Healthcare IT Thesis
— Sid Grover and Alex Kolicich
“America’s healthcare system is in crisis precisely because we systematically neglect wellness and prevention.” — Tom Harkin
The healthcare industry is one of the pillars of the American economy. National healthcare expenditures account for nearly 20% of GDP, or three trillion dollars. Furthermore, health care costs are growing around 5% per year, which significantly exceeds aggregate GDP growth. The sheer volume of money moving through the system cannot be overstated.
Yet, conservative estimates of annual slack and waste in the healthcare system exceed $900B dollars, or one third of healthcare expenditures. The United States healthcare system ranks last amongst the 11 most industrialized countries on various scores measuring the quality of care delivered. Readmission rates are significantly higher in the US than in other countries for standard episodes of care such as strokes and heart attacks. Even when medication costs are normalized, managing chronic diseases is more expensive in the US by up to an order of magnitude. American patients also see significantly higher rates of preventable comorbidities for diseases such as diabetes. On balance, Americans spend more on healthcare than any other high-income nation, yet they see comparatively poor health outcomes and significant inefficiencies in care delivery.
The problem is that America’s healthcare ecosystem suffers from a fundamental misalignment of incentives between health insurance companies (‘payers’), providers, and patients which systematically inhibits the delivery of cheap, effective care.
Until recently, the overwhelming majority of dollars flowing through the healthcare system were tied to fee-for-service (FFS) payment models. Under this paradigm, care providers are reimbursed for each service they provide, for instance a surgical procedure, the administration of a drug, a blood panel, or a pathology test. As a result, they are perversely motivated to increase care volume rather than care quality. Over the long run, they are also disincentivized to identify and deliver preemptive treatment for preventable conditions or to adopt new technology, so many industry stakeholders continue to deliver suboptimal, poorly coordinated care with little threat of disruption.
History has shown that technology needs to be tied to payor and provider incentives in order to achieve widespread acceptance in the healthcare ecosystem. For example, although the current generation of EMR systems were developed in the late 1990s, adoption growth was relatively modest until the HITECH Act of 2009, which established financial incentives in the form of explicit reimbursements for the adoption and meaningful use of EHR systems. Since then, over 135K care providers have filed for EHR incentive payments, and the percentage of physicians who adopted EHRs in their practice doubled between 2008 and 2011. In general, changing incentive structures create technological voids which allow us to predict where the health IT puck is headed next.
In recent years, the American healthcare system has slowly begun to uproot its incentive structure and transition to a new framework for reimbursement known as value-based care. Value-based care consists of a family of novel payment models that tie payments to the efficacy and quality of care delivered, instead of reimbursing care providers for the number of procedures performed. Although diverse in structure, all value-based payment models are united by their attempt to align the financial incentives of providers with better outcomes for their patients. Examples of value-based models include:
- Bundled payments (aka episodic payments) programs, which change the atomic unit of healthcare from a procedure or operation to an episode of care. In a bundled payments program, care providers are reimbursed a single lump sum per care episode (capitation). For example, payers negotiate a bundled payment to be paid to a cardiologist for treating a heart attack, regardless of the specific circumstances around the operation or follow up conditions and readmissions. If treatment costs less than the capitation, the provider gets to pocket the difference as profit. However, if care is not delivered efficiently and the total bill exceeds the capitation, the provider must pay the difference out of pocket.
- Risk-sharing agreements under which provider networks assume a fraction of the risk associated with insuring a member population and paying for necessary treatments. This incentivizes care providers to take a longer term, preventative approach to patient health, as readmissions, poorly managed chronic conditions, and comorbidities will eat directly into their bottom line.
- Integrated care delivery networks such as ACOs, which bundle reimbursements for primary care providers (PCPs) and specialty care providers together, thereby incentivizing them to deliver more integrated, coordinated care.
- Outcomes-based reimbursement models under which providers are only reimbursed when they hit well-defined outcomes benchmarks for patient recovery, such as hemoglobin A1c levels in diabetic patients. Conversely, providers are penalized for readmissions and lapses in patient health.
The financial incentives to adopt these payment models are incredibly compelling. For example, under the Merit-based Incentive Payment System (MIPS) program, care providers treating Centers for Medicare and Medicaid Services (CMS) patients are eligible to receive an extra 9% on top of standard reimbursement for hitting defined outcomes benchmarks.
Over the past few years, value-based payment models have grown to account for a significant, secularly increasing fraction of healthcare spend. For instance, 58% of Anthem’s reimbursements in 2017 were to value-based models. Anthem alone has shared risk agreements with over 64,000 providers operating in ACO structures. Similarly, health insurance giant Aetna is on track to pay out 75% of its reimbursement dollars through value-based payment models by 2020. The CMS is pushing to have 50% of Medicare dollars flowing to value-based payment models by the end of 2018.
