The New (Experimental) Ways To Tackle The Mental Health Crisis
Get Out-Of-Pocket in your email
Looking to hire the best talent in healthcare? Check out the OOP Talent Collective - where vetted candidates are looking for their next gig. Learn more here or check it out yourself.
Finance Associate - Spark Advisors
- Spark Advisors helps seniors enroll in Medicare and understand their benefits by monitoring coverage, figuring out the right benefits, and deal with insurance issues. They're hiring a finance associate.
- firsthand is building technology and services to dramatically change the lives of those with serious mental illness who have fallen through the gaps in the safety net. They are hiring a data engineer to build first of its kind infrastructure to empower their peer-led care team.
- J2 Health brings together best in class data and purpose built software to enable healthcare organizations to optimize provider network performance. They're hiring a data scientist.
Looking for a job in health tech? Check out the other awesome healthcare jobs on the job board + give your preferences to get alerted to new postings.
Last time, we talked a bit about why mental health companies were all going after the employer market and why I thought that was problematic. A lot of you responded with 😬😬😬.
Most companies in the mental health space seem to have focused on new distribution channels, increasing demand for mental health services, and matching between patients and providers in some capacity. They’re important for sure, but don’t address the root of the problem which is the lack of clinicians and how we deliver care.
But I don’t like pointing out the problems without suggesting some solutions myself.
I find myself more interested in ideas around changing the format of care itself, the people giving it, and how we measure it. This is the only way to achieve the key things we need in mental health today: increasing and bolstering the supply-side of the people delivering care, triaging people by severity, and paying for outcomes/progress.
Here are some of the areas I’m interested in.
Coaching aka increasing access to care with non-licensed folks
We need to figure out how to expand the number of people that can deliver care to people with mental health needs. The reality is that not every issue needs to be seen by a clinician.
Coaches are already becoming a more integrated part of the healthcare ecosystem. Coaches are generally less expensive than clinicians + more abundant so they typically have more touchpoints with patients.
There are a few reasons I like the use of coaches. The obvious is that they can extend the size of patient panels that clinicians care for helping with by triaging and handling less clinical tasks in a care team. Less obvious is that “coach” is a great prefix and I hope one day someone calls me “Coach Krishnan”.
Coaches can also extend the availability of contacting someone (e.g. if you want to talk to someone outside the normal care hours). One of the reasons this is possible is because coaches don’t need to go through the state-by-state licensing process, which makes it much more flexible to add coaches and support that might be in different time zones.
Plus, health coaches can specialize. For example, BetterUp uses coaches for people in certain roles like Sales Performance and you can imagine having more tailored coaching for other job-specific niches. Or, you might want to tailor coaching to specific demographic groups like MindRight, who’s working with managed Medicaid plans to offer tailored + culturally-responsive coaching to young adults. Since the barrier to becoming a coach is much lower than their clinical counterparts, this can be an excellent area to upskill people that come from backgrounds where access to care might be hardest or require the most tailoring/empathy. I need a coach who’s also an only child to tell me I’m well-adjusted…right?
One key development here is that CPT codes for coaching were introduced within the last couple of years to make it easier for coaches to bill insurers for services. And last year In April, Cigna became the first health plan to reimburse for behavioral health coaching. Companies that are on the hook for a patient’s total cost of care have already been using coaches (e.g. Iora Health) but now the payment rails exist for other companies in fee-for-service contracts.
One of the open areas for coaching is what the baseline level of training and accreditation should look like. Right now basically anyone can call themselves a “coach”, so there’s a ton of variance in the quality. Little Otter, a children & family mental health company (disclosure: I invested) includes parent specialists in their care teams that are BCBA certified and trained in-house on evidence-based practices. Ginger requires a behavioral health coach to have:
- A minimum of two years of relevant experience, 6 months of which must have occurred with direct supervision, under a qualified, credentialed or licensed supervisor
- MA degree in psychology-related field OR coaching certification from an accredited training program (as approved by the NBHWC) [NK note: This organization has created their own standards for coaching.]
I’m in the camp that more people should have coaches. Sometimes you just want to talk to someone that’s seen a lot of people like you who can tell you that things will get better, give some actionable tips, and just make you feel good about your situation. Some people’s mental health issues can be better solved by someone giving more personalized feedback and encouragement than someone clinical trying to superimpose clinical frameworks onto the conversation. Or sometimes you just want to talk to someone non-judgmentally that doesn’t cost $300/hr. Like a significant other. Sometimes you want a significant other.
