Foresight

Berkeley, CA
Leveraging Technology to Reinvent Mental Healthcare
  • $6.7M revenue run rate; 30% monthly growth
  • Treated 4,000+ patients
  • Own/run 18 clinics; plans to expand to 50 by 2022

Bioverge Participation:

Seed
Big Data
AI
Mental Health

OPPORTUNITY OVERVIEW

All information contained herein is confidential and should not be shared or disclosed to anyone without explicit permission from Bioverge


Opportunity Overview

Foresight is a mental health and wellness company based in Berkeley, California. Through a network of clinics, Foresight provides patients with a wide range of treatment and services related to improving mental health and lifestyle.



Foresight is transforming the current standard of care from qualitative (i.e. "how are you feeling today") to be data-driven and objective.  The team comprises psychiatrists, therapists, software engineers, bioengineers, data scientists, and researchers. The company’s mission is to reinvent mental healthcare through the use of modern technology — enabling the delivery of highly personalized, data-backed treatment plans to each of their patients. 



The Problem

Did you know that depression is the leading cause of disability worldwide with 1-in-5 adults in the U.S. experiencing mental illness each year? Furthermore, did you know a staggering 50% of millennials and 75% of young adults in Generation Z have left a job for mental health reason and that the second leading cause of death for young adults in suicide?



To put these numbers in perspective, that means 44 million Americans are suffering from mental illness, which costs the U.S. $193 billion in lost earnings every year.

However, regardless of this massive need for treating mental illness, standard approaches are woefully outdated, incomplete and unscientific:

1. Psychiatry Practice is Outdated/Not Tech-Driven. Psychiatrists still practice the traditional “psychiatrist in armchair, patient on sofa” model, where a clinician attempts to diagnose a patient’s condition solely by asking a list of questions from a questionnaire that hasn’t been updated for 50 years. The diagnosis is based on these narrow qualitative parameters, without a scientific overlay, or personalized in other meaningful ways. Today, psychiatric diagnosis relies on educated guesses and doesn’t leverage data and technology.



2. Sub-optimal Drug Prescriptions are Common. When it comes to prescribing medication, a physician will typically consider only a small handful of factors: known contra-indications with a patient’s other drugs; whatever symptoms the patient is reporting; and potentially (and certainly not in all cases) a subjective family history. One of the key problems with this approach is that some conditions have 10-15 drugs that are candidate treatments and the physician must mentally cross-reference each of those drug’s potential adverse reactions against dozens of factors that influence likelihood of an adverse reaction. 

The set of factors that ought to be considered – but aren’t in today’s practice of medicine – include not only a patient’s symptoms and existing drug regimen, but also dozens of personal, family and genetic parameters. There is simply no way that any physician could retain an exhaustive mental repository of every possible adverse reaction, for every condition, across each drug that might be used to treat any given condition. Given the high caseloads of modern medicine, doctors have no time to spend days conducting the exhaustive cross-referencing research needed to identify all potential adverse reactions or other inferior outcomes tied to each prospective treatment drug. 



Most healthcare providers simply prescribe whatever drug they are most familiar with and most accustomed to prescribing for that condition. From the patient’s perspective, this is a terrible result. Patients should be being prescribed the right drugs, and the optimal combination of drugs, for their condition, not merely the most common or popular ones.

3. Outcome Monitoring is Outdated/Not Tech-Driven. Treatment outcome monitoring is infrequent (there may be weeks/months between sessions) and qualitative (“I feel a little better”).



4. Psychiatry Services are Unaffordable and Inaccessible. Accessibility and affordability are real problems for mental health care. You may need to wait 3 months+ for a psychiatric or therapeutic appointment and typical prices in the US can be $500+/hr. 55% of psychiatrists in the US don’t accept insurance, a number that grows to 75% in the Bay Area.



The Solution

Foresight is building a network of data-driven, scientifically-grounded mental health clinics. Each Foresight location is a fully licensed medical facility, operated under the clinical supervision of a practicing psychiatrist. The experience is intentionally designed to be able to measure everything that may shed light on a patient’s condition and their optimal course of treatment.

