Sean Luo
Source: Sean Luo

As someone who works with mental health technology development and research in Big Data and behavioral health, I like to scrutinize new developments in this space. One development is the emergence of “online therapy” services such as BetterHelp and TalkSpace. The latter of which has raised $15 million (Tech Crunch) and garnered mixed reviews including, here, here and here.

Whatever this is, it is getting a lot of publicity and I decided to give it a try. Over the Christmas break I reviewed two companies (who shall remain nameless) in this space by taking a trial of their products. While the marketing material and “culture” of the companies are different, the business models are similar.

To understand how and why this business model works, one has to understand how psychotherapy is delivered in the US in 2016. Therapy “providers” (thank you Managed Care for this colorless moniker) can be divided into two groups: doctoral and non-doctoral. Doctoral providers include psychiatrists (MDs) and clinical psychologists (PhDs, PsyDs). Non-doctoral providers include social workers (LCSWs), other masters level (LPCs and their equivalents), as well as non-masters level clinicians such as CASACs. The difference between doctoral and non-doctoral providers is that insurance reimburses doctoral providers more consistently and at higher rates: the doctoral level training process is longer and more complex, and this limits supply. Nevertheless, anyone who is certified through any of the above can provide psychotherapy.

Since psychotherapy is variable, and the quality of psychotherapy is difficult to measure, anyone who is certified to provide psychotherapy can get away with provide whatever he/she wants. This is very different from other medical services: even if there is no legal restriction for neurosurgery, very few non-neurosurgeons will start cutting into spinal cords. Such concerns are less apparent in psychotherapy. When people get hurt (usually due to a delay of evidence-based treatment), few are accountable, and very rarely do people get sued. Furthermore, in behavioral health treatment, the likelihood of adverse events are more correlated to severity of illness than the qualifications of the provider.

This generates a large supply of individuals who are legally allowed and qualified of provide psychotherapy of various types. Large supply of service provider with an inefficient market generates a very good opportunity for “Uber-ization”, and indeed the premise of this made venture capital go giddy.

The reality is what the membership buys is not psychotherapy, in the sense of any modern, evidence-based psychotherapy for a mental illness. After you put your credit card in and start the trial, the system refers you to a provider. After a few standard prompts, I realized that there is no real time feedback: the therapists check the messages on these platforms once or twice a day, and schedule in for “real time” chat or video sessions for additional fees. This makes sense: with such low fees (as low as $5 a day, and free to start), no psychotherapist would be able to sit there answering texts all day. However, this does not feel like psychotherapy. It feels rather Victorian: one composes a well written, thoughtfully long note pouring his or her heart out to the provider and hope that she (it is often a she) will respectfully write back in about 24 hours.

This is more aptly described as a therapy referral or case management service rather than actual therapy. From a clinical perspective, there is no evidence indicating that this is equivalent to regular, evidence-based psychotherapy sessions. It is possible that this kind of service can be equivalent to commonly used comparison groups such as “typical medical management” in clinical studies. At times and for certain indications they are just as effective or only marginally less effective than actual therapy. The ethical problem here is really false advertising...I thought I would get "therapy", but not really. Secondly, the platforms are cavalier about emergency management and legal regulations (i.e. HIPPA). But since there's no insurance company or regulator with their big hammer, these companies cleverly avoided doing their homework. MDs cannot practice medicine across state lines, and real telepsychiatry without a license could be a felony. Practicing psychotherapy by a non-MD across state lines does not carry the same legal weight and ethical stigma. Again, this seems to have worked so far because the stakes are so low.

But I think the real problem is a business, not clinical issue, and the companies will realize this soon enough: they haven't push the Uber model far enough. I suspect users drop out all the time because they don’t get real time feedback and this probably will restrict growth—the product is simply not all that engaging. On the other hand, I am sure there are providers who are aching for work at all times of the day. One could imagine a world in which whenever I request, I get a list of providers and their credentials who are currently “online” and can talk in real time (like the Uber app)--maybe one even physically located close to me! I would pay my fee, bypass insurance, get to talk to someone, and if I like it, keep this person for the next session, and if not, move on. This will create even more headaches for the providers, especially if money is not the limiting factor, as competition will be fierce (as they are now for taxi drivers and restaurateurs). Secondly I suspect certain demographic features would be in higher demand even with subpar qualifications. How and if the platforms need to comply with federal labor laws would be an interesting question, since providers are in theory not employees. 

Just how cheap is talking? Shockingly, one of the companies I have tried offers a free “listening” service. You get assigned to chat with someone (unlicensed) about your problems while looking at Google ads. For free. Guess where this person is from. If you said India, you would be right. This does give me a brief moment of Black Mirror-esque chills (where are we as a society if genuine human connections are outsourced?), until I realized that this is but a step on an inevitable path that started in the 90s with AOL chatrooms. 

You are reading

Hooked on Patterns

Being Bayesian in Insane Places

A brief discussion of the mathematics of Rosenhan experiments and follow-ups

Online Therapy: How Does This Work?

A review of this rapidly moving, possibly dystopic technology

Understanding Addiction With Electronic Registries

Disruptive technology with a $40 per patient clinical trial price tag.