From the Editor
“A panel of medical experts on Tuesday recommended for the first time that doctors screen all adult patients under 65 for anxiety, guidance that highlights the extraordinary stress levels that have plagued the United States since the start of the pandemic.”
So reports The New York Times late last month with news of the US Preventive Services Task Force’s draft recommendation. The article quotes panelist Lori Pbert (of the University of Massachusetts): “Our only hope is that our recommendations throw a spotlight on the need to create greater access to mental health care – and urgently.”
In the first selection, we look at the recommendation. In making it, the USPSTF reviewed the literature and weighted the advantages and disadvantages of screening. If finalized, the recommendation would have implications on primary care in the United States – and beyond. “The USPSTF concludes with moderate certainty that screening for anxiety in adults, including pregnant and postpartum persons, has a moderate net benefit.” Is this a step in the right direction? Is this well intentioned but problematic?
In the second selection, Barclay Bram writes about his experiences with a therapist bot, working with the Woebot app. In a long New York Times essay, he talks about his depression and his therapy bot. He writes: “Using Woebot was like reading a good book of fiction. I never lost the sense that it was anything more than an algorithm – but I was able to suspend my disbelief and allow the experience to carry me elsewhere.”
Selection 1: “Screening for Anxiety in Adults”
US Preventive Services Task Force, September 2022
Anxiety disorders are commonly occurring mental health conditions. Anxiety disorders include generalized anxiety disorder, social anxiety disorder, panic disorder, separation anxiety disorder, phobias, selective mutism, and anxiety not otherwise specified. Anxiety disorders are often unrecognized in primary care settings and years-long delays in treatment initiation occur. Anxiety can be a chronic condition characterized by periods of remission and recurrence. However, full recovery may occur.
So opens the USPSTF draft recommendation statement.
They look at different aspects of screening for anxiety.
“Brief screening tools have been developed that may screen for anxiety and are available for use in primary care.”
Treatments or Interventions
“Treatment for anxiety disorders can include psychotherapy (e.g., cognitive behavioral, interpersonal, family, and acceptance and commitment therapy) and pharmacotherapy (e.g., antidepressants, antihistamines, beta-blockers, anticonvulsant medications, or benzodiazepines).”
“Adequate systems and clinical staff are needed to ensure that patients are screened and, if they screen positive, are appropriately diagnosed and treated with evidence-based care or referred to a setting that can provide the necessary care.
“Potential implementation barriers of screening include provider knowledge and comfort level with screening, provider access to effective screening instruments, and impact on care flow. Clinicians should aim to develop a trusting relationship with patients by being sensitive to cultural issues and free of implicit bias. Many individuals who may be screened for mental health conditions do not receive adequate treatment. Less than half of individuals who experience a mental illness will receive mental health care.”
Potential Preventable Burden
“Anxiety disorders have long-term effects that include impaired quality of life and functioning and sizeable economic costs.”
“Potential harms of screening questionnaires include false-positive screening results that lead to unnecessary referrals, treatment, labeling, unease, and stigma. Some anxiety screening tools used in adults emphasize sleep and other somatic symptoms that are also results of aging.”
The panel also did a literature review; we highlight:
- Benefits of early detection and treatment. “The USPSTF found two randomized, controlled trials (RCTs) (n=918) that directly evaluated the benefits of screening for anxiety disorders in an adult population in primary care settings.” They also make mention of: “Twenty-one RCTs (N=4,929) and eight evidence synthesis reviews (ESRs) (~144 RCTs; N≈11,030) assessed the benefits of treatment of anxiety disorders with psychological interventions.” They did look at subpopulations.
- Harms. “The two trials (n=918) that directly evaluated screening of anxiety did not report harms, and there was no pattern of effects indicating harms.1 None of the RCTs or ESRs of psychological treatment reported on adverse events. Three RCTs (N=669), 10 ESRs (~112 RCTs; N≈29,674), and two case-control studies (N=262,3780) addressed the harms of pharmacologic treatment. Most evidence occurred in general adult populations. Evidence demonstrated an increase in nonserious harms (defined as any adverse events and withdrawals due to adverse events). Serious adverse events were rare…”
A few thoughts:
1. This is a well-argued recommendation.
2. It is a comment on our times that a major panel in the United States is focused on a mental disorder, and that the media coverage includes an article in the front section of The New York Times. #Progress
3. Their argument in a nutshell: anxiety disorders are common and have significant morbidity; screening is straightforward; the interventions work; therefore, we should screen more.
4. But is this recommendation practical? To play the devil’s advocate: just because the screening can be done, would it actually improve outcomes? Would it instead soak up scarce resources but not result in better treatment?
5. The USPSTF has a history of suggesting that more screening would be helpful, including a 2016 recommendation that called for pre- and post-natal screening for depression. I asked Dr. Scott Patten (of the University of Calgary) for his thoughts.
“In my understanding the USPSTF position on pre- and post-natal screening remains something of an anomaly. If you look at the latest Canadian guideline, I think they do a great job in clarifying some of the terminology, which should help to sidestep some of the issues: https://www.cmaj.ca/content/194/28/E981
“They recommend against ‘instrument-based screening’ but are careful to stipulate that they are not against appropriate inquiries about the important dimensions of mental health as a component of good clinical care, or other efforts that would achieve the same ends probably much better, e.g., mental health literacy and reduced stigma among patients, better clinical skills and awareness among professionals.