The value/disruption of the Foresight solution is best illustrated by charting the patient journey:

1. Building Each Patients Genetic and Non-Genetic Profile. As each patient is unique, Foresight clinics start by creating a baseline profile for members. They collect a full suite of biometrics, genetics (saliva test), a comprehensive symptom profile, brain-chemical balance estimates, mental health history, family history, lifestyle, voice samples and social media analysis. 



The intake process is entirely digital, meaning patients don’t have to deal with voluminous paperwork, which is typical. This information is subsequently stored in Foresight’s Electronic Health Record (“EHR”) system, a data repository on patient profiles and treatment outcomes that over time will enable increasingly personalized and accurate treatment programs. This is covered in further detail below.

2. Real-Time Digital Analysis of the Intake Data/Risk Assessment. Foresight’s algorithms immediately go to work analyzing the intake information and begin to create risk scores for self-harm and violence, estimated diagnoses for disorders like depression, anxiety and ADHD and optimal treatment plans. This is all before the psychiatrist has even finished seeing the patient. In contrast to the status quo, this is all part of a drive towards making diagnosis data-driven and standardized.



Foresight’s suicide risk assessment algorithm collects 50+ features on patients and uses their in-house algorithm to assign a dynamic (changes as new info is entered over time) suicide risk score to a patient.

3. Precision Medicine Software for Optimal Drug Prescription. The different data streams are also aggregated into their proprietary Precision Medicine Software, which is a decision-support software for psychiatrists. The combination of traditional and digital phenotyping that Foresight performs enables the company’s precision medicine software to synthesize highly personalized treatment recommendations, taking into account a vast range of important data points about the patient, their symptoms and their medical history. This leads to optimal treatment and prescription decisions for patients and enables psychiatrists to avoid administering inappropriate drugs that could lead to adverse drug reactions. According to research by the FDA, adverse drug reactions are the 4th-leading cause of death in the US.

With the Precision Medicine Software, the key insight that Foresight has had is that the barrier to personalizing medical treatment is not a knowledge problem. The scientific community already possesses vast information about which drug-drug combinations, drug-genome interactions and drug-disease complications are known to give rise to sub-optimal therapies – or worse still, potentially lethal adverse reactions. Instead, the problem is how to deploy that information so it’s within the convenient reach of physicians at the exact moment they are formulating their prescription.

The Precision Medicine Software is 500k+ rows of manually structured, curated data along with Foresight’s proprietary in-house recommendation engine that they use to select optimal treatments for patients. The software analyses hundreds of features on patients (genetic and non-genetic) to power the platform and give a truly holistic and data-driven treatment decision.

Click here for a video of the Precision Medicine Software in action.

4. Outcome Monitoring is Quantitative and Continuous. Following initial treatment, members are then tracked quantitatively and continuously, using digital outcome monitoring tools, wearables and chatbots. For instance, weekly surveys via Foresight’s patient portal keeps members’ symptom profile constantly up to date. Wearable technology (e.g. Foresight has Fitbit authorization for wearable integration) is used to continuously monitor sleep and exercise. Based on this continuous feedback, a patient’s treatment plan can be modified and optimized.



Conditions like depression and anxiety are especially heterogeneous and comprised and influenced by ~10-12 overlapping symptoms. Foresight’s latest software update allows them to track the frequency and severity of these individual symptoms and gives the treatment provider and the patient a quantitative view of their situation and progress. Foresight’s team note that it is unprecedented for mental health symptoms/outcomes to be granularized this deeply. 



Next, they will be overlaying the treatment on the graph to show, for instance, how “Prozac” is influencing “Restlessness”, or how “iron supplements” impact “Fatigue”. There is deep machine learning potential in this field that the team will be developing further.

This is available in the member portal to enable members to take ownership of their mental healthcare and wellness.

5. Fast Access. Access and affordability are also keystones in Foresight’s offering. They are hiring fast enough to provide openings for patients within 1-2 weeks, although they admit that they are struggling to keep pace with the rate of demand. With their mission to re-invent mental healthcare by leveraging technology, Foresight has attracted a massive flow of employee/provider applicants in the US.

6. Affordable Prices. Foresight also want to make their product as affordable as possible. Many providers shun the need to take insurance payments in the US because of the imbalance between demand and supply: they simply don’t need to make this connection. Foresight, however, is about opening up leading mental healthcare for the wider mid-market, with prices that reflect that. Consequently, they accept insurance and currently receive approx. 40% of their referrals through this channel.

