Research

Research Objective:

An estimated 1 in 5 Americans live in rural areas and face limited access to healthcare. These challenges are often exacerbated for members of historically underserved groups, though telemedicine may facilitate access to care. The study objective was to describe telemedicine utilization by race/ethnicity group in rural areas and evaluate how its use relates to other healthcare utilization.

Study Design:

In this cross-sectional study, we used administrative claims data to compare healthcare utilization between rural telemedicine users and rural non-telemedicine users. Propensity score matching was used to match rural telemedicine and non-telemedicine users on the following characteristics: age, sex, Elixhauser Comorbidity Index, chronic conditions, race/ethnicity, and state of residence. Multivariate logistic regressions, adjusting for area-level socioeconomic status, chronic conditions, age, and sex, were used to assess the relationship between telemedicine use and the following outcomes: emergency department (ED) visits, inpatient admissions, and evaluation & management (E&M) visits.

Population Studied:

Commercial and Medicare Advantage members (age 18+) residing in a zip code or census tract classified as “rural” by either Rural-Urban Commuting Area or Frontier and Remote codes between 2019-2023.

Principal Findings:

Telemedicine use was disproportionately higher in the White population (82% of rural telemedicine users vs. 75% non-telemedicine users) and relatively lower in the Hispanic population (4.9% of rural telemedicine users vs. 5.4% of non-telemedicine users) (p <0.05). The proportion of rural telemedicine users and non-telemedicine users was similar in the Asian (0.6%) and Black (3.1 vs. 3.0%) populations.

The relationship between telemedicine use and other healthcare utilization varied by race/ethnicity. In the Hispanic population, telemedicine use was associated with 31.5% more ED visits and 39.9% more inpatient admissions, relative to Hispanic non-telemedicine users (p <0.05). Telemedicine use in the Black population was associated with 20.2% more ED visits and 72.0% more E&M visits (but was not significantly associated with inpatient admissions), relative to Black non-telemedicine users (p <0.05). Telemedicine use was associated with 23.9% more ED visits, 34.3% more inpatient admissions, and 63.3% more E&M visits among White telemedicine users compared to White non-telemedicine users (p<0.05). Telemedicine was not significantly associated with other healthcare utilization in the Asian population.

Conclusions:

Results show that telemedicine is not equally utilized across race/ethnicity groups. The Hispanic population is underrepresented among rural telemedicine users and show relatively higher ED and inpatient utilization. This population may be in relatively poorer health or may have easier access to healthcare services generally.

Implications for Policy or Practice:

Understanding the causes of race/ethnicity disparities in healthcare utilization in rural areas is important for addressing individuals’ needs. Potential underutilization of telemedicine by the Hispanic population could possibly be addressed by increasing awareness of telemedicine services available, improving language accessibility of telemedicine services, and/or increasing digital literacy. Policies that include investing in the expansion of broadband to areas without service may also help ensure equitable access to care across race/ethnicity groups in rural areas.

Research Objective:

Following the COVID-19 public health emergency, telemedicine utilization has declined but remains significantly higher than pre-COVID levels. We determined whether telemedicine use among rural residents is associated with a higher likelihood of preventive screenings.

Study Design:

Using administrative claims data from a large health insurer, we constructed a sample of adults with commercial or Medicare Advantage health insurance residing in rural areas defined by either RUCA (rural-urban commuting area) or FAR (frontier and remote area). We propensity score matched telemedicine users to non-users based on age, sex, Elixhauser comorbidity index, race/ethnicity, state of residence and several indicators of pre-existing conditions. The outcome of interest is whether an individual had a preventive screening (CPT code range 99381-99429) in either 2021 or 2023.

Population Studied:

Our sample consisted of adults (age 18 and older) with commercial or Medicare Advantage health insurance who resided in rural (defined by RUCA or FAR) areas between January 2019 and December 2023.

Principal Findings:

Among those residing in rural areas as defined by RUCA, we found that any telemedicine use in 2020 was associated with an 80.1% (p<0.01) higher likelihood of completing a preventive screening in 2021. Additionally, any telemedicine use in 2021 to 2023 was associated with an 83.9% (p<0.01) higher likelihood of completing a preventive screening in 2023. Heterogeneity of the relationship was found based on underlying conditions, such as heart failure, obesity, hypertension, and by race and ethnicity. Similar results were found for those residing in rural areas as defined by FAR.

Conclusions:

Telemedicine utilization in rural areas is associated with a higher likelihood of having preventive screening. The magnitude of this relationship varies depending on underlying health conditions.

Implications for Policy or Practice:

Our results suggest that the use of telemedicine may be one way to increase preventive screenings among rural populations. Telemedicine can facilitate access to care, which may have particular importance for rural residents who often have limited or reduced access to health services. Given prior research showing that regular visits with healthcare providers are associated with receiving recommended preventive screening, telemedicine may be a way to encourage or strengthen the relationship between providers and patients. This information can help to inform policies and future research that aim to promote telemedicine to improve access to preventive health care in rural areas.

