Death by a thousand point solutions

Why we need less hyperspecialized care delivery organizations and more generalist behavioral health models...

TL;DR…

Behavioral health has seen an explosion of builder activity this decade 🚀. And this has (for the most part) been much needed. The gap between access to and supply of psychiatric care, and the need for more of it, is massive. Now that the attention of builders (and investors) has turned to behavioral health, it is incumbent on us (the industry at large) to create well-suited care models to serve the needs of patients 🤝🏽. There has been increasing activity over the last 12-18 months in the space we will call ‘point solutions’ – e.g., companies built around care models directed at very specific disease states. While helpful in increasing access for often underserved patient populations, this approach also creates more barriers to and siloes in the coordination of care and serves to fragment an already hyper-fragmented subsector of healthcare further. As we march towards a world in which the behavioral healthcare system becomes populated with more and more point solutions (focused on treating very specific populations), we worry about the impact this will have on: 1. The supply of clinicians to address the overall healthcare needs of Americans, and 2. The increasingly complex patient journeys that will arise.

Intro…

Before delving into the business side of things, let's focus on the clinical aspect 👩🏽‍⚕️. Psychiatric diseases rarely occur in isolation; they are often interlaced with one another. In psychiatry, we employ a holistic biopsychosocial model to understand the origin of the diseases we treat. This model outlines three driving forces behind psychiatric illnesses: biological factors 🧬 (such as genetic predispositions), psychological aspects 🧠 (such as mental resilience and coping mechanisms), and social influences 💵 (ranging from positive elements like stable income to negative ones like trauma).

Psychiatric diseases frequently demonstrate significant overlaps across these domains. Consequently, these illnesses often occur in pairs or clusters, which also feed off of and exacerbate one another. This makes it exceedingly challenging to treat one without addressing the others simultaneously. For instance, disregarding a patient's complex post-traumatic stress disorder can stymie effective treatment of a co-occurring major depressive disorder. See the below fantastic work showing the time-sequence relation between 1st diagnosis and 2nd diagnosis of psychiatric conditions out of the Danish National Research Foundation. Note that this graphic has been filtered only to show relationships that have a hazard ratio of 5; that is, a diagnosis of the ‘prior’ disorder leads to a 5x (or greater) probability of being diagnosed with a ‘later’ disorder at some point in the future 😲.

Ok, so why does all of that matter?

Given: 1. The prevalent co-occurrence of psychiatric disorders, and 2. The closely entwined nature of symptoms (leading to intricately linked treatment pathways required to achieve remission), point solutions aren’t always best suited to facilitate comprehensive, whole-person healing. 

Moreover, the rapid expansion of these point solutions could potentially complicate the process for both patients seeking appropriate care and for clinicians, such as those in primary care or emergency room settings, striving to refer patients to the most appropriate care providers 👩🏾‍⚕️.

This is not to say that there is not a place for specialists in behavioral health. There is a massive gap between supply (access to care) and demand (the need patients have). Oftentimes, it is the most complex populations that have the most difficult time seeking out care. Focused clinical models can play a role in caring for these patients. For example, the care of patients under 18 🧒🏽 requires additional nuance, structure, and support, and generally should be delivered by child and adolescent-trained clinicians. For this reason, there are a significant number of care delivery organizations and companies that focus on this unique population, without adding unnecessary fragmentation to the patient journey and without potentially diluting the supply of clinicians available to the average American suffering from psychiatric symptoms.

How this fragmentation impacts the system:

However, consider the plight of the average primary care or emergency room clinician, the two settings which likely account for the vast majority of first presentations of psychiatric symptoms and from where the vast majority of referrals to behavioral health clinicians originate (PCPs treat nearly 60% of patients being treated for depression and write 79% of antidepressant medication prescriptions). Primary care clinicians are generally aware of the broad categories and specialties in psychiatry; for example, child and adolescent for patients under age 18 or addiction medicine for patients with severe substance use disorders. However, the role of a primary care clinician is not necessarily to make a highly nuanced clinical diagnosis and referral decision; it is instead to identify the need for a specialist if the presentation is complex and then make the appropriate referral. Let us use a medical specialty to make our point even stronger:

Suppose a patient presents to their primary care doctor complaining of knee pain. After completing a work-up and imaging, the PCP determines the patient has an ACL injury and likely requires surgery. In today’s world, the PCP will refer the patient to a joint specialist (or perhaps a knee specialist). But imagine orthopedists went so far as to specialize in repairing ACLs on only one side of the body (e.g., ‘I only repair ACLs of the left leg’), or would not also address damage to the MCL (which frequently co-occurs with ACL damage) if found during ACL repair surgery. And therefore, imagine that PCPs had to manage this repository of information to ensure they referred their knee pain patients to exactly the right type of orthopedic surgeon. It would be unrealistic that a PCP would have the resources, knowledge, and time to refer their patient with knee pain to just the exact superspecialist who would treat their ACL injury.

