Credit: Pixabay/CC0 Public Domain

An injury or sudden illness serious enough to require hospitalization is a high-stress experience for anyone. For most people, that stress is temporary. For others, however, the episode can lead to lasting mental health distress.

Hospital clinicians have had no easy way to predict who will bounce back psychologically after emergency care and who will continue to struggle.

Stanford Medicine researchers aimed to change that. They created a 10-question screening for patients hospitalized after emergency care to assess their risk of developing mental health problems down the road. In a study, they found the screening could identify 79% of patients who had post-traumatic stress disorder, depression or anxiety two months later.

The findings give clinicians the chance to intervene early and offer preventive resources to vulnerable patients.

Screening for the future

“If you’re admitted to the hospital after a car accident, you’re going to show stress symptoms,” said David Spain, MD, professor of general surgery and senior author of the new study published Oct. 1 in PLOS ONE. “A lot of those people will get better with time, especially if they have good social support, help at home, friends and family. But some people will not.”

In his experience treating patients with traumatic injuries, Spain knows that long-term psychological impact doesn’t necessarily correlate with the severity of injuries. “If you’re in a very bad rollover car crash, even if you didn’t really get hurt, the event could still be very scary and trigger emotions,” he said.

Currently, trauma patients are screened for pre-existing mental health problems, Spain said, but not for their future risk.

Illnesses that require emergency care, from a severe bout of pneumonia or COVID-19 to a heart attack or stroke, can also leave some patients with long-term stress. And the hospital environment can add to a patient’s feeling of loss of control.

“The problem is, if we wait a couple of months until we can tell who’s not getting better, it’s a little too late,” Spain said.

Roughly 40% of patients admitted after emergency care will have elevated levels of PTSD, depression or anxiety symptoms two months later, according to Eve Carlson, Ph.D., clinical professor of psychiatry and behavioral science, lead author of the study, and a researcher with the National Center for PTSD. “Providing preventive mental health care soon after exposure to traumatic stress can prevent or reduce later mental health problems,” she said.

Ten questions

The researchers developed the new screening tool, known as the Hospital Mental Health Risk Screen, in a study published earlier this year. They included more than 1,000 adult patients admitted after emergency care, except for those patients admitted primarily for psychiatric reasons.

They tested more than 100 questions taken from 14 existing mental health screening tools. Two months later, they analyzed which questions and which scores were strongly linked to the patients’ mental health outcomes.

They found they could whittle the long list of questions down to just 10 and still correctly identify three-quarters of at-risk patients.

“Our goal all along was to make something that could be very low resource,” Spain said. The screening can be completed on an iPad, in just five minutes, and requires no special training to administer or score.

The 10 questions relate to past, present and future emotions, including how much patients felt respected; their history of anxiety and depression; their feelings of isolation and stress since coming to the hospital; and how stressed they expected to feel over the next month.

Prioritizing sensitivity

Importantly, the latest study validates the screening tool in a new set of patients. This is essential to confirm that the screening is generalizable, the researchers said.

The new study included adult patients from three hospitals in California, Ohio and Washington, DC, who had diverse educational, economic and ethnic backgrounds. Patients took the screening within a few days of being admitted and could do so in English, Chinese or Spanish. About 450 patients completed the study, including patients from five self-identified ethnoracial groups: Asian American/Pacific Islander, American Indian or Alaska Native, Black, Hispanic, and white.

At the two-month follow-up, the patients’ mental health was assessed by standard measures for depression, anxiety and PTSD.

Overall, the screening correctly predicted 79% of patients who had a mental health diagnosis at two months—a measure known as sensitivity. It correctly predicted 72% of patients who had no mental health diagnosis at two months—a measure known as specificity.

The screening predicted well for all ethnoracial groups, with slightly higher sensitivity for multiracial, Asian American/Pacific Islander, Black and Hispanic patients and slightly higher specificity for white and Hispanic patients. There was no difference between injury and acute illness patients.

In choosing the cut-off score that distinguishes low-risk from high-risk patients, the researchers prioritized sensitivity over specificity—meaning they prioritized capturing all high-risk patients, even if some low-risk patients were included.

“There are some people you may refer for help and resources who don’t need it, but you want to make sure you don’t miss people who are going to have symptoms,” Spain said.

The next—and more challenging—step is to evaluate resources that could help those at high-risk to lower their risk.

“Very few hospitals in the U.S. have the capacity to provide professional preventive mental health care,” Carlson said. “Unfortunately, most referrals to mental health care don’t result in care being obtained.”

Instead, the researchers are developing self-help programs that patients could use at home that would direct them to professional care if needed. Coaching programs for family members to better support patients could also help.

Though designed for the hospital, the new screening could be adapted for survivors of other types of trauma—such as natural disasters or mass shootings.

“Based on what I’ve seen, I think it would work in other situations,” Spain said.