By Greg Olsen, Ph.D. and Jessica Hinman, Ph.D.
The executive summary from the 2019 OECD (Organisation for Economic Co-operation and Development) report, "Health in the 21st Century: Putting Data to Work for Stronger Health Systems," captures a significant issue in healthcare:
"Health lags far behind other sectors in harnessing the potential of data and digital technology, missing the opportunity to save a significant number of lives and billions of dollars.
A digital transformation is urgently needed and long overdue at a time of increasing pressure on health systems and budgets. This means ensuring access to the right information by the right people at the right time. The results will be safer, better and more efficient health systems and healthier populations." (1)
Digital transformation in healthcare is happening, but it is slow - slow compared to expectations and slow compared to what has transpired in other industries. The recent wave of investment in 'Digital Health' startup companies is certainly helping to accelerate this transformation, but some significant organizational and cultural roadblocks must be overcome to enable substantial change. This post focuses on the opportunity for existing healthcare providers and insurers to adopt more data-driven approaches to patient care, and looks to historical lessons from the digital commerce transformation that began more than 20 years ago.
The rise of digital commerce (starting in the late 1990s) led to fundamental changes in how retailers of all sorts use data to serve customers better and to operate more efficiently. Today, vendors (brick and mortar, online, and hybrid) aggressively and continuously capture any data about customers they can find. They then mine and analyze that data to improve their product delivery and operations.
We're unsettled by how far this has gone in some cases (e.g., tracking which online sites or physical locations we're visiting, how fast we drive, or impending life events like job changes or pregnancies, constant streams of targeted advertisements). In general, however, this data-driven approach is now an essential tool to help vendors know:
This transformation was driven by online vendors who used the insights they extracted from data to gain a competitive advantage. Ultimately, all but the smallest vendors and supply chain partners made the necessary investments in infrastructure, tools, and skills to implement more data-driven continuous improvement approaches.
Our ability to instrument people, collect health-relevant information about them, and analyze that data to guide health improvement has never been better. Health-related sensor technology has become massively more accessible and ubiquitous, e.g.:
Similarly, software tools and infrastructure used for advanced data analysis have gotten simultaneously more powerful, easier to use, and lower cost.
There is enormous potential for healthcare improvements through large-scale regular patient population assessment. The reality, however, is that outside of focused research studies, nobody is:
Most healthcare providers' interactions with patients are reactive, generally infrequent, and irregular. The information systems used by providers were designed for scheduling, billing, and medical record keeping. These systems and the generally unstructured data they collect are poorly suited for longitudinal or population-focused data analysis.
For the average person, Amazon, Google, Apple, and their credit card provider have a much better understanding of their day-to-day health than their health provider or insurer.
There are, however, signs of change. The combination of greater consolidation and centralization in the industry, competition from new digital health companies, and more information & tech-savvy patients and employers is driving more data-driven approaches. The following section describes some critical challenges that need to be addressed to accelerate the pace of change and adoption.
Healthcare faces some unique challenges that conspire to make the widespread adoption of data-driven approaches difficult, including:
The provision of healthcare to patients involves much more complex relationships than the typical merchant-to-consumer relationship in commerce. While all parties involved (i.e., patients, providers, insurers, and regulators) desire an optimal understanding of patient health, the structure of these relationships subverts paths to improvement.
Patients continue to be more data-driven, with greater access to data from wearable devices and other resources. Still, they lack access to population data, and the expertise necessary to develop deeper insights than the device companies provide (e.g., Apple, Fitbit, Oura, Whoop, etc.).
Providers focus on direct interactions between individual clinicians and individual patients for which they are reimbursed. Even in large provider networks, there is often significant independence among individual providers. Centralization has focused first on finance, billing, and medical record keeping. The incentives, priorities, and cultures haven't aligned so that the providers are willing to invest in more systematic data collection and analysis for their patient populations.
Insurers have a clear incentive to foster improved health at a population level. They, however, have more limited direct access to patients and clinical expertise than providers. Despite the limitations, data-driven approaches in the form of wellness programs where data is collected through patient surveys or incentivized tracking of wearable data are becoming more widespread. To tap the real potential of this type of data analysis, provider access and expertise need to be involved.
