Moneyball Medicine: Data-Driven Healthcare Transformation
In his 2017 book MoneyBall Medicine: Thriving in the New Data-Driven Healthcare Market, author Harry Glorikian offers a prescription for how data and analytics will shape the future of healthcare delivery in the United States. We are all familiar with the Moneyball concept and how it transformed professional sports over the course of the past two decades – with an assist from Brad Pitt. Glorikian has applied this same thinking to healthcare, noting “When we have systems in place for collecting and analyzing the data, we can use those insights to transform healthcare”.
Glorikian was motivated to write MoneyBall Medicine after observing how data and technology had transformed industries ranging from professional sports to financial services. He remarks, “Disruption is messy. There’s a lot of uncertainty in figuring out how technology like artificial intelligence and machine learning can impact healthcare. I am constantly surprised to see the resistance in healthcare to using data and technology”.
I asked Glorikian how data-driven practices can be employed to improve and transform the healthcare system today. Glorikian cites patient management as an area of visible change, commenting, “the availability of data and analytics are transforming patient management at a revolutionary pace”, starting with electronic health records (EHR’s). Physicians now have hundreds of thousands of patient data points at the touch of their fingertips. EHR’s are also becoming smarter, no longer just massive data repositories of clinical notes, lab values, and radiology reports. EHR’s are slowly being enhanced to help doctors identify patients who are due for routine cancer screenings or who are at risk for disease. Glorikian observes, “Data-driven healthcare is transforming medicine from a reactive practice to a proactive one”.
Data and analytics are also transforming how doctors practice medicine, noting advances in telemedicine, remote care, and wearable and connected devices. Glorikian notes, “We have seen a massive shift to remote care and telemedicine during COVID-19”. Data-driven initiatives are also helping healthcare systems and hospitals become more efficient as they navigate value-based reimbursement.
Drug discovery and drug repositioning are additional areas that are benefitting from data and analytics. Historically, drug development could take decades at great expense. Today, computers using AI can screen thousands of virtual molecules against libraries to identify those that have the best chances for success.
Lessons from COVID-19
Though scientists, epidemiologists, and public health experts have access to massive amounts of data, the models are only as good as the data feeding them. Data which was available at the start of the COVID-19 outbreak was both incomplete and insufficient to account for the many permutations of social distancing and shifting quarantine policies. This was magnified by a pathogen of yet-undetermined virulence, operating against a patchwork of healthcare infrastructure of varying quality. Glorikian observes, “When you consider the number of variables that must be considered to get an accurate projection of how many hospitalizations, how many deaths, how long the social distancing needs to last, when this will be over—well, it is just mind-boggling”.
One of the most important successes from COVID-19 has been the incredible speed with which the international scientific community has been able to sequence the viral genome. The response of the scientific community will help us better understand the virus and predict how it will behave based on its similarities and differences with other viruses. This will have an impact on future resource allocation and logistics.
The COVID-19 outbreak offers some valuable lessons by presenting a potential playbook for the next time a pandemic threatens the U.S. and the world. The data that we are capturing now, from how many ventilators were needed at a hospital at the peak, to the true impact of social distancing on mortality, can be used to help scientists develop more accurate models in the future.
An April 23rd article in Becker’s Health IT notes, “The COVID-19 pandemic will have a long-lasting effect on the healthcare industry, with new potential for digital health initiatives and data-sharing to help patients and public health surveillance, according to Judy Faulkner, founder and CEO of Epic”.
Yet challenges to data and analytic adoption remain. Glorikian observes that “Only recently have patients been able to access their medical records through online patient portals. Physicians remain hesitant to rely on analytic models and AI when they are perceived to be black boxes.”
Hospitals can be expected to standardize data definitions so regulators such as the CDC can access data more rapidly for monitoring public health emergencies. Epic’s Faulkner notes in the Becker Health article, “If people define the data differently, then you can’t aggregate it. And just collecting the data when it isn’t standardized doesn’t get you very far”. The classic data preparation challenge.
Glorikian is optimistic however, noting, “We see data-driven initiatives in some unexpected places, such as rural hospitals that are leveraging the power of data and analytics to help weather the incredible financial and staffing challenges they face”. He believes that data and analytics will become increasingly central to the future of healthcare management, predicting that future pandemics will play out with better models that have been built on the data and lessons of COVID-19.