The 2022 Oak Fire spread quickly, burning more than 19,000 California acres in a few weeks amid severe drought. But for some medically vulnerable residents in Mariposa County, the harm was much longer lasting.

More than one in five medically vulnerable people who responded to survey questions reported missing missed medical appointments – going without care for weeks or months in 90% of those cases. In addition, more than one in five respondents said delayed medical care after the wildfire had harmed their health.

Those are just a few of the findings reported in a study newly published in the journal Disaster Medicine and Public Health Awareness, part of Cambridge University Press. Andrew Schroeder, Direct Relief’s vice president of Research and Analysis, is one of the co-authors, along with a team of emergency medical researchers at Harvard University, and Mariposa County Health and Human Services officials. The study is one of a series of four on health and wildfires recently published or under development.

After a disaster, Schroeder explained, “one of the things people want to know is, ‘How were the most vulnerable people affected?’”

The Support and Aid for Everyone, or SAFE, program of Mariposa County’s Health and Human Services agency provided an opportunity. Participation in the program is voluntary for residents who qualify as medically vulnerable and at higher risk during an emergency. Mariposa County officials worked with CrisisReady, a disaster response initiative at Harvard University and Direct Relief, to learn more from SAFE program participants about their experiences during the Oak Fire.

“Rather than randomly sampling households, we were directed toward this group of people who had already been identified as being medically vulnerable,” Schroeder said.

Many of the findings are startling: Fifty-three percent of respondents reliant on powered medical devices weren’t enrolled in PG&E’s Medical Baseline Program, an assistance program for customers whose medical care depends on electricity.

While county officials worked quickly to issue emergency alerts, offer guidance to people in affected communities, and communicate evacuation orders and warnings, 59% of people first learned about the fire when they saw flames or smelled smoke.

Schroeder sat down with Direct Relief to talk about what the research team learned, what it means, and how new knowledge can help California navigate future wildfires.

---

Direct Relief: From your perspective: A) What are the most significant findings of this study and B) What surprised you as a researcher?

Andrew Schroeder: We commonsensically know that the most medically vulnerable people on a normal day are also the most medically vulnerable in disasters. But we often don’t quantify that in terms of, “How they are affected in particular ways? What proportion of people are likely to be medically vulnerable? Is it uniformly distributed, or clustered in some way? Does it relate to certain kinds of behaviors that people exhibit during disaster, or their [specific] medical conditions?”

The Mariposa County population’s exposure to disaster is kind of a top-heavy distribution overall: You have the population that’s affected by disaster. Then you have more medically vulnerable people who are somewhat more affected as a subset of that total.

But even 20% of that [medically vulnerable] group was disproportionately likely to have the most severe problems. And that population was also more likely to have more medical conditions [per person]. They were also more likely to have power-dependent medical devices that were subject to outages.

In a nutshell: Yes, the medically vulnerable, as a group, are more likely to have more problems than the general population. But even within that, there’s actually an 80-20 problem for exposure to risk that pops out of the data.

Direct Relief: Looking at the two earlier studies in this series, both of them have a broader lens, right? The first one is calculating the increasing proximity of wildfires to inpatient facilities, even as those facilities take on greater numbers of patients. And the second looks at exposure to power outages across California counties in 2019.

But this new study takes a narrower focus: It’s a specific fire, a specific fire-affected community. I’d love to hear a little bit about how this narrower, deeper lens fits into the larger research project.

Andrew Schroeder: Well, there are multiple levels of this problem. We want to know whether acute care facilities are affected by wildfires in a way that is significant or has spatial variance in the first article. We weren’t looking in time. There’s a fourth article forthcoming in this kind of group that will look at this problem over time, in terms of changing proximity to fire perimeters.

But basically, from the standpoint of public health, or medical resource planning, what can we say about the likelihood that an acute care facility is going to be close to the perimeters of wildfires over a particular area. Does that vary across the state? Does it vary based upon the type of facility?

The power outage study takes the same planning problem but looks at it from the standpoint of the functionality of the facility and the exposure of people with power dependent devices – not just in acute care facilities. We’re also looking at long-term care, and we’re also looking at community health centers.

So not only are you seeing fire perimeters encroaching more closely on areas where there is medical infrastructure. You are also seeing the operational status of those facilities placed at risk to some degree by the coincidence of power outages and wildfires.

Then [in this third study], you drill down into, “What does that literally mean for people who are affected by wildfires?” That’s where the Oak Fire study fits in. And it’s not uniform. It has this particular communication dimension to it. There’s a significant aspect of, “How do you promote health-seeking behavior? How do you ensure continuity of power for medical devices at home?”

There is a set of things that are being done to mitigate the problems, such as the PG&E Medical Baseline Program. But even there, the uptake is surprisingly low, at roughly 50%.

