described in the buttons that follow. PHASE 1Stability & drift PHASE 2Acute event & system entry PHASE 3Transition & handoff PHASE 4Active management Acute crisis Rising acuity Stable health Time (not to scale) Lung cancerCaregiver capacity (Lan)Type 2 diabetes
  • WHY THIS JOURNEY MATTERS
  • Minh has been eligible for annual low-dose CT lung screening since he turned 50. USPSTF recommends it for adults aged 50 to 80 with a 20 pack-year history who currently smoke or quit within the past 15 years (USPSTF).
  • Incidental lung nodules are often missed. In one study, 52% of patients did not get follow-up (PMC). In a multisystem VA study, only 46% got guideline-concordant follow-up and 17% got none (Annals of the ATS).

YOUNG ADULT CARDIOVASCULAR

Marco’s care journey

Marco is 24, Latino, bilingual, and drives for rideshare and delivery companies with no paid time off. He has no primary care clinician and is covered as a dependent on his mother’s employer plan. A smartwatch alert and a chatbot’s advice don’t get him to a doctor, and he ends up in the emergency room with a hypertensive emergency and kidney injury. There, staff call his mother, still listed as his emergency contact from his teens, rather than his partner Ana, whom they can involve if he agrees, and his test results go to a patient portal account his mother set up years ago. Insurance statements later tell his mother about his care, which pushes Marco to decide who gets to know what. At 26 he loses her coverage and his pharmacist-led blood pressure program with it, which shows how a young adult can hold the legal rights to his care while the systems around him still treat his parent as in charge.

Age 24 at the start; the journey runs to about 27
Coverage A dependent on his mother’s employer plan until 26, then his own Marketplace bronze plan
Who helps His mother Lucía, the policyholder. His partner Ana, day to day.
What the chart tracks Hypertension, kidney health, and sleep apnea

Hover over or tap a moment number to see Marco’s experience:

Not yet recognizedBeing managedMissed moment
Marco’s care journey Chart of Marco’s hypertension, sleep apnea, and kidney health over time, from stable health to acute crisis. Colored bands show the four phases of the Unified Patient Journey. Numbered points match the journey’s 14 moments; each moment is described in the buttons that follow. PHASE 1Stability & drift PHASE 2Acute event & system entry PHASE 3Transition & handoff PHASE 4Active management Acute crisis Rising acuity Stable health Time (not to scale) HypertensionSleep apneaKidney health Age 24, on Mom’s plan Age 26, ages off Mom’s plan
  • WHY THIS JOURNEY MATTERS
  • Among adults aged 18 to 39 (2021–2023), 21.3% had hypertension, only 28.3% of them were aware of it, and only 5.6% had BP controlled below 130/80. The gap was linked in part to not having a usual place of care (American Journal of Hypertension). CDC reports control at 4.5% in this age group (NCHS data brief).
  • Secondary causes of hypertension are more common when onset is before age 30. Obstructive sleep apnea is present in 25% to 50% of people with hypertension, and the 2025 AHA/ACC guideline recommends screening for it (2025 AHA/ACC guideline; Current Cardiology Reports summary).
  • Coverage has a built-in cliff at 26. Young adults can stay on a parent’s plan until 26 even if they are married, live elsewhere, or are financially independent (HHS). In 2014 Census data, the uninsured rate rose almost 4 percentage points at age 26 (CMS).

PEDIATRIC & ADOLESCENT

Kayla’s care journey

At 13, Kayla’s type 1 diabetes is first mistaken for a growth spurt, and she is diagnosed only once she’s in diabetic ketoacidosis (DKA). She is transferred to a children’s hospital two and a half hours from her rural home, and her mother Rachel, a night-shift nursing assistant, has three days to learn everything. Back home, Rachel copes with 2 a.m. low blood sugars, a school nurse shared across three schools, and CHIP refusing to cover glucose sensors at first, until remote monitoring by her care team steadies things. As Kayla takes over her own care in high school, she starts skipping insulin to lose weight, and a second DKA lands her in a local emergency room that can’t see her history. At 18 she never calls the adult clinic she’s referred to, and at 19 CHIP ends and her sensors run out, which shows how care breaks down when responsibility, privacy, and coverage all shift in the teen years.

Age Diagnosed at 13; the journey runs to 19
Coverage CHIP (children’s health insurance) until her 19th birthday
Who helps Her mother Rachel, a night-shift nursing assistant. Her grandmother drives her to appointments.
What the chart tracks Type 1 diabetes, Rachel’s caregiving load, and Kayla’s mental health

Hover over or tap a moment number to see Kayla’s experience:

Not yet recognizedBeing managedMissed moment
Kayla’s care journey Chart of Kayla’s type 1 diabetes, her mental health, and her mother Rachel’s caregiving load over time, from stable health to acute crisis. Colored bands show the four phases of the Unified Patient Journey. Numbered points match the journey’s 14 moments; each moment is described in the buttons that follow. PHASE 1Stability & drift PHASE 2Acute event & system entry PHASE 3Transition & handoff PHASE 4Active management Acute crisis Rising acuity Stable health Time (not to scale) Type 1 diabetesMental healthCaregiving load (Rachel) 13 · new portal rules 14 · high school 16 · driving 18 · legal adult 19 · CHIP ends
  • WHY THIS JOURNEY MATTERS
  • Diabetic ketoacidosis (DKA) is common when type 1 diabetes is diagnosed in children. In the national SEARCH registry (2010–2016), 39.5% of youth diagnosed at ages 10 to 14 were in DKA at or near diagnosis (SEARCH, Diabetes Care). When the diagnosis was missed at first, 68% of children had DKA, compared with 42.8% when it wasn’t (Clinical Diabetes).
  • Blood sugar control is worst in the teen years. In the T1D Exchange registry, mean HbA1c rose from 8.1% at age 5 to 9.3% at ages 15 to 18 (T1D Exchange, PMC).
  • The move to adult care is a known drop-off point. In one U.S. survey, 34% of young adults with type 1 diabetes reported a gap of more than 6 months between pediatric and adult care (PMC review). Children’s coverage also ends on schedule: CHIP eligibility ends at age 19 (Medicaid.gov), and the uninsured rate jumped almost 7 percentage points at age 19 in 2014 Census data (CMS).

AI OPPORTUNITY LIBRARY

AI opportunities to close patient care gaps

The Unified Patient Journey shows where patients struggle to get from one step of care to the next. These AI Opportunity Cards show where AI can help close those gaps today, and where the evidence is still emerging.

CLOSING NOTE

Care is a continuous journey

Active management doesn’t necessarily have a finish line. For some patients, it’s a durable return to stability. For others, what looks like stability is fragile: the conditions behind the last acute event are still there, better managed for now, waiting for the next lapse in attention. For patients living with chronic conditions, stability and active management blur into the same ongoing state. As one participant in our clinical decision support roundtable put it, “healthcare doesn’t have to end.” The journey is a loop, not a line. Every return to stability is also the start of the next cycle, and the next chance to get navigation right.

Next steps

Now that you’ve identified relevant opportunities, test them with the Impact-Utility-Trust Checklist. Before you pilot, confirm the opportunity is worth doing, will work in your workflows, and can be done responsibly.