Outreach via electronic patient portal ineffective at reducing self harm, study finds

ROCHESTER — A large randomized trial has found no benefit to contacting mental health patients through electronic patient portals in an attempt to reduce self harm.

The disappointing results, reported earlier this month the journal JAMA Network Open , come as health officials continue to sound the alarm over a national mental health crisis which has worsened during the pandemic.

"In the last 30 years there's been questions about how do we screen for mental health in primary care," says Dr. Mark Williams, a Mayo Clinic psychiatrist who was not involved in the study, but who has overseen similar low-intensity interventions for depression.

"Which of course that means you find that people have depression or high suicidal thinking. A lot of health planners ...want to find something they can use to help those patients. The challenge is that simple solutions aren't always the best."

The study, conducted by researchers from HealthPartners in Minneapolis and Kaiser Permanente in Seattle, Denver and Portland, was conducted between March of 2015 and September of 2018, and followed until March of 2020.

The trial recruited over 18,000 adults undergoing outpatient or primary care treatment for mental health problems.

Those who answered a screening question affirming frequent suicidal thoughts were randomized to one of two intervention groups or usual care, then invited via portal or follow up calls to participate.

Of those who took part, some received portal or telephone contacts from an outreach worker operating within a program known as care management, while others received a condensed version of an effective suicide prevention treatment known as Dialectical Behavior Therapy. Both methods studied were delivered in addition to ongoing mental health care.

Care management, also sometimes known as care coordination, is used widely in the treatment of depression. It involves the utilization of a non-clinician who periodically checks in with patients through the portal to gain insight on their progress, offer suggestions, provide questionnaires and update a clinician.

In the latest study it was delivered to 6,230 patients and entailed words of care and encouragement through regular online contact but little structured treatment.

"These were not therapists," says Williams. "They couldn't see people in person and they didn't have a specific-evidence based approach to suicide. It was mostly to get people to do whatever was available."

The Dialectical Behavior Therapy arm of the trial sought to teach 6,227 patients four subset skills of the therapy, utilizing video training modules and periodic portal contact to offer words of care and encouragement but none of the face-to-face, telehealth or group therapy elements of standard Dialectical Behavior Therapy.

Compared with usual care alone received by 6,187 patients, the care management group saw no significant reductions in self-harm, while those who took part in Dialectical Behavior Therapy skills training had a higher risk of self-harm than those receiving usual care.

The study had a low recruitment rate — just over 30 % offered care management took part, while the percentage of those who acquiesced to trying Dialectical Behavior Therapy skills training was only 39%.

Once recruited, patients were even harder to engage. Care management participation dropped to 17% after 9 months and over 90% of the Dialectical Behavior Therapy group stopped after the introduction to the skill modules.

Williams says the study may have illustrated the fact that effective treatments cannot be easily condensed without study, and that patients present with unique circumstances, requiring in the case of care coordination outreach that is tailored to the prevention objective and their unique situation.

"I think the challenge with suicide is that there are so many paths leading to that behavior, that you're always oversimplifying in some way without understanding the person in front of you...there's so many variables."

"I was concerned it was the kind of approach that an administrator might say, 'let's do something about suicide,' but not the kind of approach a clinician might do, which is to say, 'let's make sure...that we have ingredients that have been shown to reduce suicide.'"

On Twitter, lead author Dr. Gregory E. Simon of Kaiser Permanente Washington Health Research Institute cautioned that the results should not be taken as a broad rejection of the treatments studied or the technology utilized.

"We can't know if these results apply to any other supported online interventions," Simon wrote. "One point to emphasize," he added, "most studies of supported online interventions include only people who volunteered to receive them. That's very different from the group of people we studied."