Clinical trial sponsors will increasingly rely on radiomics, the science of advanced imaging analytics, to enrich and inform clinical trials strategies, leading to deeper insights into patient populations. Using AI and machine learning to extract new types of data from traditional images such as CT and PET scans, radiomics enables researchers to develop more quantitative and robust inclusion and exclusion criteria, as well as predict patient outcomes. Additionally, radiomics offers tools to objectively quantify the features of tumors and lesions that are predictive of future biological behavior, enabling an earlier understanding of a disease’s likely path of progression.
In 2022, healthcare organizations and life science researchers will continue to seek new ways of extracting insights from real-world unstructured data. This data, often trapped in the notes sections of electronic health records (EHRs), may include critical information such as symptoms, diagnoses, or outcomes. Artificial intelligence, such as natural language processing (NLP), will increasingly be relied upon to make sense of free-form text in EHRs to inform real-world evidence for researchers and quality care reporting for clinicians. With NLP helping to make sense of real-world data, we will likely see a more profound understanding of patients and diseases in a shorter period of time than possible via traditional approaches.
Looking ahead to 2022, we expect that the continuation of both a tight labor market and cost containment measures will make intelligent process automation and leveraged staff augmentation key alternatives for the provider community. More hospitals and health systems that are dealing with severe staff shortages will look to automate previously manual services as a way to cut costs and remain competitive. Additionally, workforce performance management, gamification, and other non-traditional tools will play a larger role in managing remote workforces that are not physically connected to the organization. This will be especially impactful for administrative roles at hospitals, such as in the IT and revenue cycle departments.
Pharma and the life sciences will continue to seek artificial intelligence (AI) and machine learning (ML) technologies that help them solve data challenges from bench to bedside more quickly. Look for organizations to embrace new cloud-based approaches for tools like natural language processing (NLP) that can be easily plugged into data scientists’ existing workflows without requiring the implementation of an enterprise solution. Companies require agile tools that can be built into existing processes to find the answers they need—and don’t have months or years to implement potentially cumbersome enterprise software. With a cloud-first strategy, AI and ML technologies escalate drug discovery and development because data scientists can rapidly find answers for specific tasks or for multiple issues across the organization.
Healthcare will see escalating rates of adoption for text analytics tools such as natural language processing (NLP) due to three driving factors:
- With the rapidly approaching deadline for the Interoperability and Patient Access mandate, which includes requirements for the interoperability of full-text medical records, organizations will need ways to harness the coming deluge of unstructured data. Solutions such as NLP will help provider and payer organizations process the data to enrich predictive algorithms and drive better clinical and financial outcomes.
- The serious emergence of large tech and cloud vendors into the healthcare text analytics space has increased the spotlight on tools such as NLP, which can now be easily consumed in a convenient way through cloud-based approaches, instead of traditional large-scale software deployments.
- The pandemic has shaken up healthcare in several enduring ways, including the increased acceptance of cloud-based technologies that allow users to gain access to data while working remotely. The pandemic has also highlighted population inequities that have impacted outcomes and raised awareness of the importance of looking beyond information in patients’ structured data and surfacing social determinants of health, which can be done using technologies like NLP.
Data show that while hospitals have allocated more resources to infection prevention and control efforts to contain the spread of COVID-19, it has largely come at the expense of controlling other far too common healthcare-associated infections (HAIs). It’s true that a larger volume of sicker patients at higher risk of infection and sepsis have been admitted to the hospital over the last year, but the CDC concluded that 2020 increases in HAIs were also a result of lacking surge capacity and other operational challenges. Looking ahead to 2022, as hospitals take aim at controlling all HAIs in addition to COVID-19 with more resilient care teams, they will be looking more closely than ever at AI-powered technology to support proactive and real-time monitoring of patients to empower staff with quick risk identification abilities and opportunities for earlier clinical intervention.
Accelerating changes to clinical practice with new evidence – Health systems are still grappling with the far-reaching effects of the pandemic, yet their focus on quality improvement amid the broader shift to value-based care must continue. Quality improvement research initiatives at these organizations hold the key to better patient outcomes and financial performance, but these are time-intensive programs that make it difficult to efficiently surface and implement new evidence into clinical practice. In the wake of a pandemic that laid bare weaknesses of our current delivery system, I anticipate an accelerated uptake of tools and solutions designed to shorten the cycle between identification of clinical problems and implementation of clinical solutions based on evidence.