A version of this article was originally published by Becker’s Hospital Review on April 25, 2017.
The U.S. healthcare system has been plagued by rising costs and inefficiencies attributed to the fragmented nature of care delivery and communication. As frameworks for interoperability and interconnectedness evolve to eliminate healthcare siloes, there will be opportunities to more quickly aggregate historically fragmented disparate data sources – which, when applying advanced parallel processing, the ability for real-time machine learning and NLP and other artificial intelligence (AI)-related approaches can be realized. The goal is to make that enhanced, patient-specific analytical derivative available to the healthcare delivery system so it can change how payers, providers, and other healthcare organizations engage with patients and drive better outcomes under value-based care.
As such, scalable big data platforms can empower the future of machine learning, impacting health systems and providers, enabling real-time patient specific analytics at the point of care. The application of these expanded data and analytic models derived in part through machine learning technologies will also drive improvements and efficiencies elsewhere across the healthcare ecosystem, including reducing medical errors while improving cancer detection and diagnosis of mental health conditions.
But what’s coming is even more interesting. Take for example telemedicine; it is now very much conceivable that in the not too distant future, clinicians will be able to access in real-time longitudinal patient clinical data during virtual visits including clinical recommendations guided by an individual’s unique genotype, phenotype and socioeconomic circumstances. This is a further extension to what many are speaking of with precision medicine – yet another buzz phrase growing in relevance in recent years.
Methodologies under the AI umbrella, such as machine learning and NLP are slowly becoming part of the care delivery continuum, and a growing number of stakeholders across the healthcare ecosystem. Many organizations are making investments in these technologies which rely heavily on an interoperable system that can integrate and aggregate disparate complex data inputs. This is true within pharmaceutical discovery and commercialization where data is changing the traditional paradigms for clinical trials and drugs are now accompanied by companion diagnostics that ensure the right drug reaches the right patient at the right time.
The healthcare delivery ecosystem is preparing for machine learning and other AI-related technologies by increasing its investment in scalable, modular platforms that deliver critical patient-level insights directly into clinical workflows in real time. Seamless, real-time clinical data retrieval helps meet the increasingly tighter response turnaround times of regulatory bodies. And with a process that is more transparent and efficient, less intrusive and burdensome for providers, and less taxing on provider site personnel, healthcare organizations experience increased provider satisfaction. Hospital systems, integrated delivery networks, accountable care organizations, independent physician associations, and other physician networks now realize that they must embrace technology platforms that enable large-scale data integration and aggregation which fuel analytical competencies that rely heavily on mastery of large comparative datasets.
Ultimately, clinicians want to – and should – spend more time with patients. Relevant and focused data-driven insights can enable this, but there is a steep learning curve for some. Perhaps it’s a stretch goal, but wouldn’t it be nice that by the end of the next decade, nearly every “data point” related to a patient — whether that be from their genotypic and phenotypic profile, their self-reported commentary or a wearable device, or even insight into what the patient purchased at the convenience store the night before — were aggregated and analyzed to inform and thus improve the patient-specific quality of care delivered? When this happens, we will undoubtedly be far along in the transition to value-based care, and machine learning, AI-related technologies will be a big reason for its progression.