Background: The reports of side effects on clinical trials describe expected toxicities of chemotherapy. However, these side effects, also called adverse events, are globally under-reported on trials, which means that clinicians do not have an accurate sense of adverse event rates. In the current system, adverse events are identified through time-consuming, manual medical record review. This study aims to develop a system that uses electronic medical record data to capture complex adverse events of chemotherapy for pediatric cancer and to prove that this method is more accurate than the current adverse event reporting system.
The specific aims of this application are to 1) develop algorithms to identify 10 complex adverse events using electronic medical record data from children treated for acute leukemia at two large children's hospitals, 2) compare the accuracy of this new method to that of the current system, and 3) demonstrate the utility of electronic adverse event capture in answering clinical questions by defining the incidence and risk factors of acute kidney injury. Methods: This study will use data from 1900 children with acute leukemia treated at the Children's Hospital of Philadelphia (CHOP) or the Texas Children's Hospital (TCH) from 2002 through 2017. Algorithms will be developed to identify adverse events by extracting electronic medical record data at CHOP. Once finalized, the algorithms will be tested at TCH. Using chart abstraction data as the gold standard, the accuracy of electronic ascertainment and of trial adverse event reports will determined, and the relative accuracy of each method will be compared. Lastly, algorithms will extract creatinine results from the electronic medical record and the incidence of acute kidney injury will be determined for each leukemia type and by chemotherapy regimen. Risk factors for acute kidney injury will be explored. Career Goals and Environment: With the support of this K07 award, the applicant, Tamara P. Miller, MD, MSCE, will learn how to use electronic medical record data for clinical research, obtain formal training in clinical informatics and implementation science, develop expertise in clinical trial design, and improve her knowledge of pediatric oncology and skills in scientific writing. To complete these training goals, Dr. Miller has assembled an experienced, complementary, and nurturing mentoring team led by her primary mentor, Richard Aplenc, MD, PhD. Her training plan includes formal coursework in informatics, tutorials, national conferences, and research progress and writing groups. She will benefit from the outstanding depth of resources and opportunities at CHOP and the University of Pennsylvania. Her long-term goal is to integrate the novel system of adverse event ascertainment she creates into pediatric oncology trials and to use the accurate datasets she develops to answer clinically important questions. With this award, Dr. Miller will be well-positioned to transition to her goal of an independent clinical research career focused on improving adverse event reporting and supportive care practices in pediatric oncology.
Children with cancer experience significant side effects from their treatments, but these side effects are under- reported on clinical trials and no work has been done to improve the current system of side effect reporting. This study will create a new system of identifying and reporting side effects that is more accurate and efficient than the current system. The results of this research will provide clinicians, patients and their families with a true understanding of potential side effects of therapies used to treat childhood cancer.