I wouldn’t necessarily say that you could straight-line generalize from Mariposa County to everywhere in California – everyone will have some variation in these problems – but Mariposa is not a bad place to treat as a kind of representative study of rural counties in California. It’s a piece that does give you, I think, a broad view of some of the problems that California faces around this particular type of hazard.

Direct Relief: So much of the research that has been done on disaster impacts does something similar to what you’re describing here: It has a focused lens looking at a particular disaster, or maybe a specific community that’s experienced multiple disasters. We’re slowly building a body of those studies, but a lot of them are specific to place and time.

My question is, how much of a disaster’s impact and aftermath are unique to these places and to these individual events, and how much can we extrapolate to prepare for a future event in a different place?

Andrew Schroeder: Ideally, you would do this all the time. You would regularly study the impact on the general population, and clearly on the high-vulnerability population, so there’s an effort to learn from the past. Often that is not done. So, there’s a need for more research.

But you can compare the characteristics of medical vulnerability to a lot of different places. There’s similarity at the population level. There’s similarity at the geographic level. There’s similarity at the policy level.

The exact numbers would be different, but at the agenda-setting level I think [the study] does point to a lot of similarities. I would be surprised if you were to go up to, say, Siskiyou County and get really drastically different results for a similar population.

Direct Relief: Direct Relief often hears from partners that, “We don’t have the kind of data we need to create further preparedness and response plans, and we can’t wait for it. We need to know what to do now.” And often there’s this informal network where a particular community or a particular response organization will have wisdom to share, and they’ll just call up another county or organization and say, “Hey, we think you should be ready to do X.”

In this case, let’s say a rural county or community planning for disaster came to you and said, “OK, you did this study. What should we do know?” What would you say to them?

Andrew Schroeder: One [recommendation] is to just steer people towards programs like the Medical Baseline Program. There’s a need to create connectivity between people and available programs. There’s a need to build preparedness for interruptions to care, which means having a specific mitigation plan for your medicines, your care provider.

Obviously, there’s a need to improve outreach and risk communications. There needs to be more creative thinking about how to get reliable information into the hands of people who have demonstrated issues receiving that kind of information.

Direct Relief: It is so striking that the majority of respondents learned about the fire when they saw plumes or smelled smoke. That’s just not what we hope to hear.

And many respondents said they found the evacuation process uncertain. They didn’t know if they could bring belongings, they didn’t know when they could come back, and it was stressful.

Andrew Schroeder: They also said there’s a gap between creating a preparedness plan and actually evacuating. And the experience is different in that you’re told what you need to take with you, lists of medication and key contacts. And these are not necessarily super tech savvy people, so they’re writing a lot of this down on paper.

Then when you have to put that into practice, reality is always different from your plan. One of the things that was consistently reflected back is that [study respondents] were surprised by how different the experience of evacuation was from anything that they had predicted or put into their plan.

Something is needed to close that gap.

Direct Relief: So this is a serious issue for people who did make those plans as instructed but felt that it didn’t serve them.

Andrew Schroeder: Yeah. There’s no single answer to that, but I do think you could try to solve it through some kind of exercise, or some kind of discussion group or simulation. I would imagine that’s pretty hard to do, actually, and I would imagine there aren’t great [existing] resources for it, but just raising the question would be an important start.

Direct Relief: It seems like sharing lived experience could be a good thing: people who have made plans and lived through disasters coming in to talk about what they did that helped, and what no one could have prepared them for.

Andrew Schroeder: I think that’s a great idea, and I think you should quote yourself on that one.

Direct Relief: I’m not doing that.

You know, I just said that the study was hyper-local and hyper-focused. But it also covers a breadth of information: You talk about missed medical appointments, and the health impacts those caused. You talk about preparedness, you talk about information needs, evacuation needs, how and if people evacuate at all.

What is the benefit of looking across the board at these different kinds of data together as opposed to more siloed information?

Andrew Schroeder: Well, it’s a complex experience. The causes and consequences of medical vulnerability and disasters are not one thing. One of the sources of medical harm is that, if you’re on the [Mariposa County] SAFE list, you’re more likely to need regular contact with your physician. If that is interrupted, then your probability of medical harm, by definition, goes up.

That’s why I think one of the important findings is that, if you missed one appointment, you were likely to miss contact [with a medical provider] for months. That’s a long duration of interruption to care. And interruption is only one dimension. At the household level, the loss of power that changes your ability to utilize home medical devices is a different dimension of your medical vulnerability.

You have to be able to look at this as a problem that has more than one facet to it, or you’re not going to actually capture the lived experience of people who are subject to these kinds of conditions.

And if you’re going to try to formulate policy or improved public health practice, you won’t be accounting for the factors you need to if you don’t see this as an integrated problem.