Why the next generation of public-health infrastructure must connect healthcare, social care and coverage and follow people all the way to resolution.

by Michael Hancock, Former Mayor of Denver and Nima Roohi, CEO of Blooming Health

There is a moment that plays out every day in emergency departments across America. The patient is stabilized. The discharge instructions are printed. And then the harder part begins.

What if the patient has no stable place to recover? Not enough food? No transportation to a follow-up appointment? What if Medicaid coverage has lapsed - or a notice requiring action is sitting unopened at home?

The clinical emergency may be over. The social emergency is not.

Denver understands this problem better than most cities. We have each seen, from different vantage points, what happens when healthcare, government and community organizations work across traditional boundaries. Denver has spent years building a more integrated safety net across housing, behavioral health, crisis response, community care and benefits navigation, anchored by Denver Health.

The next challenge is to make those pieces behave more like one system.

Denver Has Already Changed What a Health Intervention Can Be

Denver Health's responsibilities extend far beyond those of a traditional hospital. Its system includes a Level I trauma center, neighborhood and school-based clinics, behavioral-health and addiction services, correctional healthcare, paramedic and public-health functions, and a health plan. It provided roughly 1.4 million patient visits in 2025.

Nearly one in five adults hospitalized at Denver Health is unhoused. Through its HOPE program and partnerships with the Colorado Coalition for the Homeless and Denver Housing Authority, Denver Health reports reducing the average length of stay for unhoused patients from 11 days in 2022 to eight days in 2025, while reducing uncompensated-care costs by more than $6 million in the past year.

This is good healthcare - and good economics.

The evidence extends beyond the hospital. A randomized Denver evaluation of supportive housing found participants averaged six fewer emergency-department visits two years after entering the program. A Colorado initiative providing coordinated community support to frequent emergency-care users produced 27.9% fewer subsequent ED visits and 114% more primary-care visits than among comparison patients.

Denver has learned an important lesson: the goal is not to keep people away from emergency departments when they need them. It is to prevent the ED from becoming the only part of the system they can reliably reach.

From Resources to Resolution

Denver has built an impressive network: community resources through Denver Health, a "no wrong door" approach to addiction treatment, STAR's alternative response to appropriate 911 calls, and major investments in addressing homelessness. The city's 2026 Point-in-Time count identified 518 people living unsheltered - 64% fewer than in 2023.

But having resources and successfully connecting people to them are different things. A screening is not an outcome. A referral is not an outcome. A phone number is not an outcome. The outcome is that the need is resolved.

Denver Health's 2026 community-health assessment reflects this distinction. Residents called for easier scheduling, help maintaining Medicaid enrollment, transportation assistance and better navigation between services. Denver Health wants to know whether patients receive timely follow-up after discharge and actually receive the community services to which they were referred.

These are not fundamentally referral problems. They are execution problems.

The Medicaid Test

Beginning in 2027, federal changes will require many working-age adults in the Medicaid expansion population to demonstrate qualifying employment, education, work-program or community-service activity equivalent to 80 hours per month unless exempt. Eligibility reviews will also become more frequent for affected populations.

There is a crucial difference between being eligible and successfully proving eligibility. Someone may be working but struggle to document it. Someone may not realize an exemption applies. A notice may go to an old address. A resident may complete four steps but miss the fifth.

Colorado Health Institute, citing Urban Institute modeling, estimates that paperwork barriers associated with the new requirements could cause approximately 95,000 to 108,000 Coloradans to lose Medicaid coverage.

For Denver, this is a public-health and fiscal question. Nearly half of Denver Health's payer mix is Medicaid, and the system already absorbs more than $140 million annually in uncompensated care. Denver residents recently approved a dedicated sales tax to help sustain the institution.

Denver stepped up when its safety net needed help. Now we should prevent eligible residents from unnecessarily falling out of coverage.

Make Follow-Through Infrastructure

For most of modern healthcare, information systems have been designed primarily to record what happened. The next generation should help make things happen.

Imagine a virtual care manager working alongside the human care team after a resident leaves the emergency department. It reaches them through the channel and language that work for them. If today's outreach goes unanswered, it tries again tomorrow. It schedules the primary-care appointment rather than providing a number. It identifies transportation barriers, guides someone through Medicaid renewal, coordinates with community organizations, confirms whether services were received and escalates complex situations to a human.

AI is not the point. Follow-through is.

Automation should not remove people from care. It should remove repetitive administrative work that prevents extraordinary people from providing it.

Denver already has much of what this model requires: Denver Human Services, public health, Denver Health, community health centers, housing and behavioral-health providers, and extraordinary nonprofit and neighborhood organizations.

What Denver needs next is not necessarily another organization. It needs a way for these organizations to execute together.

That is the difference between a safety net and a safety network. A net catches someone after they fall. A network recognizes risk earlier, activates the right resources, knows whether the connection worked and continues until the problem is resolved.

Denver Can Lead Again

Anyone who has led a city knows that the hardest public challenges rarely fit neatly inside one agency, budget or institution. Progress comes when we organize around the outcome rather than the bureaucracy. The answer today cannot simply be asking already overstretched clinicians, social workers, community health workers and public servants to do more. We have to give them leverage.

Denver has done hard things before. It has demonstrated that housing can be a healthcare intervention, that emergency response does not always have to begin with law enforcement, and that persistent challenges like unsheltered homelessness can move when a city organizes around an outcome.

The next frontier is less visible.

It is the patient who leaves the ED with five things to do and completes none. The resident who qualifies for coverage but cannot navigate the process required to keep it. The community organization that receives a referral but cannot reach the person. The care manager with 300 names and enough time to truly help 30.

Those cracks are where tomorrow's emergencies begin.

Denver can build something different: a model in which healthcare, government and community organizations share responsibility not simply for identifying needs, but for getting people all the way to resolution.

A citywide care-execution infrastructure. A safety network that follows through.

Ultimately, the measure of a healthcare system is not how many messages it sends, referrals it makes or programs it creates. It is whether people get the care they need.

Denver can make that the standard.