There is a long history of data collection and analysis related to human health - in research studies in laboratories of universities, pharmaceutical companies, and medical device companies. These research studies typically include a high level of rigor, including randomized controlled trials and independent reviews. As a result, the path from research study to operational use for a new diagnosis, treatment, medication, or medical device by a provider may take many years. That path is outside the control of that provider.
The culture of separation between the research lab and the operational world of the provider runs deep. Many people in healthcare view all data analysis through the lens of peer-reviewed clinical research studies and randomized controlled trials (RCT).
This separation and approach are appropriate and necessary for many things like the development of new medications or innovative new procedures. There are countless opportunities for valuable medical insights, however, where a much more direct approach is more appropriate. For example, providers and insurers can use their ability to collect and analyze data about patients to:
Data collection now is the lifeblood of future understanding. We simply can't wait for the serendipitous research project or government mandates to build the datasets and knowledge we will need. Providers and insurers must proactively and innovatively collect and analyze the data they can access directly through their operations to better understand their patients' health needs.
To illustrate, consider an alternate world where an eCommerce vendor's path to innovation had to go through external research studies. How would one construct a study to determine which products are best to offer a particular customer at a particular moment?
Clearly, the "wait for external research study" strategy would have been unable to deliver the results that many vendors continuously measuring and analyzing their customer behavior at scale were able to provide.
Almost every industry that encountered the information age started with data collection and analysis concentrated in research laboratory settings. Though it has occurred at different rates, data analysis capabilities and associated technologies have made their way outward toward the consumer. Whether it is manufacturing, transportation, retail, telecommunications, or nearly anything you could name, companies have built the infrastructure and developed the skills needed to use data intelligence technology to do what they do better.
Healthcare is not an exception, but for the reasons mentioned here and others, it has moved at a much slower pace.
Despite the challenges, an acceleration of digital transformation is happening in healthcare. Several factors are contributing to this change:
Consolidation: Centralizing capital and operational resources in provider organizations enables better investment in information technologies and the benefits of scale.
EHR (Electronic Health Record) infrastructure maturation: For the past several years, the healthcare industry has been distracted by the buildout of basic EHR infrastructure and HIPAA compliance. As this capability matures, the capacity for future digital transformation will improve.
Tele-medicine growth: The rapid growth of telemedicine during the pandemic had the side effect of bringing down many cultural and psychological barriers to the integration of technology with healthcare. The limitations of remote interaction increase the need for instrumentation and data collection.
Digital health investment: Large venture capital investments into digital-first companies that seek to use data and network effects to advantage are providing a disruptive competitive force to existing healthcare practices and business models.
Wellness programs & remote monitoring: Fueled by employers and insurers seeking to reduce costs, various forms of "wellness programs" have increased in recent years. These programs have focused on proactive and continuous efforts to assess and maintain individual health through exercise and scalable screening. Over time these types of efforts are becoming more integrated with healthcare delivery through mechanisms like remote physical and therapeutic monitoring
Consumer health tech proliferation: Healthcare consumers increasingly expect and demand digital transformation. Higher and higher percentages of consumers possess smartphones, watches, rings, hearing aids, and other devices and use them to better understand their health through available applications.
Venture investors may have hoped for a digital health wave that resembles the digital waves of the past where startup companies emerge as industry-disrupting unicorns ala Amazon, eBay, Uber, or AirBnB. Healthcare, however, has proved more challenging to disrupt in this manner - with many of the most hyped, highest-valuation digital health provider startups needing to retrench.
The delivery of healthcare to people is significantly more complex than the delivery of packages.
The deep domain expertise and established relationships with patients embedded in existing networks of providers are of central importance. Patients will rely on providers to:
Likewise, the complex relationships between providers and payers cannot be easily disintermediated or replaced. Digital transformation will be more about evolving the digital capabilities of existing providers and payers than replacing them with new entrants.
The consensus-driven, regulated clinical research path will continue to be an essential pipeline for new treatments, devices, and medications. Digital transformation, however, will be driven by independent actions, experimentation, and competition among all market participants (i.e., patients, providers, payers, and tech vendors). How well participants use data and data analysis to improve patient health and healthcare delivery will become an important source of differentiation and will change the healthcare delivery landscape.
OECD (2019), Health in the 21st Century: Putting Data to Work for Stronger Health Systems, OECD Health Policy Studies, OECD Publishing, Paris, https://doi.org/10.1787/e3b23f8e-en.