Rethinking POLST: How Early ACP Bridges the Gap and Elevates Patient Voice
May 8, 2026
Most POLST conversations start in crisis. The median interval between POLST completion and death is 6.4 weeks. The average patient completing a POLST form in a hospital setting is 73 years old and the clock is already running.
That timing isn’t a design choice. It’s a default. And it points to a structural problem that the field has been slow to name: the traditional ACP and POLST process is entirely provider-dependent. A patient can only get that conversation if a clinician decides it’s time to have it, which means it happens late, inconsistently, and often in a moment of crisis.
That model encodes a quiet power imbalance: the provider decides when you’re “sick enough” to talk about your own wishes. Most patients are ready for that conversation long before their doctor initiates it. And when it only happens inside a clinical encounter, it gets crowded out as there’s always something more urgent on the agenda.
It also creates enormous equity gaps. Patients with more frequent provider contact, more health literacy, and more assertive self-advocacy are more likely to get the conversation. The patients who most need it are often the least likely to be offered it.
At Koda, we’ve been asking a different question: what if patients could do meaningful groundwork before the clinical trigger? What do they actually want, and what does the data tell us when we give them the structure to answer that themselves, on their own terms, years before a crisis?
We presented some of those findings at the National POLST Collaborative Conference in New Orleans this April. Here’s what the data shows.
Patients Are Getting to Conversations Sooner and on Their Own Terms
The average age of a Koda member completing their advance care plan guide is 69, a full decade earlier than the published mean age at POLST completion. That’s not a marginal difference. It means patients are articulating their preferences, naming a decision-maker, and communicating their values to their families years before a crisis forces the issue.
Just as striking: 68% of Koda members complete their guide without a physician referral. They initiate on their own, in their own time on evenings, weekends, outside of clinical hours enitrely. That tells you something important: they’re not engaging because a provider put it on the schedule. They’re engaging because they wanted to.The completion rate is 76%, more than double the 30-40% industry benchmark for digital health tools.
When patients initiate, they also come in with more clarity. They’ve had time to think, to talk to family, to process. The conversation is different in quality, not just in timing.
Patients aren’t afraid of this conversation. They’re ready for it. They just need a process that meets them where they are.
What Patient Selections Reveal
When patients are given a structured, accessible process as opposed to a rushed hospital intake), their preferences are nuanced and specific. Here’s what Koda’s population data shows:
On CPR: The largest single group (46%) wants CPR only if their treatment team believes it’s appropriate and not if quality of life would be unacceptable. Only 10% want CPR no matter what. Most patients are expressing something more sophisticated than a binary yes or no. They want shared decision-making. And notably, younger patients (under 65) are nearly twice as likely to want aggressive intervention as older patients, which has real implications for how we design age-aware ACP conversations.
This is also where ACP starts doing real work toward clinical documentation. When a Koda member selects no CPR today and no CPR if quality of life becomes unacceptable, the platform automatically flags them as potentially eligible for an Out-of-Hospital DNR (OOH-DNR). An OOH-DNR is a medical order, distinct from but related to POLST, that instructs emergency responders not to attempt resuscitation outside a hospital setting.
That flag triggers a follow-up step: Koda’s patient advocates have a dedicated field to document that the conversation was had and that facilitation steps were completed. 14% of Koda members meet that eligibility threshold. That’s not a documentation exercise, it’s a direct pipeline from expressed preference to actionable medical order, initiated years before a crisis would otherwise force it. And for many of those patients, it’s also the natural opening for a POLST conversation: the values are already on the table, the hard question has already been answered, and the clinician isn’t starting from scratch.
On life-sustaining interventions: Across long-term feeding, ventilation, and dialysis, the pattern holds: older patients consistently prefer less aggressive treatment, while younger patients skew higher. The age gaps, which are 4 to 5 percentage points across each intervention, reinforce why one-size-fits-all ACP tools fall short.
On hospice and palliative care: 85% of Koda members express interest in hospice, and 82% in palliative care, if quality of life becomes unacceptable. That’s not a small signal. For context, POLST completion is associated with a 49.9% hospice admission rate (vs. 27% with advance directives alone) and an 85.7% out-of-hospital death rate (vs. 72%). Early, structured engagement moves those numbers.
On where they want to spend their final days: 73% of Koda members want to die at home. 15% prefer a hospital. 12% prefer a facility. Another 16% aren’t sure yet, but they’re engaged in the process, which means there’s still time to have the conversation.
On what matters most to them: 95% of members cite family. 67% cite independence. 55% cite religion. 38% cite pets. These aren’t clinical data points. Rather, they’re the context that makes goals-of-care conversations meaningful rather than transactional.
Equity in ACP Engagement
The equity picture in ACP (and by extension POLST) is well-documented and still urgent. Published research shows DNR completion rates of 45% among non-Latino white patients vs. 25% among Black patients and 20% among Latino patients. Younger adults, people with lower educational attainment, and rural residents are consistently the least likely to complete any form of advance care planning. That means they’re also the least likely to ever have a POLST.