Abstract: Many hourly paid and low wage earners choose to waive enrollment in commercial health insurance coverage. This can be problematic because low wage earners tend to work in conditions that are more prone to injury and illness. Nevertheless, potential motivations not to enroll may be the high cost of premiums, excellent health, and health literacy. However, little evidence documents employee enrollment dynamics focused specifically on those who waived and then enroll in commercial coverage, especially among low wage earners. We estimate the change in health care utilization among those who enroll in health insurance following a year of waiving employer health insurance. Our samples consist of 10,292 employees at a large nationwide restaurant chain employed between 2019 and 2020. In 2020, 17% (1,753 employees) of those enrolled waived coverage in 2019. We study the impact of the enrollment variation across time combined with employee job-categories tied to wage – salaried employees who are generally higher income, hourly employees making below $15 an hour, and hourly employees making above $15 an hour. Among those employees enrolled in both years, hourly employees (both below and above $15 an hour) utilize less outpatient care than salaried employees. Salaried employees enrolled in 2020 were about 7 percentage points less likely to utilize outpatient care and utilized about 2 fewer outpatient visits in 2020 compared to salaried employees enrolled in both 2019 and 2020. Compared to salaried employees who enrolled in coverage in 2020, employees making less than $15 an hour who enrolled in 2020 were about 5 percentage points more likely to utilize outpatient care and utilized 2 more outpatient visits in 2020. Hourly employees making more than $15 an hour who enrolled in 2020 did not utilize outpatient care significantly different than salaried employees who enrolled in 2020. For all employees who enrolled in 2020, we did not observe increased emergency department use or hospitalizations. However, both low and high hourly paid employees had higher primary care physician visits than salaried employees who enrolled in 2020, while salaried employees who waived coverage in 2019 had fewer primary care physician visits than salaried employees enrolled in both years. To explore whether the identified differences in health care utilization could be explained by a lag in utilization once enrolled due to learning, we also look at the difference in utilization between the newly hired and the newly insured. Both groups are insured through this employer in 2020 and not insured through this employer in 2019. When compared to newly hired employees, hourly employees (below and above $15 an hour) are more likely to utilize outpatient care. Those newly insured making below $15 an hour are about 12.5 percentage points more likely to utilize outpatient care and those above $15 an hour are about 9 percentage points more likely to utilize outpatient care. There is no statistically significant difference between newly hired and newly insured salaried employees, suggesting that waiving health insurance coverage may lead to pent-up demand among hourly employees but not among salaried employees.

Abstract: I estimate the impact of air pollution events caused by wildfire smoke on respiratory and circulatory health outcomes. Utilizing a combination of California health data and NOAA wildfire smoke data I can estimate the impact of exposure to wildfire smoke on health outcomes for all individuals in California. Using inpatient data I am able to construct a measure of exposure to wildfire smoke prior to the hospital visit, this allows for the identification of the impact of wildfire smoke exposure on different health outcomes. I find that an additional day of smoke exposure in a month leads to on average 11.38 additional hospital admissions for respiratory diagnoses and an additional 3 hospital admissions for circulatory diagnoses. This translates to an annual cost of wildfire smoke exposure in California due to respiratory and circulatory hospital admissions of $192,316,498.

Abstract: How does the sharing economy affect traditional lodging markets? The advent of platforms such as Airbnb in 2008 has introduced a new channel of market interaction between those with space and those who seek it. This allows for transactions of lodging services that might otherwise be underutilized. This paper develops a framework to help think about how peer-to-peer transactions interact with traditional rental mar- kets, and what this means for property managers and tenants. Specifically, we examine how the introduction of sharing platforms (e.g. Airbnb) affect the listing decisions of managers of vacant properties and the lodging choices of dwelling seekers. The model features landlords who choose where to list vacant properties and renters who search for lodging. Renters can be either short or long-term, referencing how long they wish to occupy the property. Sharing platforms give landlords the option of accessing these short-term renters who would otherwise occupy hotels, affecting traditional, long-term renters. We find that Airbnbs decrease hotel prices by about $24 per night (about a 10% reduction) while they increase average rents by $39 per room, per month (about a 2.5% increase).

Abstract: We study the behavior of cluster-robust test statistics in models with instrumental variables when cluster heterogeneity is present. Inference in a large number of papers using two-stage least squares regressions published in American Economics Association journals are driven by the presence of one or two influential clusters. We link a measure of cluster heterogeneity, the feasible effective number of clusters, to measures of influence. Using simulations, we demonstrate that high levels of cluster heterogeneity lead to coverage of less than 95% for 95% confidence intervals when using instrumental variables with panel data or with data that can be grouped into clusters. Using data from papers with two-stage least squares regressions published in American Economic Association journals, we show that the feasible effective number of clusters can be used as a pre-test to the sensitivity of two-stage least squares inference to influential clusters.

Abstract: I study the impact of changing weather patterns on the incidence of malaria and the effectiveness of malaria initiatives in Sub-Saharan Africa between the years 2000 and 2015. Combining malaria incidence with climate reanalysis I estimate the impact of malarious weather on malaria prevalence and analyze the changing efficacy of malaria prevention and treatment.