Research

2026

Circumstances Preceding Firearm Suicide Deaths in a Large Sample of Hispanic/Latino Adults: Findings from a Content Analysis

In 2026, this study I assisted with was published. It looked at circumstances preceding firearm suicide death among U.S. Hispanic/Latino adults and whether the circumstances varied by sex. The National Violent Death Reporting Systems collects data from law enforcement and medical examiners on all violent deaths, including all suicide deaths, in the U.S. The data include qualitative summaries of narratives by both law enforcement and medical examiners. We performed a content analysis of a random sampling of those narratives.

We found that there were six categories of recurring circumstances preceding death in the study population: behavioral health problems, physical health problems, relationship issues, socioeconomic instability, characteristics of the scene of death, and information about firearm access and storage. For example, known mental health conditions and depressed/melancholy mood were frequently mentioned in narrative summaries. Many of the sampled decedents had relationship and socioeconomic stressors, including romantic relationship challenges, interpersonal arguments, and legal issues. Female decedents more frequently had prior suicide attempts and access of another person’s firearm mentioned in their narrative summaries.

The full study can be read here.

Cross sectional examination of public perceptions around substance misuse using state behavioral risk factor surveillance system data

This paper, published in 2026, was part of a larger project I worked on around the social and economic impact of Huntsman Mental Health Institute at the University of Utah. In this study, we focused on examining public perceptions of individuals who misuse substances, in Utah. The question was what characteristics would be related to negative perceptions, because that information might be useful to groups in Utah who are trying to decrease stigma around substance use disorders. There was a seemingly unique question on the Utah Behavioral Risk Factor Surveillance Survey that asked about this–just in Utah and just for one year–that allowed us to examine this.

We found that older age, Hispanic race/ethnicity, affiliation with several Christian religions, residence in rural counties, and lack of health insurance were negatively related to sympathy toward individuals misusing substances. Those with past-month marijuana use, higher education, and current poor mental health were significantly more likely than their counterparts to express this sympathy.

Based on those findings, we suggest that the former list should be targeted in anti-stigma efforts and also that substance use disorder-related sympathy or stigma should be studied in nationally-representative datasets.

The study can be read here.

Associations among Epilepsy Phenotypes and Pain, Sleep Interference, and Dissociative Symptoms

When I was doing VA work, I worked on a project about epilepsy phenotypes. We looked at useful ways to group veterans that share certain characteristics. The “phenotypes” are the shared characteristics. Once developed, you can see if those groups have different health than each other. A study I wasn’t involved in first developed those groups: (1) relatively healthy, (2) post-traumatic stress disorder, (3) anxiety and depression, (4) chronic disease, (5) bipolar/substance use disorder, and (6) polytrauma.

These groups were formed by looking at veterans health before and after they were diagnosed with epilepsy, then grouping veterans together based on shared patterns.

When I joined the project, our task was to figure out whether these groups, or phenotypes, were related to other health problems or to quality of life issues. The first of those studies, which I helped with, was published in 2026. The question there was whether veterans in different groups had different likelihoods of experiencing pain, sleep problems, or dissociative symptoms.

As you might expect, the answer was yes. Veterans in the PTSD, polytrauma, and chronic disease groups were more likely to have problems with pain than veterans in other groups. Veterans in the polytrauma and bipolar/substance use groups were more likely than others to have problems with sleep.

One takeaway is that these groupings of veterans with epilepsy might one day be useful for health care providers. For example, if a doctor knows that their patient who has epilepsy belongs in the polytrauma phenotype, they might be more aware of the patient’s risk for pain and sleep problems.

The study can be read here.

National Trends in Telehealth Utilization, 2020-2023: Post-Pandemic Trends from the Medical Expenditure Panel Survey

In 2026, a telehealth study that I helped with was published. It was rare to use telehealth before the COVID-19 pandemic, and then its use positively exploded during the pandemic (with an assist from a bunch of legal changes). Somewhere between 15-50% of visits were via telehealth at the peak. A lot of this was in mental health specialties, gestroenterology, and endocrinology. Our question was this: lots more people used telehealth during the COVID-19 pandemic, but did they keep using it afterward?

What we saw was that some people definitely have. The percentage of health care visits that were done using telehealth went from 1.84% in 2020 to 4.53% in 2021, and then remained relatively steady in 2022 and 2023. Most visits were for outpatient care—including mental health services.

It seems like telehealth use belongs on the list of things that the pandemic might have changed forever.

The study can be read here.

2025

The Relationship Between Mental Health and Employment Status Among United States Veterans: A Systematic Review

As another entry in the large, Congressionally-mandated research project on veterans’ health an employment, I led a study on mental health and employment in post-9/11 U.S. veterans. It was published in 2025.

In this case, the study is a review of other people’s science; it can be very important to go back and summarize other peoples’ work in order to spot problems with it, things that are missing, and places where agreement is developing. It’s clear that, in general, unemployment is higher in people with mental illness than in people without it. So we tried to see what the research is saying about this relationship in veterans.