7. Electronic Health Record/Database for Ongoing Treatment Insights. As previously mentioned, Foresight has built its own EHR. This is a data repository that structures and analyses all data from all sites on patient profiles, treatments and outcomes (aggregated and anonymized). Quantifying outcome data in this way allows Foresight to analyze how patients are doing over time on different treatment programs. 

This is in contrast to the status quo of qualitative data collected in paper notes and siloed in each individual provider’s mental healthcare practice. Virtually no other mental health company has their own EHR, which hugely limits their ability to customize treatment, eliminates development opportunities and prevents data collection. With their EHR, Foresight is aiming to create a feedback loop of data about the ongoing health outcomes of patients that will continuously improve the accuracy and efficacy of their treatment programs, including drug selection, drug dosage and the reduction of adverse drug reactions. 

Over time, these insights should provide an even deeper moat around the business and, as a data set, is potentially extremely valuable in its own right. Similar data was collected by Flatiron Health in the field of oncology. They were acquired by Roche for $1.9 billion after just 5 years of operation.



8. Additional Product Features – Coming Soon

a. Precision Medicine Software: the fact that Foresight has spent over 2 years developing their own technology platform to analyze and house all patient data has built a big defensive moat around the business. Foresight’s R&D and engineering team is continuously deploying updates, which is unprecedented for mental healthcare. The platform they have built is flexible and extendible, meaning they can build clinical modules on top of their software. Nutrition, exercise and mindfulness will all have personalized recommendations in Foresight’s Precision Medicine Software by Q2 2020, at the absolute latest. They are most excited about the nutrition module. There are others active in this space, but there are none that have a fully integrated software and clinic model. This allows Foresight to deploy tools directly into the clinical setting much faster than any alternative offering. The goal is to have patient data inputted and personalized guidelines for nutrition, exercise and mindfulness generated with reasoning for the recommendations. In time, they will also consider a neurological module and in-patient programs, but the current focus is on scaling the current mid-market proposition.

This is again all part of building a holistic treatment program that recognizes that mental health can’t be fixed once a month on the sofa. Personalized therapy, precision medicine, improved sleep, custom nutrition and exercise are all valuable components in enabling patients to reach their optimal mental health.

b. Nutrition/Supplement Store (recently launched!): Foresight hired their first nutritionist in September, who will initially be covering all sites via telehealth. Whilst the team builds the nutrition software module, this will enable nutrition counseling from October. This will be a new treatment line that will complement existing clinical offerings. There are multiple circumstances where a provider may wish to refer a patient for a nutrition assessment: where they display a hormone imbalance, chronic fatigue, migraines, insomnia, MTHFR genetic mutations, ADHD, and not least because many medications deplete basic nutrients in the body. Nutrition counseling is not covered by insurance companies, however, so will be offered “out of pocket”. In addition to nutrition counseling fees, nutrition supplements will soon generate revenue through a white- labelled online e-store that is in the process of going live. They have partnered with Designs for Health to launch this.

c. Neuropsychology: Integrating neuropsychology at Foresight was identified as an important need and the company has already hired and launched 2 neuropsychologists who are both fully booked 3 months out, as of today. Neuropsychology is intertwined with mental health and assesses problems in brain functioning related to behavior, emotion and cognition. Many mental health conditions can be traced back to neurocognitive disorders. Hiring neuropsychologists enables Foresight to do the following as they extend this service offering:

i. provide clinical decision support software around when to refer to a neuropsychologist to confirm or rule out diagnoses;

ii. digitise neuropsychological assessments to examine brain functioning and aggregate neurocognitive biomarkers (memory, processing speed, concentration, motor function, etc.) to have a more comprehensive “mental state” picture, before treatment and to measure outcomes; and

iii. once enough data is gathered, train a model to recognize new digital biomarkers based on neurocognitive assessment data, clinician diagnoses, and symptom scores (Mindstrong did this, see Competitive Landscape section below).

d. Voice Analysis: with consent, Foresight has begun to record the audio of its patients. Using standard software, they can characterize features imbedded in speech (prosody, velocity of speech, tone, volume, pauses, and much more). With this data, they are able to train machine learning algorithms to identify biomarkers in speech for conditions such as depression and anxiety. This is very valuable as, along with the labelled diagnosis, clinical survey, risk scores, and other data, Foresight will provide a highly standardized diagnosis process across providers, increasingly removing subjective decision-making from the process. Additionally, voice data at scale could be immensely valuable to players in the smart speaker or smart assistant space who are interested in getting into mood analysis and/or treatment (digital therapist). Some of these players include Google (Home), Amazon (Alexa) and Apple (Siri). The algorithm should have enough voice samples from patients to be sufficiently trained by the end of the year/early 2020. 33% of patients are currently consenting to audio collection in their appointments.

