Data Professionals in Healthcare: Rising Demand for Statisticians & Analysts
The 21st century runs on data. By 2025, it’s estimated that 463 exabytes of data will be created each day (World Economic Forum, April 2019). That’s roughly 212 million DVDs worth of information, for those who remember DVDs. What are we going to do with that? First off, we’re going to give it to quants—quantitative analysts, data analysts, and data scientists who can turn massive pools of data into actionable insights through a blend of industry knowledge and rigorous statistical modeling.
The World Economic Forum’s 2018 report on the future of jobs found data scientist to be the number one emerging career, and the Bureau of Labor Statistics (2019) projects “statistician” to be one of the fastest growing careers of the next ten years. The problem is that there’s just not enough available talent yet. According to Indeed (2018), an employment data aggregator, data scientist job postings jumped by 75 percent over a recent three-year period. No wonder McKinsey & Company reported that the demand for data scientists outstrips demand in every sector across the world. IBM calls this widening gap “The Quant Crunch” and in healthcare, it’s even more pronounced.
Healthcare lags even further behind other major industries in the use of cutting-edge tools. A 2017 survey from Dimensional Insight found that barely more than half of all hospitals have a set strategy for data analysis. But the possible benefits of merging healthcare with data science can lower operational costs, improve patient outcomes, and benefit global public health. To achieve those benefits, however, will require quants in healthcare to navigate the contextual complications that are unique to the industry.
The Challenges to Quants in Healthcare
The Heightened Complexity of Healthcare Decisions
Data issues in healthcare are more complex than they are in a strictly business setting. Healthcare decisions are based on sensitive information, and have to be made quickly in scenarios where there’s little room for error. That means that decision-critical healthcare data needs to be collected constantly and accurately, and shared in real-time. The needs for high-quality technical infrastructure and data professionals are extremely high. Plug-and-play solutions such as IBM’s Watson don’t yet work and carry unacceptable levels of risk.
A further hindrance is the question of privacy. If patients understandably fret about the security of their most sensitive health data, how can data analysts and scientists collect a full picture from which to derive meaningful insight? Tomorrow’s quants in healthcare need to understand not only data, but the contexts in which it’s being applied.
The Unique Characteristics of Healthcare Data
Data integration is a major hurdle for quants in healthcare. Medical data is disaggregated across different entities, many of which house it in different formats which can’t easily translate to one another. Gaps exist in the record. Housing and sharing data is more expensive than it is in other industries.
Designing a universal, updated, and sharable database comes with serious concerns about privacy. But another concern is the veracity of the data that’s collected: analyses and predictions have no value if the data is tainted in some way. Without rigorous standards shared across a wide network of healthcare providers, a master database could cause more problems than it solves. Quants in healthcare have to work with what they have, but also work towards building an interoperable environment that functions for a variety of stakeholders.
The Structure of Healthcare Delivery
One of the more promising aspects of data analytics in healthcare is providing decision support in clinical settings. In theory, a physician could consult a dashboard that balances the physician’s opinion against data insights.
In practice, however, the workflows of healthcare make this ineffective. Clinical decisions often involve several parties and are guided by institutional practices. The result is a physician receives yet one more opinion—this time, from the decision support software—that needs to be integrated into the institutional workflow. Bringing this supposedly helpful tech into the clinic is actually increasing the workload for the physician, and it’s one of the reasons why physicians are resistant to it. Quants in healthcare need to work to smooth the physician’s existing workflows—not disrupt them.
The Varied Motives of Healthcare Stakeholders
Healthcare is a large and crowded industry, with a wide array of stakeholders. Unfortunately, their motives don’t always align and that has major impacts on the flow of data between parties. Insurers hold claims data that can tell a caregiver whether their new patient has recently been to an emergency room. Clinicians hold data that could help insurers adjust a patient’s costs. Patients change clinicians (and insurers) with some frequency. In some cases, the lack of data sharing is intentional. In others, it’s merely a side effect: critical data may go unseen, marooned on some third-party vendor’s software platform.
But everyone is looking to someone else for a solution. Insurers focused on cost are less likely to invest heavily in health-focused data initiatives that yield insights for patients who may take their business to a competitor. Healthcare institutions like hospitals and clinics are unlikely to saddle the costs for such health-focused data initiatives themselves, and a continued lack of data interoperability helps them keep top-performing caregivers in-house.
The vendors of healthcare data software aren’t particularly interested in a standardized format that hurts their market share. And even patients will make data decisions that prioritize privacy over their own health. Quants need to work within this fractured and sometimes schizophrenic data landscape to deliver actionable insights that result in win-win scenarios.
The Future for Quants in Healthcare
While federal policy should focus on the interoperability of health data and incentivizing healthcare professionals (insurers and providers) to make collaborative use of it, quants can look toward the future of healthcare. Their decision support systems should work with clinician workflows to make them more efficient instead of less. Medical imaging, drug discovery, genetic research, and preventative medicine can make great strides when data analysis is tailored to the populations and infrastructures that they serve.
The future of healthcare depends on an increased supply of capable and visionary quants. Their work to advance patient care and improve organizational efficiencies will represent the same windfall quants do to other businesses: an increased ROI and a more desirable final product. In this case, however, the final product is improved global public health.