3 things all health tech entrepreneurs need to know
The Great Recession sparked the health tech revolution
Until very recently, paper was the primary storage medium for our nation’s healthcare system. In 2009 the American Reinvestment and Recovery Act (ARRA) earmarked $30B in federal funds to incentivize health care providers to adopt electronic health records (EHRs) in an effort to modernize our nation’s healthcare system. By most measures, this program was wildly successful: in 2008, only 9.4% of health care providers used EHRs, by 2015 EHR adoption jumped to 83.8%. Today, 96% of community hospitals rely on EHR technology.
From an insider’s perspective, EHR adoption felt like it progressed at breakneck speed. In 2010 I was a clinical intern at Santa Clara Valley Medical Center (SCVMC) where I spent hours handwriting patient notes and relied on a pneumatic tube system to send orders from one wing of the hospital to another. In 2011 — then an anesthesia resident at Stanford Health Care (SHC) — I used an EHR to place orders for patients in the ICU but spent the bulk of my time in the operating room keeping meticulous records of my patients’ heart rate, blood pressure, and respiratory rate using pen and paper. By the time I finished my training in 2014, the housestaff at SCVMC and SHC were relying almost exclusively on EHRs to fulfill their clinical duties.
EHR transition teams were lead by “clinical champions” and “super users” — physicians and nurses with backgrounds in clinical informatics and/or EHR-specific training and certification. Once EHRs were in the picture, these hybrid clinical-technical leaders became increasingly influential figures within their respective health care organizations, and many went on to serve as Chief Medical Information Officers.
Today, clinicians are more involved than ever in all matters health IT. CMIOs are business leaders that play a key role in decisions involving software licensing and data governance. This means that companies whose software generates, processes, or consumes clinical data need to adopt sales, marketing, and engineering strategy that addresses the broad set of concerns raised by clinical leaders. Doing this effectively requires a deep understanding of how policy translates not only to health IT needs, but also into practice. It also underscores the significance of the ARRA’s EHR incentive as a quintessential case study for all health tech entrepreneurs.
Meaningful Use is a clinician’s enemy, and an entrepreneur’s best friend
In order to protect the government’s investment in health IT, lawmakers require health care providers to use federal incentives toward EHR software that “offers the necessary technological capability, functionality, and security to help them meet Meaningful Use criteria.” EHR software that meet these criteria is known as certified EHR technology (CEHRT). The Office of the National Coordinator for Health IT (ONC) is the entity that is responsible for managing the EHR incentive and CEHRT programs on behalf of federal lawmakers. MU criteria are the knobs and levers of authority granted to the ONC to achieve their legislative mandate.
Since the EHR program’s inception, MU requirements have drawn criticism from health care providers. They argue that ongoing changes make compliance into a moving target. They also criticize requirements for being exceedingly specific, painfully vague, or costly to implement. The following examples illustrate these points:
- “For 2016, a [health care provider] must send a secure message using their [EHR’s] electronic messaging function to at least one patient (or the patient-authorized representative) the [health care provider] sees during the EHR reporting period, or must respond to a secure message that the patient (or the patient-authorized representative) has sent during the EHR reporting period.”
- “The [health care provider] is in active engagement with a public health agency to submit immunization data”
In both cases, the first challenge to achieving compliance is ensuring that one’s existing IT stack is able to capture and transmit the necessary data. The second challenge is training clinical staff to use new software and/or features — a process that often involves altering clinical workflows to accommodate the need for proper documentation. Given that doctors already spend twice as long entering data into EHRs as they do seeing patients, it’s no wonder that health care providers have grown weary of ever-evolving MU criteria. But it is precisely because MU requirements create pain points for clinicians that they are an entrepreneur’s best friend.
Simply put: Meaningful Use is a category creator. Nearly every doctor in America has to comply with MU, this is a fact that all-but-guarantees a large market for a broad spectrum of health tech products. The challenge for entrepreneurs is therefore to deliver software solutions that help clinicians achieve MU compliance without a) disrupting clinical workflows or b) creating administrative overhead.