Traction

Foresight launched with 4 offices at their first Berkeley clinic and have consistently underestimated how much space they'd need due to overwhelming demand.



Foresight is now up to 12 room clinics in some of their new LA locations. Importantly, margins are better in larger locations: 6-8 rooms: 20% margin; 8-10: 25% margin; 10-12: 30% margin. They’re all laid out the same and look/feel nearly identical, except for the increasing number of provider rooms.



To provide a sense of how fast Foresight is growing, total revenue since inception is just over $2.5M and Q1 2020 revenue is already at $1M+. The growth shown in these figures demonstrates the deep vein of demand the Foresight team has uncovered. The company is positioned to grow even faster as a result of surging mental health issues attributable to the COVID-19 outbreak. Further, leases have already become significantly cheaper as demand drops dramatically for new commercial spaces. 

The pandemic also served as an accelerant for the company's planned foray into telehealth services, helping speed the learning curve and rollout around launching a virtual offering. They plan to continue growing the virtual clinic business after the shelter-in-place is lifted to complement their physical presence.



The company’s performance metrics for March included 32% MoM revenue growth and a 40% increase in monthly visits, fueled by the surge in demand from the COVID-19 outbreak. This brought Foresight to a $6.7M revenue run rate. 

Foresight currently has eight clinics live in the Bay Area, with an additional three opening in May 2020 and one opening in San Diego, bringing the the company to a total of 18 clinics. The Series A roadmap through to Q4 2021 is to scale this to 50 sites in 7 U.S. markets.

Unit Economics. A typical Foresight clinic becomes breakeven after six months and hits a steady-state revenue of circa $90-$100k per month at month nine to ten. The EBITDA margin is approximately 20% at the steady state. Therefore, at a steady state, a clinic will have an annual revenue close to $1m and an annual contribution of approx. $200k. Establishing larger clinics once demand is proven and less hiring is required through recruiters should see a meaningful improvement in these figures over time. Foresight just received a substantial increase in reimbursement rates from several insurance payers in Q1 bringing the estimated contribution margin to 30-35%.



Foresight maintains a headquarter office that acts as a cost center and includes the R&D team, management and marketing. The annual burn rate of this is estimated to be $1.5m by the end of 2020 and therefore the company is already projecting breakeven by mid Fall with their existing clinics.

Financial Projections. The initial focus is on growth in the California, although by the end of 2020 the team will begin pursuing a national roll-out strategy and project comfortably growing to 200 sites nationwide. At say 200 locations across the US, with each site turning over ~$1m per annum at steady-state, there is therefore a roadmap to scaling to $200m+ in annual revenue, with a consolidated EBITDA contribution of over $35m+. This is just the cashflow portion of the business, with significant additional value also coming from the deep data set they will collect and analyze on patients, treatments and outcomes.

TEAM

Thumb screen shot 2019 12 30 at 10.30.32 am
Matthew Milford
Co-founder

Bioengineering - UC Berkeley

Thumb screen shot 2019 12 30 at 10.31.11 am
Douglas Hapeman
Co-founder

Electrical Engineering and Computer Science - UC Berkeley

Thumb screen shot 2019 12 30 at 10.32.13 am
Jerry Gelbart, MD
Medical Director

20+ years of psychiatry experience and formerly Clinical Adjunct faculty - Stanford University

Thumb screen shot 2019 12 30 at 10.33.13 am
Keely Strong
Director of Operations

Hired from One Medical, a network of primary care clinics (most recently valued at ~$2B with 70+ clinics in 9 cities), where she held the same role and scaled the organization from 4 to 32 sites.

RESOURCES

I Can Help