“With respect to anxiety I haven’t gone through the evidence review for their draft anxiety screening recommendation – the relevant section in the public comment document starts with ‘The direct evidence for anxiety screening was extremely limited and did not suggest a benefit.’ So, I suspect that the draft recommendation is based on indirect evidence – which is an argument that we hear often: ‘anxiety is common, it is a negative factor vis a vis health, it is often overlooked and can be treated’ – all of which is taken to mean that it is self-evident that screening should be implemented. However, universal screening consumes vast resources (usually directed from those with greater need to those with lesser needs), is a logistic nightmare and has never been sustainable in any setting that I’m aware of, let alone primary care. I think you need direct evidence of benefit to be able to say that formal screening would be the best strategy to deal with the problems listed above.”
The full USPSTF draft recommendation statement can be found here:
Selection 2: “My Therapist, the Robot”
The New York Times, 27 September 2022
“I first met Woebot, my A.I. chatbot therapist, at the height of the pandemic.
“I’m an anthropologist who studies mental health, and I had been doing fieldwork for my Ph.D. in China when news of the coronavirus started spreading. I left during Chinese New Year, and I never made it back. With my research stalled and my life on hold, I moved back in with my parents. Then, in quick succession, I lost a close family member to Covid and went through a painful breakup. I went months without seeing any of my friends. My mental health tanked, as it did for so many.
“I was initially skeptical of Woebot. The idea seemed almost too simple: an app on my phone that I could open when I needed it, type my hopes, fears and feelings into, and, in turn, receive A.I.-generated responses that would help me manage my emotions.”
So begins an essay by Bram.
He notes the critiques of apps: “How could an algorithm ever replace the human touch of in-person care? Is another digital intervention really the solution when we’re already so glued to our phones? How comfortable was I being vulnerable with an app that could track my data? Spending time with Woebot didn’t really bring me answers to these important questions.”
He writes that he was “depressed and anxious” and unable to afford a private therapist. “So, despite my doubts, I reached for the algorithm.”
“Over time, I noticed various exercises I did with Woebot rubbing off in my daily life. Woebot taught me how to set SMART goals – specific, measurable, achievable, realistic and time-limited. Out went ‘I need to finish my Ph.D.’ In came ‘Let’s write 500 words every day for the next six months.’ Out went ‘I have to find a way to get through this lockdown without killing my parents.’ In came ‘I’m going to go for extremely long solo walks in the park.’ Woebot isn’t the kind of mystical guru you go on an arduous journey to consult. Its guidance was practical and grounded to the point of feeling obvious. But through repetition and practice, it did start to amount to something more than just some prosaic words. It felt clichéd sometimes, but maybe that was the point. Perhaps everyday healing doesn’t have to be quite so complicated.”
The essay observes that “A.I. chat therapists have been rolled out in settings as diverse as a maternity hospital in Kenya and refugee camps for people fleeing the war in Syria, and by the Singaporean government as part of its pandemic response.” He provides examples:
- “In Britain, bots are being trialed to bridge waiting times for people seeking therapy but unable to get appointments and as an e-triage tool.”
- “In the United States, some apps are getting recognized by the F.D.A. and are in trials to be designated as clinical interventions.”
- Chatbots have also been used in maternity wards in Kenya. “When I suggested this to Eric Green, an associate professor of global health at Duke University who ran the A.I. chatbot trial in Kenya, he was unfazed. ‘You can’t replace something that doesn’t exist,’ he said. As he pointed out, globally, more people have access to phones than to mental health professionals.”
“I don’t believe that Woebot is the cure for serious mental illness, nor do I think that an algorithm is ever going to fix my daddy issues. But that doesn’t mean there isn’t a place for it. As Alison Darcy, Woebot’s creator, told me, it is not designed to displace other forms of healing. ‘In an ideal world, Woebot is one tool in a very comprehensive ecosystem of options,’ she said.”
A few thoughts:
1. This is a well-written essay.
2. The author makes good and interesting observations. He doesn’t see Woebot as a human-like therapist, but a guide to his CBT work. Is Woebot a techie version of a self-help book? Is this a meaningful new approach to engaging patients?
3. The essay is personal and largely focused on the writer’s experiences. Wired just published a less enthusiastic article on bots. The author notes various problems, including a general lack of evidence. “Research to support their efficacy is scant and has mostly been conducted by the companies that have created them. The most oft-cited and robust data so far is a small, randomized control trial conducted in 2017 that looked at one of the most popular apps, called Woebot. The study took a cohort of 70 young people on a college campus, half of whom used Woebot over a two-week period, with the other half given an ebook on depression in college students. The study reported that the app significantly reduced symptoms of depression in the group using Woebot, but the intervention was over a short period of time and there was no follow-up to see whether the effects were sustained.”
Dr. John Torous (of Harvard University) comments: “We don’t want to be too cynical – we’re excited about innovation, we should celebrate that. But we certainly don’t want to celebrate too early.”
That article can be found here:
4. Chatbots have been considered in past Readings. For example, we looked at the Vaidyam et al. review from The Canadian Journal of Psychiatry. The authors (who include Dr. Torous) find that participants were generally accepting of the bots. You can find that Reading here:
The full NYT essay can be found here:
Reading of the Week. Every week I pick articles and papers from the world of Psychiatry.