Alcoholics Anonymous is one of the most successful interventions, especially relative to the cost. There’s a certain irony that Employee Assistance Programs started with trying to employ this peer model in the workplace but has morphed into this jank version of itself.
We talked previously about patient communities being effective but outdated in how they think about business models so I won’t rehash that here. However, I do still think there’s much more room for peer support to help solve some of these issues (hence my investment in Most Days). We’ve already seen this work in more complex behavioral health issues like substance use disorder. But I think there’s more room for experimenting tackling lower acuity issues which may actually be issues of loneliness, motivation, and general venting that peers similar to you that have gone through the issue themselves might be better suited to help.
One area that seems to be interesting are popular social media or cultural figures leading their own groups and workshops. For example, Schuyler Bailar (a formerly D1 trans athlete) hosts several workshops and support groups for people that identify as trans, LGBTQ+, etc.
Group Therapy + 1-1-Many Formats
We might be able to better match supply and demand by making it possible for clinicians to “see” more patients at once. One way to do that is by combining aspects of peer support with the guidance, oversight, and structure of therapy. Group therapy companies like Grouport, Sesh, and Pace are experimenting here, and Groups has been doing this in the opioid use disorder space where it’s a bit more tried and true.
Another way to “see” more patients is broadcasting sessions to a wider audience who might actually benefit as voyeurs to the session. I think of these as 1-1-many care delivery formats. Mindset uses celebrities to talk about some of the everyday struggles they go through which listeners might as well. Couples Therapy the show is helping people understand issues and coping mechanisms they can use themselves. Therapists on Tik Tok give tips on identifying or managing different mental health issues.
My favorite is Dr. K, who has a Twitch Stream where he has “discussions” that look like therapy with lots of other people watching in the chat and getting involved in small ways.
“ [Dr. K is] talking to someone who goes by Mini, who's got on similarly big headphones (hers have cat ears on them). Mini's talking to Kanojia — Dr. K, as he's known to his fans — about how she's suffered depression, low self-esteem, and panic attacks all her life. The thousands of people watching this on the live streaming platform Twitch offer a steady flow of heart emojis and condolences in the chat box. Mini reveals that she's talked with six therapists, but none have given her satisfying answers. "What do you want to hear?" asks Dr. K. She pauses and fidgets and then answers, "what's wrong with me?" The heart emojis come even faster now.
This type of exchange is common on HealthyGamer_GG Twitch channel, where Kanojia will talk to people about their mental health issues, lead group discussions and guide viewers through various meditation techniques. The channel's rising popularity coincides with the growth of Twitch as a platform, the pandemic forcing us all inside, and growth in mental health issues.”
The line between entertainment and help is blurring. It’s becoming increasingly clear that there are lots of downsides and negative mental health effects to social media. But if people are going to keep spending more time on it, we should experiment with more ways to create positive mental health experiences for users.
Digital Biomarkers For Mental Health Tracking
IMO one of the biggest reasons mental health is difficult to innovate on is because it’s basically the last specialty to actually have a biomarker to track improvement. The scales we use today for mental health like PHQ-9 for depression or GAD-7 for anxiety are outdated. Plus, they rely largely on self-reporting from patients. Who doesn’t love regularly filling out surveys with uplifting questions like this? There aren’t enough $10 Amazon gift cards in the world to motivate me to fill these out.
These issues manifest very differently for each patient and we try to shoehorn them into standardized scales because they’re easier to measure vs. create personalized baselines per person that can help us actually understand improvement. Think about the fact that major depressive episodes are characterized by any of the 5 symptoms below. There are so many permutations that patients could experience and they’d all be classified in the same bucket. Some of the symptoms are literally opposites, and many of these symptoms might overlap with other behavioral health issues which makes accurate diagnoses very difficult.
I’m interested in new digital biomarkers that might give us more insight into how individuals are progressing with their mental health, give us more specific information to better diagnose patients, and potentially create individualized baselines for patients. While the verdict is still out on whether any of these biomarkers will be usable, I’ve generally seen the approaches fall into a few different camps or some combination of them to produce composite scores:
- Voice biomarkers - Companies like Kintsugi, Ellipsis, and Sonde are examples here. Your voice can be a window into how you’re feeling, so why not build a Shazam…of sadness. Psychologists say when you’re depressed it can manifest in your voice as being more monotone, softer, more pauses, etc. Anxiety manifests as tension which also can change the quality of your voice. Companies like Winterlight Labs are trying to do this for more severe issues like dementia.