The Next Generation of Health IT Infrastructure
The advent of value-based care is immensely exciting for technology investors because legacy healthcare IT infrastructure is completely molded to fee-for-service care delivery. Fee-for-service care involves a relatively linear workflow: enroll a patient, make a diagnosis, treat them, and bill the patient and payer for each incidental expense. Consequently, current technology only addresses the lowest level of the care delivery stack — static snapshots of post-admission patient health and disparate treatment logs found in EHR/EMR and revenue cycle management systems.
The manner in which value-based payment models are attempting to align the interests of payers and providers with the long-term wellness of their patients is completely changing care delivery and, by extension, the health IT technology which they are incentivized to adopt. In a value-based world, the patient lifecycle becomes much more dynamic, resting on four primary pillars:
- Population health management: Continually monitoring patient health to identify high-risk patients and deliver care preemptively
- Care coordination: synchronizing the treatments and procedures performed by stakeholders such as PCPs and various specialty care providers
- Tech-enabled care delivery: Leveraging technology to bring an unprecedented level of scientific precision to clinical workflows in order to achieve better outcomes more efficiently
- Outcomes Measurement: Continually reevaluating patient progress against established outcomes benchmarks, and adapting procedures until satisfactory progress has been realized
Unfortunately, the technology powering these four workflows is either nascent or completely absent. Value-based care today is still being administered and serviced using legacy technology, resulting in suboptimal decision making and care. There are immense opportunities for today’s technology companies to build the next generation of infrastructure that will support the industry going forward. We discuss these opportunities in detail below.
Population Health Management
Value-based care refocuses the lens through which providers view the lifecycle of patient engagement. While fee-for-service models engender a reactive attitude towards disease treatment by rewarding care providers for temporary solutions, value-based care models such as risk sharing agreements and outcomes-based reimbursement financially align providers with the long-term health of their patients, and encourage preventative medicine. This proactive approach to patient health is known as population health management, and it is born out of the need to identify high-risk patients and gaps in care delivery before they turn into larger financial liabilities.
In concrete terms, providers now have compelling financial incentives to harness data platforms that will give them insights into their patient population, continually rank and stratify individuals at risk, and personalize treatment for individuals who need it.
Population health management systems for various specialties and conditions comprise some of the most compelling technology platform opportunities that we see in the healthcare space. These platforms must be built on the back of significant data infrastructure and analytical machinery well beyond the scope of what even the most sophisticated EMR and EHR systems offer today. A comprehensive population health management system must integrate and normalize disparate data systems from hospitals, PCPs, pharmacies and labs, in addition to claims and real-time clinical data. The dominant EMR and EHR systems today are clunky, unscalable, and only capture a sliver of patient data.
On top of this data layer, population health management systems must offer scalable and customizable analytics that can handle a complex feedback loop with constantly updating patient data.
Depending on the relevant specialty care and patient condition, such systems should flag high-risk patients for whom reactive treatment will likely be expensive and smoothly coordinate preventative care measures. As a result, we expect that population health management systems will serve as a valuable middleware layer between provider systems and patient facing technology, such as messaging tools and medication adherence trackers. At an infrastructural level, this means that significant value will accrue to companies building technology for interoperability and data warehousing.
Value-based care will also force providers to bring technology to one of the most historically neglected elements of the care delivery process — care coordination. Primary care physicians are the gatekeepers of the healthcare system in virtue of their immense leverage over the patient relationship. Until now, primary care physicians had little incentive to optimize the manner by which their patients were connected to specialty care. Novel value-based payment models are changing this.
One such example is the Accountable Care Organization (ACO) model, under which doctors and hospitals come together and form risk-sharing care delivery networks to improve the quality of patient care. ACOs make providers jointly accountable for patient health and encourage more care coordination between various stakeholders. They also include strong financial incentives to avoid redundant testing and unnecessary procedures. In the spirit of value-based care, eligible ACOs are incentivized to assume some of the balance sheet risk of insuring their member populations, further aligning their incentives with their patients’.
Another example of a value-based program that incentivizes better care coordination is Medicare’s aptly named coordinated care management model (CoCM). It is estimated that 60% of Americans with a diagnosable mental health condition go untreated because it’s difficult for their PCPs to orchestrate mental health treatment across a siloed network of mental health care providers. CoCM directly addresses this problem by reimbursing PCPs for connecting their patients to mental health therapists, supporting the mental health care delivery process, and reporting patient progress relative to benchmarks.
At a technological level, payment models such as ACOs and CoCM are united by a common technology platform challenge. In a world in which a single hospital can be running as many as ten different EHR deployments, it’s important to have technological plumbing that helps stakeholders exchange patient data, synchronize treatments and procedures, and better manage the transfer of patient custody and responsibility.