These are infrastructure and access problems, not patient motivation problems. The traditional model amplifies them: patients with less frequent provider contact, lower health literacy, and less assertive self-advocacy are exactly the ones least likely to be offered the conversation in a clinical setting. If POLST completion depends on a provider initiating the ACP conversation first, the same populations who are underrepresented in ACP will be underrepresented in POLST, and underserved at the moment it matters most.
Koda’s own engagement data reflects a different pattern. The data from our work at Houston Methodist Coordinated Care ACO suggests what’s possible when you change the model. In a published study of 454 patients (JABFM, Vol. 36, No. 6, 2023), Koda demonstrated consistent ACP completion rates across every demographic group measured — race, gender, and socioeconomic status:
- Black patients: 50.0% / White patients: 47.6%
- Lower SES: 42.1% / Upper SES: 40.8%
- Female patients: 40.2% / Male patients: 45.1%
Across all six groups, completion rates ranged just 9.8 percentage points, from 40.2% to 50.0%. In a field where equity gaps routinely exceed 20 points, that’s a meaningful result. It suggests the gaps we’ve accepted as inevitable aren’t inherent to the conversation. Build a process that patients can access on their own terms, and the disparities start to close. More importantly for POLST: it means a far broader and more representative population arriving at the clinical encounter already prepared with values documented, decision-maker named, preferences considered. The POLST conversation becomes the next step in a process that’s already underway, rather than the first and only chance to have it.
| Group | Completion Rate |
|---|---|
| Black patients | 50.0% |
| White patients | 47.6% |
| Lower SES | 42.1% |
| Upper SES | 40.8% |
| Female patients | 40.2% |
| Male patients | 45.1% |
JABFM Vol. 36, No. 6, 2023
The Bridge to POLST: A 10-Year Head Start
ACP and POLST have traditionally been treated as separate conversations. I see them as part of the same continuum.
ACP sets the stage. It asks people what matters, who speaks for them, what care they want when they’re seriously ill. POLST translates those values into medical orders when the moment comes. One without the other leaves gaps.
POLST was designed for a specific moment: a patient who is already seriously ill or facing functional decline, sitting across from a clinician, making decisions under pressure. It does that job well. But it shouldn’t be the first moment a patient has thought about what they want. By the time POLST is appropriate, the window for a thoughtful, unhurried conversation has often already closed. Families are stressed. Patients may be sedated or cognitively impaired. The decisions that should have been made together, over time, are now being made in a hallway.
There’s also a harder question worth asking: if the form is only completed under clinical pressure, in a hospital room, how authentic are those preferences really? Removing provider dependency from the front end doesn’t remove clinical oversight. It removes the bottleneck that was preventing the conversation from happening at all. A patient who has reflected on their values at home, talked it through with their family, and arrived at a considered position is expressing something more meaningful than one who’s handed a form during a 15-minute appointment.
That’s the gap ACP is built to fill, and where Koda lives.
When a patient has already named a medical decision-maker, articulated their values, and shared their preferences with their family years before a hospitalization, the POLST conversation shifts from orientation to confirmation. From crisis to care. 94% of Koda members name a medical decision-maker as part of their guide. Identifying a named, aligned surrogate is one of the primary functions POLST serves. We’re making sure it happens at 69, not 79.
Early ACP integration is also linked to measurable downstream outcomes: favorable POLST completion rates, reduced 60-day hospital readmissions, lower terminal hospitalizations, and higher rates of death in the patient’s preferred setting.
69
73
wks
What This Means for the Field
The data points to a few things we think deserve more attention:
Completion timing matters as much as completion itself. A POLST filled out six weeks before death is better than nothing. A care plan completed a decade earlier, shared with family, updated over time, and integrated into clinical workflows, is fundamentally different.
Age-stratified preferences require age-aware tools. The consistent gaps between patients under and over 65 across CPR, feeding, ventilation, and dialysis preferences suggest that population-level ACP programs need to account for this, instead of applying the same conversation to everyone.
The MDM relationship is the variable the field is underweighting. Goal-concordant care doesn’t happen in a document. It happens when the right person is in the room, knows what the patient wanted, and feels confident acting on it. Koda tracks MDM alignment rates because we believe that’s one of the most meaningful (and least measured) indicators of whether ACP actually worked.
Scalable ACP (and scalable POLST) cannot be gated behind a provider visit. There aren’t enough clinicians, there isn’t enough appointment time, and the incentives aren’t there to prioritize it. If we want ACP completion to move from single digits to meaningful population penetration, the patient has to be able to be the initiating party with structured support, clinical guardrails, and a clear pathway back to their care team. That’s not replacing the physician. It’s building the infrastructure that lets the physician’s time and expertise be used where it actually matters.
We talk about patient-centered care constantly. But the ACP process has never actually been patient-centered. Instead, it’s been provider-initiated, provider-timed, and provider-documented. Giving patients the infrastructure to drive this themselves isn’t a workaround. It’s what patient agency actually looks like in practice.
The work ahead isn’t just getting more patients to complete more documents. It’s building infrastructure that reaches patients earlier, captures their preferences with enough depth to be actionable, and ensures those preferences are connected to the people and systems that will need them.
That’s the bridge we’re building.
Koda Health presented this data at the National POLST Collaborative Conference, April 16, 2026, New Orleans. For questions or to learn more, reach out at jen@kodahealthcare.com.