- Mobile Activity Mining - Several companies are trying to monitor things like scrolling patterns, how you type/swipe, etc. to see if it indicates changes in your mental health status. This is where Mindstrong started (though seems to have expanded into other services). Not sure why you need to do research on this, just monitor if there’s an increase in “swipe left” actions or an increase in typing “t-w-i-t-t-e-r” on the phone keyboard and you’ll know if they’re depressed.
- Activity Tracking - With existing wearables and other sensors, we can measure things like physical activity, locations traveled, etc. which seem to have some correlation with mental health status. This meta-analysis points out all the massive issues with reproducibility and lack of standardization across the wearables/data outputs, but notes some common features in areas like sleep and sedentary lifestyles that seem to be connected with worsening PHQ-9 or other mental health areas. I’m curious how they control for me sitting in one place to binge “is it cake?” on Netflix when it releases…unless that IS depression.
- Facial Recognition - There’s been research in more serious mental/behavioral areas like autism and schizophrenia to use facial recognition software combined with some exercises to understand changes in eye contact, muscle movement, etc. to understand disease severity and progression. Cognoa has done a lot of cool work here. Could this be expanded to other areas?
- Brain Data - Researchers have been trying to throw machine learning against existing datasets to see if there’s any signals we can use to connect brain activity with these mental health disorders. Using machine vision on MRIs is one area being examined. For example, this pretty large and impressively designed study saw thinner cortical gray matter in certain parts of the brain for patients with major depressive disorder vs. the control groups. Another is EEGs, those electrode brain cap things that every movie uses when someone uses “100% of their brain”. They produce a ton of data, but new biotechs are using them as biomarkers to understand if their drugs are working or using them as inclusion/exclusion criteria for patients in trials.
All of these seem to be in the research phase and there are still very important questions around things like data privacy, algorithmic bias, and how this kind of data gets factored into downstream care. Pretty much everyone I talk to in the field (who don’t work at companies tackling this problem) are pretty pessimistic we’ll ever actually be able to use tools like this. But if we actually can figure out better measurements for tracking mental health progress, we might be able to experiment with new payment models like quality bonuses and also triage patients appropriately to the right level of care like a coach, therapist, psychiatrist, etc.
On top of that, this data can help keep patients engaged between sessions and help clinicians tailor treatments better, especially if the data actually informs workflows. Already, some companies are incorporating these tools for risk-assessment. Kintsugi and Mindstrong combine their biomarker with care delivery. Sondermind, a company that helps patients get matched with a therapist that’s in-network, acquired Qntfy, which uses machine learning on different mental health + biometric data to make treatment recommendations. I’m sure we’ll see more companies that incorporate these digital biomarkers into care.
Timeboxed Mental Health “Courses”
I’m interested in the idea of “tracks” for patients that want to tackle some specific area of mental health within a specific time frame. This isn’t going to be for everyone, obviously, but I think courses like this can combine a lot of the stuff above and also requires input from patients themselves (e.g. journaling, doing mindfulness exercises, etc.). By combining different modalities together into a course, higher level clinical staff can step-in in specific parts of a patient’s care journey when it’s most relevant while a patient is still doing activities themselves to feel progress. This is another way of letting clinicians see more patients at once and also gets more information from patients between visits that keep them engaged + can help understand their progress. Real is an example with different tracks + events + broadcasted content. Charlie Health is a 6-10 week intensive outpatient therapy program for teens and young adults who need therapy more than once a week that also combines aspects of group therapy. Meru has a 12-week program that combines biomarker monitoring with peer-to-peer support and therapists. This might be the first bootcamp I actually get something out of (looking at you, General Assembly front-end web development course).
Infrastructure For Virtual Therapist Practices
I think we could expand the size of patient panels for clinicians if we made their tooling better. This means building infrastructure for clinicians to:
- Start their own practices.
- Add tools to choose how they deliver care (e.g. asynchronous care, patient at-home exercises + tracking, digital therapeutic prescribing, video, integrating care teams, etc.)
- Reduce the amount of time spent on back-office processes like dealing with insurance, payment, accounting, etc.