Today referrals are far from seamless — they are still run on telephones and fax machines. Patient data transfer involves paper-based transactions that are fraught with error. It is estimated that PCPs spend as many as 10 hours per week preparing prior authorizations for referrals, and that as many as 40% of initiated prior authorizations are done incorrectly or abandoned. As payment models that reward integrated care force industry participants to coordinate care delivery more effectively, the need for better middleware technology to handle care transitions, data exchange, and integrated decision support will become all the more pressing.
Tech-Enabled Care Delivery
Another family of compelling business opportunities is centered around clinical workflow software which will overturn conventional modes of practice by encouraging novel clinical workflows, more optimal use of patient data, and deeper patient engagement. These three facets of care delivery present the opportunity to build a new wave of enterprise software infrastructure that captures and owns clinical decision making workflows in their entirety, collecting valuable data on treatment procedures and leveraging it to provide decision support. We call this new paradigm tech-enabled care delivery.
Such technology will allow health systems to scientifically analyze variations in procedures along parameters ranging from personnel to supplies. Furthermore, data analytics platforms built on top population health management software will serve as the backbone for precision medicine and personalized care delivery. This data will eventually be used to scientifically determine optimal treatment and procedures, where before medical decisions relied upon the fallible intuitions of doctors.
Value-based payment models are explicitly incentivizing tech-enabled care delivery in various ways. For example, bundled payments programs are compelling care providers to adopt software for clinical workflows by forcing them to iterate on their clinical pathways in order to realize procedural efficiencies and lower their costs of treatment.
Another example is CMS’s Oncology Care Model (OCM), a novel payment model for cancer treatment. In broad terms, OCM is encouraging a more personalized, agile approach to cancer treatment by mandating that oncologists implement real-time performance measures tied to patient well being, and document feedback loops between these measures and further clinical decision making. In addition, OCM relies on patient-reported outcomes to engage patients and uses reported symptoms to guide preventative treatment. Each of these elements requires robust software solutions that provide seamless workflows and improve the quality and scientific precision of care rendered.
Companies that power tech-enabled care delivery will become increasingly lucrative as the value-based care landscape matures because they can establish brands around the provision of a certain standard of care. Branded technology companies can become origination channels for specialty care providers working in partnership with PCPs and ACOs — this is a strategy that Flatiron Health is taking with OCM, for example. In a world in which reimbursements are tied to outcomes and better care delivery, the technology companies setting those standards and controlling the infrastructure will wield significant clout.
Outcomes Measurement and Reporting
Ultimately, the backbone of the value-based care paradigm is the notion that providers should be reimbursed for outcomes realized and not merely for services rendered. To this end, a central pillar of the modern health IT stack will be technology enabling precise outcomes measurement and reporting.
We expect that outcomes measurement software companies for various care verticals will occupy a point of immense leverage, as they will serve as the source of truth on patient health post treatment, thereby driving reimbursement under value-based payment models. Furthermore, these companies alone will control the data necessary to set outcomes benchmarks in the first place. Finally, outcomes data will inform several ancillary workflows, such as performance assessment and procurement processes.
Outcomes measurement and reporting platforms will also become particularly important in the context of real-world evidence (RWE) — the growing school of thought that the efficacy of clinical protocols, drugs, and medical devices should be tested and validated in real-world clinical contexts rather than artificially structured clinical trials. RWE studies are a massive area of spending for pharmaceutical companies (who rely on them to assess comparative efficacy with respect to competitor drugs and develop marketing strategies) and payers (who are financially aligned with promoting the least expensive and most effective drugs). As providers assume patient insurance risk under value-based payment models, their need to gather, collect, and analyze RWE data will grow.
RWE studies have been incredibly expensive to run because there is little software or technology which measures and reports patient outcomes under fee-for-service models. Value-based models, on the other hand, require such technology by their very nature, and this infrastructure is likely to capture a hefty chunk of the value associated with RWE data.
In summary, the American healthcare system is in the early stages of the most revolutionary change in its history — a shift to value-based care. Early value-based payment models in the US such as OCM, bundled payments, and risk-sharing agreements, have achieved substantial success and many have received widespread acclaim. The number of dollars tied to value-based structures is on an accelerating upward trend.
We believe we are in the early stages of a new wave of health IT enterprise software companies building the next-generation infrastructure that is likely to power the ecosystem for decades to come. The value that such companies will capture by virtue of their stickiness, deep integration with clinical workflows, and proprietary data relevant to decision making makes them a compelling opportunity for entrepreneurs and investors alike. Most importantly, such technology will play an important role in reshaping a new paradigm for healthcare in which patients pay less for a higher quality of care.
 Kocher, Robert. “How to Make Our Crazy, Expensive, Amazing, and Uneven Health Care System Better Faster.” New England Journal of Medicine “Catalyst”. October 24, 2016.
 Average Annual Percent Growth in Health Care Expenditures per Capita by State of Residence.” Kaiser Family Foundation, 2014.