How to tap into ‘small’ data to achieve bigger results

A focus on small data—one number for one patient—could revolutionize healthcare; however, clinicians and caregivers don’t always have access to small data when they need it.


Data analytics gets a lot of hype, but the analysis of big data generally only points us to larger patterns and trends. So while it has implications for our broader understanding of preventive healthcare, it doesn’t directly impact an individual patient’s personal health. On the other hand, a focus on small data—one number for one patient—could revolutionize healthcare.

There’s just one problem: Clinicians and caregivers don’t always have access to small data when they need it. Luckily, that’s changing fast.

Advances in real-time biometric technology are enabling providers to administer remote patient monitoring (RPM) through nursing staff and extenders rather than through perhaps the more costly confines of an office, clinic or hospital.

According to a comprehensive market report that analyzes current RPM market trends, the RPM market was valued at almost $16 billion in 2017, a figure that’s projected to double by the end of 2023. This isn’t a coincidence. The reality is that RPM solutions enable truly proactive care—the kind that wasn’t possible even a few years ago.

Patients treated remotely often receive a more holistic, consumer-centered experience. At the same time, providers’ time is freed up for more complex, time-intensive cases. Data from connected devices like scales, blood pressure cuffs, glucometers and other technology that’s part of the Internet of Medical Things can reveal a change in compliance or condition in real time, enabling clinicians to engage at the exact moment of need and drive patients to the right next step in care.

For RPM and other data-driven approaches to work, the data that fuels them must be reliable and accessible. At a minimum, care providers and payers need access to daily touchpoints with the client.

This may include anything that serves as a proxy for a desired compliance measure—for example, imagine compliance to SSRIs (variable) as measured by glucose test-strip utilization (proxy). A daily patient check-in not only provides longitudinal tracking for the care team, but also a 24-hour compliance check that immediately warns when compliance is waning or the patient’s condition has deteriorated.

Small data that provides clinicians with actionable insights at the patient level can be harnessed to provide better care in multiple ways. Research from Cornell University found that small data generated by walking and location patterns, communication habits and online behavior can yield valuable insights to providers that enable more personalized treatment.

One Cornell researcher developed an app to help people manage chronic pain—through the collection of small data on patient movement and location—and applied what she learned to develop customized suggestions in real time. Similarly, this app could be useful for detecting the circumstances in which a condition like rheumatoid arthritis typically causes pain.

Still, it needs a better mechanism for filtering useless data and deriving meaning from the incessant background noise. Pinpointing the data that matters is still a challenge for most AI-powered machines, and it’s one that we must overcome.

Nurses and physician assistants would rather not engage multiple disparate portals while caring for one patient—and physicians simply won’t. All relevant patient data must be consolidated and easily accessible in one place as part of a streamlined daily workflow for all providers caring for a patient.

As data is consolidated, tools and technology solutions should be, too. Interoperability—the ability for different programs and devices to work with the same datasets in a coordinated manner—is becoming a requirement as startups, health systems and providers attempt to provide more personalized care to a growing number of people.

Thanks to the new Interoperability and Patient Access Proposed Rule (created through the Centers for Medicare and Medicaid Services), there are now many opportunities to “make patient data more useful and transferable through open, secure, standardized and machine-readable formats,” according to the CMS.

Solutions that don’t integrate with others aren’t going to be useful for long. In the future, healthcare’s highest value commodity will be a patient’s clinical chart—as long as that chart includes accurate, relevant, real-time data.

Real-time biometrics is still nascent as a medical discipline, but it’s already making an impact on the way we provide care. What we’ve learned so far is that when patients are engaged in programming (and have access to resources so they can get answers to their health-related questions), they visit their doctors more often, stay out of the ER and avoid inpatient stays.

Readily accessible patient data has the potential to become the norm as the adoption of electronic health records expands. As this occurs and clinicians access and analyze data from disparate sources in one place, they’ll be able to not only provide care in the present, but also help educate and treat for the future.

More for you

Loading data for hdm_tax_topic #better-outcomes...