Earlier this year, Ira R. Katz, M.D., Ph.D., and his colleagues at the Office of Mental Health and Suicide Prevention at the U.S. Department of Veterans Affairs (VA) released a ground-breaking study, finding that VA patients were at high risk for premature mortality after discharge from psychiatric hospitalization, especially in the first one to three months after discharge. The researchers looked at all-cause and cause-specific mortality of 3,829 VA patients who died within the first year after release from VA inpatient health units in 2013 and 2014. Deaths were classified as mortality due to suicide, external causes, or natural causes. External causes were those deemed accidental or undetermined, and fatalities that fell under natural causes included cancer, circulatory and lung diseases, dementia, and infectious diseases.
Dr. Katz says that while he expected suicide rates would be higher within 90 days post-discharge as compared to the remainder of the first year, he was surprised to learn that there was a disproportionate number of deaths from other external causes during that same timeframe. Slightly more than 19% of suicides after discharge occurred in the first 30 days, and 39.9% happened within the initial 90 days. In terms of other external-cause deaths, 13% took place in the first 30 days, and there were high rates throughout the year. “We were looking at what else happens to our patients and found that the rates of other external deaths in the year after discharge were greater than the rate of suicide among younger veterans discharged from our facilities. In fact, deaths from accidental overdoses and injuries were more frequent than deaths from suicide.” In older veterans, the rates were comparable, though the researchers did find an increase in natural-cause mortality in older patients with dementia.
Examining mortality after psychiatric hospitalization gleams opportunities for both suicide prevention and overall health care. Dr. Katz agrees with Michael Schoenbaum, Ph.D., Senior Advisor for Mental Health Services, Epidemiology, and Economics at the National Institute of Mental Health, that determining whether survival rates are improving in mental health over time requires tracking survival in the first place. While not often recognized as such, Dr. Katz notes that mental health illnesses are real diseases and that they can be fatal. Not only are there increased physical chronic health risks, including Alzheimer’s disease, diabetes, cardiovascular disease, and stroke, but also numerous other mortality risks such as overdose and motor vehicle collision. The results of Dr. Katz’s study paints a much more diverse picture of the increased mortality risks people face after a mental health crisis leading to hospitalization.
The study of deaths after hospitalization is part of a broader VA effort evaluating the increased mortality associated with mental health conditions in veterans. Americans are experiencing an uptick in what have been called deaths of despair—suicide, drug, or alcohol-related deaths. According to Katz, “It’s not just suicides that are on the rise. Deaths of despair, the aggregate of suicide together with drug and alcohol-related deaths, is going up, especially in unstable communities, and it’s having an impact on overall life expectancy in America. The VA is interested in how it affects veterans.” The term deaths of despair originates from the work of Anne Case and Angus Deacon, economists who have studied the rise of mortality due to suicide, drug and alcohol poisoning, and alcohol-related liver disease among Americans over the past two decades. A recent study published in JAMA supports many of their key findings, showing a rise in deaths of despair among Americans ages 25 to 64 across all racial and ethnic groups as well as an increase in deaths from a range of organ system diseases. In 2017, more than 152,000 Americans died from alcohol- or drug-related deaths or suicide, which is the highest number in U.S. history. Nearly half were a result of drug overdoses. Young people ages 18 to 24 are also particularly at risk, with an increase of 108% in drug-related deaths, a 69% jump in alcohol-related deaths, and a 35% increase in suicides between 2007 and 2017.
Whether accidental deaths are indeed accidental is challenging to determine. Dr. Katz says it can become a tangled web to try to differentiate between the two. Katz’s study found that in cases where the hospital noted suicidal ideation, then the official cause of death was likely to be suicide. This provides some degree of validation for the distinction between suicide and accidental causes in death records, but the question about individual cases remains. He says it’s not just overdoses that increase during the three-month timeframe after hospitalization but also motor vehicle and pedestrian accidents. Determining whether a death was the result of distraction, accidental in some other way, or intentional can become subjective. “There used to be a major debate in suicide prevention about whether the coroner or medical examiner’s death certificate diagnosis was reliable enough to support research and program planning. For me, this question is still not fully addressed. We have to recognize that what is a suicide, and what is not, is fuzzy. Was the person distracted when stepping in front of a car, or was it intentional?”
Most of the 106,430 VA patients discharged from psychiatric hospitalization in 2013 and 2014 were admitted to acute care primarily for a mental health condition and not a substance use disorder, yet they faced a high risk of overdose. Dr. Katz says this gives the mental health field hints about opportunities for prevention, including how dire it is that substance use disorders are addressed during hospitalization and after discharge, even if the person was admitted based on another condition. He thinks the field should take it a step further and, similar to the trend of hospitals giving naloxone kits to overdose patients, provide naloxone kits to patients discharged from psychiatric hospitalization who are at risk. “For those who are taking opioids for pain or have a history of misusing them, this is a life-saving, actionable step we can take. We should also teach them—as well as a close family member or friend—how to use the kit.”
Before discharge, many emergency departments and psychiatric hospitals do a personalized Safety Planning Intervention (SPI) with patients, helping people identify steps they can take when feeling suicidal. The intervention includes essential information such as warning signs that a crisis may be developing, coping strategies, people to turn to, and ways to make an environment safe. It’s been shown to be life-saving, reducing suicidal behavior by 50% and making it twice as likely that people will get treatment. Dr. Katz says that SPIs should expand to include steps for preventing overdoses and accidental mortality. “People are more at risk in the first 90-days after discharge. What can we build into the safety plan that may increase their awareness? We should be using an SPI also to create clear steps people can take when they experience substance cravings or if they are going to use, not to do it alone. This would give them a tangible way to keep themselves safe.”
Follow-up after hospital discharge is a widely accepted practice, but, similar to an SPI, it may need to be individually tailored. Dr. Katz and his colleagues were surprised when their study revealed that the VA’s policies about follow-up after psychiatric hospitalization were not as effective as they thought they would be. In 2009, Dr. Marcia Valenstein, a researcher at the VA, found higher risk periods for suicide among their patient population in the first three months after discharge. This led to a VA policy change, requiring that providers contact patients within one week after release, and for those recognized to be at high risk, providers must make contact with the patient four times within the next month. “Our policy adherence is close to 100 percent, and we wanted to see if follow up was making a difference. What we found is that the policy change helped to mobilize care, but it wasn’t enough to mitigate the increased risk of suicide after hospital discharge.” Dr. Katz says that while he and his colleagues were disappointed that the follow up done at the VA hasn’t been making the marked difference they’d hoped, it may mean that it needs to be more robust. “We may need to include other evidence-based interventions like telephone care management until the discharged veteran is fully engaged in treatment.”
In terms of what’s next, Dr. Katz and his team are further examining mortality differences after discharge across their 140 medical centers nationally to evaluate differences between facilities to see what’s working and what’s not. Disparities, he says, may result from the quality of care and available resources in the community. He says one finding thus far is that “it’s beginning to look like community support is far more important than we ever imagined.”
