Higher Rate of ED Visits in Mental Illness Patients Explained

Batya Swift Yasgur, MA, LSW

October 22, 2018

Individuals with mental health diagnoses make 25% more visits to the emergency department (ED) than those without mental illness; increases in frequency correspond to illness severity, new research shows.

Investigators analyzed data on more than 3.5 million individuals. These persons accounted for more than 7 million ED visits. The investigators found that patients with mental illnesses visited the ED more frequently and that the increase correlated with illness severity, prior patient visit patterns, previous hospitalization, and high rates of lagged ED visits.

Socioeconomic factors also played an important role; neighborhood poverty as well having Medicaid or no insurance increased the likelihood of future ED use.

"We studied which factors were associated with higher future ED use among patients with and without mental health diagnoses [because] identifying characteristics related to increased ED use among patients with mental illness is an important step towards improving their healthcare access, quality, and costs," study coauthor Hemal Kanzaria, MD, associate professor of emergency medicine, University of California, San Francisco, School of Medicine, told Medscape Medical News.

"I believe patients would benefit from enhanced community and ED-based mental health treatment options, improved physical and mental health integration, and programs that address their health-related social needs," he said.

The study was published online October 19 in JAMA Open Network.

Understanding the Profile

Increases in the rate of visits to US EDs have led to increased costs, which in turn has led policy makers to "concentrate on ways to reduce potentially avoidable utilization" by understanding the profile of those patients who use the ED frequently, the authors write.

Approximately half of patients who visit the ED frequently have been diagnosed with a mental iillness. For these patients, rates of morbidity and mortality are higher, as are medical costs over time.

Moreover, up to 80% of patients with mental illness seek care in medical rather than behavioral settings, "where they often leave without treatment for mental illness," the authors note.

To help inform efforts to address high ED utilization among patients with mental illness, "it is important to understand factors associated with their increased ED use among patients with mental health conditions and also to recognize how they differ from factors that affect patients without mental illness," the study authors write.

No prior studies have examined this particular issue, they note.

To investigate the question, regression analysis was used to estimate which medical and social factors are associated with future ED utilization and how these factors differ between those with and those without mental illness.

Data were taken from the previous year's ED records for patients who presented to any California acute care hospital in 2013.

The researchers drew on data from California's Office of Statewide Health Planning and Development, which provides detailed information on all ED visits at licensed California hospitals and is a census of all hospitals and all payers in California.

For each patient, the first ED visit in 2013 was considered to be the index visit. Data from all ED visits in the prior 365 days were used to build a regression model that predicted ED use. Such use included inpatient hospitalizations originating from the ED at any California hospital over the following year.

The study focused on adults aged 18 to 64 years. The final sample included 7,678,706 ED visits by 3,446,338 patients during the 365-day look-back period.

The primary covariates of interest were related to the patient's mental health diagnoses during the year prior to the index ED visit. The researchers controlled for patient age, sex, race, insurance status, previous hospital admission during the look-back period, the poverty rate in the patient's ZIP code area, and whether the patient lived in an urban or rural county.

The researchers used a model in which the mental health diagnosis indicator variable interacted with all the other covariates to determine whether associations differed between certain factors and future ED use, depending on whether the patient did nor did not have mental illness.

Nonclinical Factors

Of the 3,446,336 patients, a little fewer than half (44.6%) were men, 36.6% were aged 18 to 30 years, 40.3% were aged 31 to 50 years, and 28.1% were aged 51 to 64 years.

The mean number of ED visits per patient per years was 169 (SD, 2.56); 29.1% had ≥1 mental health diagnoses.

The volume of visits was generally higher for patients with a mental health diagnosis than for those without such a diagnosis (~42% vs ~60%).

Among those with a mental health diagnosis, those with ≥4 ED visits (16.8% of patients) represented 39.6% of the total visits.

Admission of a patient to any hospital during the previous year was one of the strongest factors associated with ED use during the next year (incidence rate ratio [IRR], 4.88; 95% confidence interval [CI], 4.83 - 4.93), followed by high volume of ED visits in the previous year (IRR, 1.64; 95% CI, 1.62 - 1.66 for those with four to 10 visits; vs IRR, 1.97; 95% CI, 1.86 - 2.08 for those with 11 to 17 visits; and IRR, 5.91; 95% CI, 5.00 - 6.98 for those with ≥18 visits).

The researchers assessed nonclinical factors, including insurance coverage and neighborhood poverty, which also had a significant association with future ED visits.

Patients with Medicaid, Medicare, or no insurance in the previous year made more visits in the following year than those with private insurance.

Patients who lived in moderately poor and the poorest regions (middle and lowest terciles for median ZIP code area income, respectively) were predicted to have 11.1% (95% CI, 11.0% - 11.2%) and 11.7% (95% CI, 11.6% - 11.7%) more visits in the following year, relative to patients who lived in ZIP code areas where the groups with the highest incomes resided.

Mental health conditions were associated with the second largest increase in ED use, second only to diseases of the skin and subcutaneous tissue.

Those who had a mental health diagnosis in the prior year made more ED visits in the following year (IRR, 1.256; 95% CI, 1.22 - 1.29) than patients who did not have a mental disorder.

Additionally, an ED visit with a primary mental health–related discharge diagnosis in the previous year was associated with more ED visits in the following year (IRR, 1.309; 95% CI, 1.30 -1. 32) in comparison with a secondary mental healthy discharge diagnosis.

More severe mental health disorders were associated with more ED visits in the following year. Diagnoses classified as mild, moderate, and severe were associated with a 3%, 12%, and 23% increase, respectively, in future ED use.

"The locus of factors associated with ED use, while not necessarily causal in our analysis, suggest that prior patient patterns and patient illness severity could be important contributors to increased ED use," the authors comment.

"We also found that nonclinical factors matter; Medicaid coverage and lack of insurance were associated with greater ED use than continuous private insurance," they report.

"Our results are particularly relevant in states like California, where the state's Medicaid program has instituted changes in how it delivers and pays for mental health services under the Affordable Care Act," Kanzari observed.

A Bigger Problem

Commenting on the study for Medscape Medical News, Ron Honberg, JD, senior policy advisor, National Alliance on Mental Illness, said the study is "certainly focused on something that we've been aware of for some time, which is the disproportionate number of people with serious mental illness going to ERs, often spending long periods of time in those ERs because their symptoms are severe but there are no beds available."

He noted that the "ER personnel are doing the best they can and dealing with a very serious problem, and spending a lot of time and resources trying to help these folks in crisis, but are often ill equipped to respond effectively."

The crisis is "part and parcel of much bigger problems in our mental health field, specifically, the fact that we really have across the country severe gaps in our capacity to respond to crisis situations and in the availability of services for people once the acute crisis has been alleviated," said Honberg.

Kanzari agreed, adding, "Our results suggest there is room to improve care coordination, particularly for individuals with severe mental health diagnoses."

No funding for the study has been reported. Dr Kanzari, the study coauthors, and Dr Honberg have disclosed no relevant financial relationships.

JAMA Open Network. Published online October 19, 2018. Full text

For more Medscape Psychiatry news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.