Not all data are created equal
My co-founder Marquesa Finch and I hosted Protecting Health at the Intersection of Tech and Data at McKesson, a panel event that explored how new data sources, data science and improved analytics in tech can leverage product design and development to advance public health innovation.
Our featured speakers included:
- Sue Grinnell, Director of Business Strategy and Technology at the Public Health Institute’s Population Health Innovation Lab
- David Vivero, CEO at Amino
- Sai Moturu, Head of Data Science, Ginger.io
This marks our third public health and tech event and also marks the official launch of P2Health, an initiative that supports innovation to foster the protection of population health and promotion of disease prevention.
Overview
This event recap below captures some of the key themes of our conversation as well as future opportunities that leverage data for better population health.
Data has been the subject of nearly every health care and technology conversation but rarely with an eye toward how different stakeholders in the health care system use and view data. The following questions were essential for framing our conversation about data, health and technology:
- What application are you trying to build?
- What questions will available data answer?
- How will you close the feedback loop?
- What data are you missing and why?
Challenges
- Outdated or restrictive methods of data collection. The public health system relies on costly but time-tested phone surveys. These surveys administered via landlines, time consuming, and generally are not representative of the underlying population due to the data collection methods. Additionally, the lengthy process makes real-time feedback and response is impossible.
- Data xenophobia. Greater collaboration has been slow moving. Where it is occurring, friction has emerged because stakeholders speak different languages, use different data to define problems, and tend to be overly protective of sharing data, impeding the process of addressing complex problems. Additionally, HIPAA still frightens people into inaction.
- The healthcare/community divide. It’s a well-known and documented fact that health care (ie. clinical care) drives roughly 20% of what makes us healthy while social determinants of health (ie. our behaviors and environment) drive 80% of health outcomes. However when you take a look at overall health care spending, these percentages flip. Closing the healthcare/community divide would be especially impactful for underserved populations. Access to Medicaid data is extremely difficult to access and most EHR companies do not collect data on social determinants such as housing status that could affect care delivery. Even if this data were accessible, most providers lack the knowledge of how to refer their patients to the proper services
Opportunities
- The paradox of data drought and data flood. Health data collection is a space ripe for innovation. There is a deep unmet need for more inclusive, faster, easier and/or cheaper data collection. Given the novelty of these methodologies, funders such as the NIH have a strong need to foster research to validate these methods. At the same time, we are swimming in so much data especially considering cell phone data and metadata as well as quantified self data. We still haven’t found ways to integrate, analyze and communicate data for intervention and feedback.
- Public health trend spotting. Real-time valid insights to drive partnerships and intervene sooner. Think what could have happened in Flint if lead poisoning could be monitored on a heat map with predictive analytics, machine learning, and data science that incorporated various sources of data.
- New care delivery models (ie. ACOs, PCMH). Leverage the momentum and financial incentives of these models to fight data xenophobia and accelerate collaboration by standardizing and validating data sharing agreements, measures, and best practices.
What’s Next
Analytics and actionability must go together for meaningful improvement. Data has the potential to deliver proactive, personalized care to the right person at the right time in the right place.
These models often bridge the gap between the health care delivery system and community. Public health is ideal to bridge the healthcare/community divide given their systemic perspective and mandate to support community while promoting health, making them a trusted advisor to both stakeholders. Making the public health system the facilitator allows each stakeholder to do what they do best: the health care system delivers clinical care, community-based agencies and organizations address the social determinants of health while the public health system provides navigation and translation and analyzes data to determine the ROI.
At the end of the day, behavior change is what drives improvements in health outcomes. Most of a person’s health is tied to ex-health care. If the best data is collected and analyzed perfectly, that won’t matter if individuals are not motivated or able to perform the actions that will make them healthy. This requires a constant, consistent investment of time and resources in understanding not just what requires change, but also how these innovations are embedded and connected to existing systems and practices that are known to work.
Resources
Transdisciplinary Collaborative Center (health disparities in Health IT research)
