I remember one cold winter in Boston when I was helping a patient transition from a medical respite program to a shelter a few streets away. The timing was already tight. Beds were opening and closing on a schedule we had no control over, and transportation had been arranged for a narrow window in the afternoon. By the time we went outside, the weather had already started to change the plan.
It was raining steadily, with strong wind pushing against us as we loaded a few belongings onto a wheelchair. What should have been a straightforward transfer became slow and uneven. We moved block by block, stopping often because the sidewalks were already collecting water and visibility was getting worse.
The patient I was with used a wheelchair. Ramps emptied into puddles, making it hard to pass through. Curb cuts that were technically "accessible" became difficult to use because water had collected in them. Some intersections were only passable if we went out of our way to find a different route. Nothing was officially closed, but parts of the city were no longer realistically usable.
We eventually made it, but what stayed with me was how quickly everything became harder — not because of a breakdown in care, but because the conditions outside changed. The discharge plan was still technically intact. The shelter bed was still available. Transportation still existed. And yet, in real time, the system was barely holding together.
That moment changed how I think about access to care.
As medical students, we usually talk about healthcare access in terms of insurance, transportation, or appointment availability. But for patients experiencing homelessness, access is also dependent on something much less stable: the environment around them. And that environment is changing.
More than 770,000 people experienced homelessness on a single night in the U.S. in 2024. Many of the patients I worked with were already navigating fragmented systems of care. Getting to an appointment might involve multiple bus transfers, long walks, communication with shelters, and transportation of everything they own.
There is very little margin for error, and climate change removes that margin even further.
I started noticing it in smaller ways after that storm. Heavy rain meant bus stops that were technically "operational" but hard to reach. A flood warning meant routes that suddenly took twice as long. A heat advisory meant patients had to choose between traveling to an appointment and finding a safe place to cool down.
On paper, these might look like missed visits. In reality, they are often situations where getting to care becomes impossible or unsafe. And the consequences are not small. Missing a wound care appointment can lead to worsening infection. Missing a psychiatric follow-up could destabilize treatment. Missing a dose or a visit in addiction treatment can interrupt recovery.
But these disruptions rarely show up as system failures. They show up as "no-shows" or "missed appointments," without the context of what it actually took to try to get there. What often gets lost is that getting to care is part of the care itself.
For patients experiencing homelessness, there usually is no backup option. No car waiting outside. No ability to reschedule around work or weather. No safe place to wait out a storm and try again later. If the bus isn't running, or the sidewalks are flooded, or the heat is dangerous, the appointment is gone. Climate change is making these situations more common.
Weather patterns are becoming more volatile: stronger storms, heavier rainfall, and more frequent heat waves are making parts of cities temporarily unreachable. Even when services stay "open," access to them becomes inconsistent.
What stands out is that climate change is not creating these problems from scratch. It is exposing how fragile access already was.
In healthcare settings, we often focus on what happens inside clinics. But for many patients, the most important barriers are outside the clinic doors. As a medical student working in these settings, I've started to think less about whether care is available and more about whether it is actually reachable.
Healthcare systems can respond to this. Flexible scheduling during extreme weather. Transportation support that accounts for actual disruptions, not just ordinary delays. Mobile outreach when travel becomes unsafe. Stronger coordination between shelters and clinics so patients are not navigating everything alone. These are not radical ideas; they are extensions of what good care already asks us to do.
I think back to that rainy day often. Not because it was extreme, but because it wasn't. It was a fairly normal day in Boston that became much harder than it should have been. Nothing "failed" in an obvious way. And yet, for the patient I was with, the system only worked because we were able to push through conditions that were actively working against us.
For patients experiencing homelessness, climate change is not a future threat to prepare for. It is already testing whether the system holds up at all.