There is likely a relationship between depression and unemployment in veterans. We found that the evidence was much weaker for relationships between other mental health conditions and employment. There were a few potential reasons for this:

• Not very many studies in the first place

• Studies looked at really different groups of veterans

• Studies aren’t using a standard definition of employment — For example, are students employed, unemployed, or not in the job market?

• Studies aren’t clearly referencing a theory that tells them which other factors to account for (like education, combat exposure, etc.), so they’re all accounting for different sets of factors

The takeaway was basically that there needs to be more research (the perennial suggestion in research!) and that researchers need to be more strategic about their methodological choices in this area.

The study can be read here.

2024

Chronic Physical and Mental Health Conditions Associated With Employment Status in Veterans

In 2024, I helped with a study about chronic conditions and employment in post-9/11 U.S. veterans. It was part of a larger, Congressionally-mandated research project on veterans’ health and employment.

Many veterans have health problems, and many veterans have trouble with employment. It’s well known that, for all people, having a chronic disease makes it more likely that they’ll be unemployed. This is somewhat less well-studied in veterans than in the general U.S. population, though—which is the point of having the Congressionally-mandated project. We tried to learn exactly which conditions might be related to unemployment in veterans.

A lot were, even when we took into account things like age, sex, race/ethnicity, and socioeconomic status:

• Diabetes

• High blood pressure

• Stroke

• Emphysema

• Arthritis

• Serious hearing loss

• Poor self-reported mental health

• Poor self-reported physical health

• Depression

• Psychological distress

Hearing loss is an interesting one. Hearing plays an important role in the workplace, like helping with communication and following safety regulations. Hearing problems are common in veterans. Their hearing is often damaged by exposure to blasts.

There are government programs to help veterans with employment, but they are mostly restricted to veterans with mental health problems or disabilities that are a direct result of their military service. These findings suggest that it might be useful to make those programs available to more veterans.

The study can be read here.

2023

Poverty and Suicidal Ideation Among Hispanic Mental Health Care Patients Leading up to the COVID-19 Pandemic

This study, published in 2023, was similar to an earlier study I helped with. Again, suicide rates have increased recently in the Hispanic community. Suicide has complex causes, but poverty might be among them. So we explored the relationship between poverty and suicidal ideation among Hispanic patients who had received mental health care.

We saw that Hispanic mental health patients who had experienced poverty had 1.55 times the odds of also experiencing suicidal ideation, compared to people without poverty experience.

The study can be read here.

2022

Discrimination and Suicidality Among Hispanic Mental Health Patients, 2010–2020: A Natural Language Processing Approach

Another study I worked on was published in 2022. We wanted to understand if there was a link between discrimination and suicidal ideation for Hispanic patients receiving mental health care. Hispanic suicide rates have increased recently, and so have reports of discrimination against Hispanic people.

In our data, Hispanic patients who had experienced discrimination were 1.72 times more likely to also experience suicidal ideation. It was even higher among younger patients.

The study can be read here.

Vaccine Uptake in the US After Full Food and Drug Administration Approval of the BNT162b2 mRNA COVID-19 Vaccine

In 2022, I led a study to try to see if the FDA’s approval of Pfizer’s COVID-19 vaccine changed COVID-19 vaccination patterns in the U.S. If you remember, the vaccine was available before approval. The pandemic was considered such an emergency that the vaccine was allowed to be used in the U.S. before full approval and with only most of the safety and effectiveness trials complete rather than all of them. In that case, the first two phases of study were complete and the third phase was partially done. Even in that third phase, though, information was already available from thousands of study patients.

This was a relatively unique situation–to have a drug available before the government fully approved it for use. I thought I’d see whether approval impacted people’s vaccination behaviors. Did they change at all? If yes, did vaccinations go up or down after approval?

To figure this out, I forecasted out what vaccination rates would have been without approval and then compared that to actual vaccination rates after FDA approval.

The results?

If we didn’t separate out first and second doses, vaccination rates went up 36%. If we did separate doses, there was a 16% decrease in first doses and a 77% increase in second doses.

So, if we imagine things from an individual’s perspective: someone who hadn’t yet gotten the first dose became less likely to get it (at least during the period we looked at), and someone who already had the first dose but not yet the second became more likely to get the second.

The study was just on rates, so it didn’t give us any information about why this might be, but some guesses:

Negative public perceptions of government vaccine information were common at the time. For people with those feelings, government signals of vaccine approval might discourage them from getting a vaccine.

Alternatively, people who had plans to get a COVID-19 vaccine at the time of approval made by a different manufacturer might have paused to consider whether they should get the brand which had been approved.

For people who already had one dose, the approval may have reminded them to schedule the second.

We hoped that this study would inspire more work into strategies for helping move people from no vaccination to one to two (and boosters afterward).

The study can be read here. I was first author, and the study was published in JAMA Network Open.

This research was also covered in the national news by organizations including USA Today, MSN, and Poynter.