Precision medicine has enormous potential to change cancer outcomes for >500,000 Americans annually by targeting the genetic mutations of their tumors with FDA-approved drugs known to more effectively treat their disease. Thus, accelerating the use of cancer genomics is a national priority with combined public and private investment topping $8 billion a year. Despite high significance and investment, uptake of precision medicine in clinical practice is low. Tumor genome sequencing is not widely used and treatments based on molecular profiling are infrequently implemented. Implementation science is an emerging field which offers a theoretically-informed, evidence-based approach to accelerate the translation of evidence into practice, but has yet to be applied to precision medicine and lacks tools to rapidly diagnose organizational challenges to innovation adoption. Using this approach, we have identified a number of critical gaps in current research on the barriers to precision medicine adoption. Focusing on the needs of community oncologists, who deliver the majority of cancer care in the US, we will: 1) Survey oncologists to identify precision medicine adopters, assess community oncologists' motivations for innovation adoption, and evaluate the degree to which precision medicine aligns with community practice. 2) Conduct linked, semi-structured qualitative interviews of physicians, staff and administrators involved in precision medicine implementation. Using the Theoretical Domains Framework, we will identify constructs key to implementation success and further describe the strength, frequency, and type of implementation strategies used by successful organizations. 3) Assess the feasibility of using natural language processing to more rapidly diagnose adoption and implementation barriers. We will apply an ontology of barriers and facilitators to data collected in Aim 2 and three extant qualitative datasets exploring innovation adoption. We will develop and train an automated feature extraction system to code data sets and compare congruence of results from human coding and artificial intelligence. This work is expected to advance precision medicine, implementation science and cancer outcomes.
Aim 1 will allow the first estimate of precision medicine adoption in community oncology. Barriers to precision medicine implementation identified in Aims 1 and 2 will be mapped to strategies known to be effective in addressing them, enabling a randomized trial of the comparative effectiveness of precision medicine implementation strategies. Ontologies and machine learning developed in Aim 3 will contribute to a larger machine learning effort launched this year to expedite the selection of effective, tailored implementation strategies. Ultimately, this work is expected to expedite society's return on investment in the precision medicine initiative and contribute to cancer patients' longer survival and enhanced quality of life.