Companies like DocSpace are helping with several steps of this process (disclosure: I’m an advisor). Headway/Alma/Grow Therapy are working on the credentialing and insurance parts, while Heard is working on the accounting piece.
If you build a new infrastructure for therapists you can use modern tech tools to give them more time in the day to potentially see more patients than a traditional practice by making them more efficient and making it easy for them to integrate asynchronous + synchronous tools into their workflows. On top of that, if you make it really easy for them to start a practice then it’s also easier for them to moonlight on off hours.
Plus, the practices can then be more customized. Maybe the clinician wants to choose a very different care setting based on the target demographic they want to serve. A totally virtual clinic that uses avatars instead of real faces might feel more accessible to some people. Therapy during a hike or walk can help show therapists different parts of body language, something that became popular during COVID. Not only could this reduce no show rates if patients are more excited about the activity during therapy, but it can also be a much lower operating cost for the physician who no longer needs the expense of the office itself or the patient acquisition costs. All of this needs infrastructure that’s flexible, customizable, and easy-to-start for mental health clinicians.
New Pharmaceutical Therapies and Medical Devices
Mental and behavioral health issues really haven’t seen that many novel drugs hit the market in the past few decades. We’re just finding new ways to distribute the same drugs. Today you can get Adderall on the web instead of that friend of a friend you met at that party one time and don’t remember their name but now it’s too awkward to ask so you just say “what’s up man”.
Traditionally it’s been difficult for drugs to target these diseases because the placebo effect is strong (especially in low-moderate severity issues), making it tough to demonstrate efficacy. Plus, we don’t have a great understanding of the underlying issues that cause these diseases and the side effect profiles of drugs in this area can be pretty gnarly.
However we’re starting to see a wave of new developments that might be changing things:
- There’s a recreational drug renaissance. Lots of really promising therapies using MDMA, psilocybin, and ketamine seem to be working well against different mental issues like PTSD, treatment-resistant depression, and more. Spravato is a ketamine spray that’s already on the market. The most significant side effect here seems to be all my microdosing Williamsburg friends not shutting the fuck up about these developments.
- Digital therapeutics seem to be working in some capacity, especially in areas where the existing drugs have a strong side effect profile or there aren’t many options for low-moderate severity of a disease (e.g. ADHD). We talked about this in a previous post + the issues around payment.
- Transcranial Magnetic Stimulation (TMS) seems to work pretty well for patients. About 50% of patients with treatment-resistant depression have some clinically meaningful improvement, and now the protocols for care are getting more targeted with even better results. TMS is something that needs to be given semi-regularly, but making this more accessible can meaningfully help people with depression and potentially expand to other areas.
The nice part of these interventions are that they’re scalable, have strong effects, are sometimes actually curative, and cheaper forms eventually become available in the form of generics. Compared to services which get more expensive over time and considering we’re already suffering a serious shortage, that seems like it could be pretty helpful.
However, understandably, there’s worry that these drugs could potentially have longer-term side effects, could be abused or overprescribed, they have overdose potential, and could potentially push patients to more harmful drugs/doses. Plus we still need to figure out a pricing mechanism that makes sense and doesn’t incentivize unnecessary prescribing.
There’s a tradeoff here that has to be made, patients will likely have to be very closely monitored, and abuse prevention mechanisms need to be well thought out.
The mental health crisis seems to be getting worse and we need to figure out ways to increase access to care and improve the existing treatment protocols.
I think also beyond these healthcare-specific fixes, a lot of mental health issues seem to be byproducts of loneliness. Before Out-Of-Pocket, I originally was planning to start a company that created an online-offline social network monetized by events to help combat this (turns out that’s a bad COVID business lol). But I think there’s something to be said about creating better ways to help people socialize/find their people. More public spaces and accessible events? Group-planner-as-a-business to help people congregate in more intimate spaces? Or maybe someone else can run with my old idea :)
Bringing fulfillment and community to people might end up having as large of an impact on mental health as any of the interventions we talked about, but outside of the traditional healthcare sphere.
Nikhil Krishnan aka. “N K-hole” aka. “future Coach Krishnan”
Other posts: outofpocket.health/posts
Thanks to David Ricupero, Morgan Cheatham, and Dr. Arpan Parikh for looking through drafts of this
If you’re enjoying the newsletter, do me a solid and shoot this over to a friend or healthcare slack channel and tell them to sign up. The line between unemployment and founder of a startup is traction and whether your parents believe you have a job.