I should note that some in the industry speculate that Meaningful Use is on its way out while others argue that it is here to stay. The reality is likely somewhere in between. Even if the MU program ceases to exist in name, it is likely to persist in practice. There are simply too many stakeholders that benefit from the role that MU plays in driving health IT policy, not the least of which is Congress. In the last two years alone, Congress has leveraged the information gleaned through the MU program to craft two major health care laws: MACRA (Apr 2015) and the 21st Century Cures Act (Dec 2016). Both laws were passed with overwhelming bipartisan support. So, if recent history is any indication, MU (or a functionally-similar program by a different name) will continue to play a key role in health IT policy and category creation well into the future.
Electronic health records are the solution to (and cause of) many problems
The transition to EHRs was not easy. Data entry proved to be a particularly stressful pain point as clinicians struggled to write patient notes and place orders. EHR vendors helped to ease the transition by adding features that allowed clinicians to create note templates and to bundle related orders into “order sets.” These features made data entry easier than ever, but they also gave rise to a wave of problems related to information retrieval.
While note templates streamlined clinical documentation, they did so at the expense of clarity in communication. If you were to review the daily notes written by a single doctor for a given hospital patient, you’d discover that the notes differed by no more than a few sentences here and there. Moreover, if you reviewed all the notes written by a single doctor for all of their patients, you’d find that 70–90% of the content is nearly identical. The problem with this practice is that it drowns salient points in a sea of redundant content. This makes it increasingly difficult for anyone other than the note writer to answer simple questions such as:
- Why was this patient seen by this clinician? Was it at the request of another clinician? If so, what was the specific clinical question that was beyond the scope of training/experience of the referring doctor?
- What did this clinician do for this patient on this occasion? How much of the copy-and-paste content relates to this interaction vs. yesterday’s vs. last week’s vs. last month’s?
- Is the author of the note the clinician best suited to oversee this patient’s care from this point forward? Are they actively involved in the patient’s care today, or has primary responsibility for this patient been transferred to another clinician since the note was written?
These are examples of real-world information gaps that reflect a systemic problem rooted in clinical data systems that are poorly interoperable. A report commissioned by the Institute of Medicine concluded that “30 percent of health spending in 2009 — roughly $750 billion — was wasted on unnecessary services, excessive administrative costs, fraud, and other problems”. Related studies pinpoint the source of waste to technical barriers limiting clinicians’ access to a patient’s complete medical history. The importance of these findings cannot be understated: not only are information gaps exposing patients to unnecessary harm, they contribute substantially to wasteful health care spending.
Along the same line, order sets solve one problem while creating another. In theory, order sets are a convenient way of ensuring that clinicians adhere to well-established guidelines when managing the care of patients with conditions such as shock, chest pain, and stroke. In practice, the data that speaks to the intent of each order is separated from data about the order itself. Lacking the appropriate clinical context, it’s difficult if not impossible to safely interpret study results, and to measure clinical outcomes in a meaningful way.
A common example of the problem created by order sets involves the use of Heparin — a blood thinning medication — in conjunction with a lab measurement known as “PTT”, which is used to monitor whether the patient is receiving the appropriate dose of the drug. Heparin is indicated in patients suffering from a heart attack, certain types of stroke, and other conditions. It is critically important to maintain the patient’s PTT within a desired range, but the target range varies according to the drug’s indication and the patient’s condition. A PTT value below the desired range indicates that the patient is at risk for ongoing damage to vital organs and/or death. When PTT values are above the safe threshold, patients are at increased risk for developing internal bleeding. Although leading hospital EHRs ship with pre-built order sets to help manage Heparin dosing, they make it incredibly difficult to understand why a patient is receiving heparin and/or undergoing PTT testing. This is problematic when multiple doctors are involved in the care of the same patient (a very common scenario) because each doctor is making clinical decisions based on their interpretation of the patient’s condition.
Barriers to information retrieval caused by EHRs affect other products in the health tech economy. For example, digital and consumer health companies will need to find ways to surface their data through existing EHR systems. Meanwhile, companies that are building enterprise software — especially those involving precision medicine, population health, and care coordination — will need to engineer their products to be compatible with the dizzying matrix of clinical software that comprise health IT stacks.
Companies that fail to mitigate the impact of clinical information gaps risk losing their strategic position in the market. I expand on this idea in another post: The Health Tech Economy.