Where’s the upshot here?

Given the potential for worsening fragmentation in what is already one of healthcare’s most fragmented and siloed subsectors, we see the proliferation of hyper-specialized point solutions driving two trends moving forward (which we believe will continue to proceed in tandem, not necessarily in competition):

1. Increased focus on and value from generalist behavioral healthcare delivery organizations that can address the needs of wide swaths of the population and deliver value to all the key 🔑 stakeholders (the patient, the primary care complex, the healthcare system at large, and payors and employers). By definition, these generalist behavioral health delivery models are most ready to serve as the first point of contact for Americans before escalating care to super specialists, if necessary. We see some specialization as natural and necessary when the specific patient population requires a dramatically different type of intervention and clinician – for example, the care of patients under 18 suffering from specific diseases, the care of patients who require more intense treatment (e.g., intensive outpatient [IOP] or partial hospitalization [PHP]), or the care of those suffering from severe and persistent psychiatric conditions (such as schizophrenia). However, we see these as the exception and not as the rule.

2. With the continued proliferation of point solutions, their success depends on the appropriate referral of patients who warrant a super-specialized approach to care. We suggest that no organization whose primary function is care delivery is equipped (nor should it be) to maintain this constantly updated repository of information. As such, we see the need for robust care navigation platforms whose primary function is to catalog the point solutions available in the market, their unique inclusion and exclusion criteria, and facilitate the referral of patients from generalist outpatient behavioral health settings to these more specialized levels of care. However, we see this as a potentially short-lived need.

Our hot take 🌶 is that these super-specialized platforms for care should (in an ideal world) rely not on referrals from primary care or direct-to-consumer efforts as their top of funnel but instead on referrals from generalist mental health platforms that have completed the appropriate initial triage, screening, and assessment to identify the need for such specialized care.

Not to be shills, it is worth noting that we’re not just talking the talk 🗣 here but also walking the walk 🚶🏽. We’ve both made very intentional decisions to join organizations - SOL Mental Health 🌞 and Bend Health 🌲 - that deeply believe in a comprehensive, generalist approach to addressing mental health. And we both continue to beat the drums in our respective organizations of expanding access and expanding care by widening the breadth of services we deliver (which we ought to note, resonates deeply with almost every potential strategic partner and stakeholder we speak with).

Our prediction of the future (we are psychiatrists after all, right?)…

Ultimately, we predict that the hyper-specialized behavioral health platforms will become acquisition targets for the generalist providers of care, which is when their true power will be unlocked as they will be available to increasingly broader swaths of the population. And perhaps this is one rational mechanism by which generalist organizations can develop the expertise to care for complex populations.

We’ll pose a provocative question: if the ultimate goal of creating a hyper-specialized behavioral health platform is an eventual acquisition by a generalist, are we missing the mark in building what is best for patients? Of course, one of the hallmarks of a free market is the motivation that future successful outcomes create for builders to start and grow new companies, so the answer is not black and white. However, if the proliferation of point solutions creates confusion in the market when it comes to patients seeking care (and results in worsened access to care since we’re all hiring from the same finite resource of mental healthcare clinicians), perhaps the means don’t justify the ends.  

So, is the short-to-mid-term pain we predict (increasing fragmentation and decreased supply of generalist clinicians) worth it for the long-term gain? Time will tell…

Two Questions:

1. For folks building and/or operating in other healthcare specialties, do you see analogous forces playing out in your vertical?

2. If you are a clinician or a patient, how has the proliferation of point solutions impacted your experience with the healthcare system (positively or negatively)?

✌🏽 A + A

To read more about our vision for the Stack, check out our intro post here.

Feedback? Find us on Twitter: Amit & Arpan