For Atiye Cansu Erol, the numbers are both stark and alarming: Mental health issues are responsible for 23% of all pregnancy-related deaths in the United States, according to the U.S. Centers for Disease Control and Prevention, with postpartum depression a leading contributor.
“But nearly half of new mothers with depression go untreated,” she says.
With the massive growth of telehealth since the COVID-19 pandemic, Erol wondered whether virtual mental health treatment options could increase access to care for new mothers and whether there could be barriers to such care. In a resulting working paper, the assistant professor of Information Systems at Johns Hopkins Carey Business School identifies two competing privacy concerns at play. And she offers insights for how telehealth services for maternal mental health could most effectively be implemented.
By analyzing detailed insurance claims data, Erol found that the availability of mental health services via telehealth increased utilization among new mothers, notably among those living in communities where societal stigma related to the use of mental health services exists. But new mothers living in crowded households were less likely to pursue the telehealth option.
“When people think about privacy issues in telehealth, they think about data privacy, and most research until now has focused on health data security,” says Erol. “However, the use of telehealth completely changes your physical presence during the time of care, raising different concerns about privacy that need to be explored.”
Measuring stigma through underreporting
For her study, conducted with Lynn Wu of the University of Pennsylvania’s Wharton School, Erol analyzed administrative claims from a database that included detailed data covering more than 280,000 women who gave birth between 2017 and 2021. Of that group, 35,058 new mothers have a mental health condition, and are also insured — and were thus observable — in the six months after birth.
“This data set is uniquely granular since it includes detailed sociodemographic variables,” says Erol. Among these variables: education, household income, household size, race, and age. “Such demographics are usually not provided in combination with health utilization data,” she notes.
The researchers defined a “crowded” home as having three or more adults residing there. Measuring stigma surrounding mental health services was trickier.
“Unfortunately, while such stigma is widely acknowledged, there has been very little research done to measure and quantify it,” says Erol.
For their study, the researchers computed stigma as the degree of mental health underreporting in a community. They did so by comparing the rate of mental health claims data by state in 2018 and 2019 to the self-reported rate of receiving mental health services in government surveys. “A high rate of underreporting in a state thus indicates stigma — many more people receive mental healthcare than those willing to say that they do,” explains Erol.
A more targeted approach
In their analysis, Erol and Wu note that there are many well-understood barriers to mental healthcare; among them are inadequate supply of providers and services, difficulties with entering the healthcare system, high out-of-pocket costs, and lack of insurance.
The researchers underscore that the surge in telehealth has the potential to expand access, particularly, as they found, for new mothers dealing with mental health concerns. Their study, they believe, offers “actionable insights” for policymakers in deciding which telehealth services should be reimbursed.
“In general, our results suggest that telehealth availability is promising in increasing access to treatment for individuals with mental health conditions, but that physical privacy concerns can have contrasting effects — depressing utilization in areas where patients lack physical privacy for telehealth visits but improving utilization where there are concerns about physical presence in a healthcare setting,” they note in their paper.
Rather than a general approach to reimbursement decisions, Erol and Wu conclude that health system managers should target allocation of their resources “according to the areas and the patient pool they serve.”
Moreover, the study findings “are likely to extend beyond the maternal mental health setting,” Erol notes, since the advantages and disadvantages surrounding telehealth and associated privacy concerns “are not exclusive to mothers or mental health.”
