Lessons Learned from Mapping the Process of De-Implementation of Routine Cytogenetics in the Diagnosis of Pediatric Acute Lymphoblastic Leukemia

Introduction: The diagnosis of pediatric cancers has traditionally relied on a cascade of testing approaches that allow for cancer identification. The advancement of diagnostic capabilities using genome sequencing has led to redundancy in information reported at some centers. Consolidation of the diagnostic approach has the potential to improve resource allocation and decrease costs. At St. Jude Children's Research Hospital (SJCRH), conventional cytogenetics in the diagnosis of pediatric acute lymphoblastic leukemia (ALL) was discontinued with the introduction of clinical genomics testing (including whole genome, whole exome, and whole transcriptome sequencing). Integrated results showed comparable and superior findings when genomics testing was compared with cytogenetics, resulting in the de-implementation of cytogenetics in 2018. De-implementation provides an avenue for removing, replacing, reducing, or restricting a healthcare practice. De-implementation of a routine component of pediatric cancer care and its influence on care delivery has been previously unexplored. In this qualitative study, we aimed to understand and map the process of de-implementation of routine cytogenetics in the diagnosis of pediatric ALL at SJCRH.

Methods: This qualitative study consisted of interviews with current or former multidisciplinary staff of SJCRH who participated or were impacted by the de-implementation of routine cytogenetics for diagnosis of pediatric ALL. Rapid qualitative analysis was used to analyze interviews. Analysis approach was informed by the consolidated framework for implementation research and data was organized to develop a process map using swim lanes to demonstrate activity sequence. The process map was iteratively revised by the research team prior to member-checking with participants, to ensure findings aligned with their experiences.

Results: Thirteen individuals participated in the study. Of these, 54% (n=7) worked in the pathology department (faculty or lab support) and 46% (n=6) worked in oncology. 62% (n=8) had over 21 years of experience. Overall, participants described de-implementation of cytogenetics in pediatric ALL successful, with no impact on patient outcomes. Four swim lanes were generated relating to processes within the departments of pathology and oncology, division of cytogenetics, and institutional partners (i.e. computational biology). Process decision-making began within the department of pathology who ran internal correlation studies (comparing outcomes of cytogenetics with clinical genomics testing) prior to de-implementation. Motivation for de-implementation ran alongside development of a new clinical trial for upfront treatment of pediatric ALL. While some participants described hesitation with cytogenetics de-implementation, most considered the process change to be a natural evolution with potential to improve molecular classification of leukemia, improve resource allocation, and reduce costs. Perceived risks included loss of competency in cytogenetics, initial delayed turnaround time, and career insecurity, which were addressed institutionally. Participants believed that with similar infrastructure and approach, cytogenetics could be discontinued for pediatric ALL at other centers, recognizing that clinical genomics testing is not necessarily accessible broadly. Lessons learned for future intervention de-implementation processes includes the need to improve planned discussion about purpose, logic, and evidence supporting decision-making particularly during the planning phases of de-implementation.

Conclusions: De-implementation of routine cytogenetics in pediatric ALL in favor of clinical genomics testing has been successful at SJCRH. Drivers of success included access to advancements in diagnostic testing, innovation leaders, and internal correlation studies confirming appropriate test characteristics. Staff impacted by this process change sought more thorough communication strategies. Findings will be used to generate an approach for future de-implementation of a chosen diagnostic intervention and provide an example of de-implementation in pediatric cancer care.

No relevant conflicts of interest to declare.