Learning Methodologies Core
The Learning Methodologies Core (LMC) is dedicated to advancing clinical research and practice through the development and dissemination of novel statistical methods, decision-making tools, and research designs. With the unprecedented amount of electronic patient health information currently available, the LMC recognizes the need to funnel relevant, evidence-based, up-to-date clinical information to healthcare providers and patients in accessible, user-friendly formats. With expertise in biostatistics, bioethics, psychology, clinical research, and behavioral sciences, LMC investigators are devising innovative solutions geared towards the individualization of patient care, the respectful consent of patients for clinical research databases, and the dissemination of decision-making tools for healthcare providers and patients. See below for more information on LMC projects.
Pneumonia Etiology Research for Child Health (PERCH): Pneumonia, a leading contributor to global childhood mortality, can be caused by over 30 different pathogens. Correct identification of the pathogen causing pneumonia is critical for effective treatment of the disease. However, the most widely used tests to identify the causal pathogen in pneumonia cases are imperfect. LMC investigators are designing statistical tools that combine the results of several diagnostic tests to predict the most likely infecting pathogen for each patient diagnosed with pneumonia, enabling more effective treatment of each patient.
Prostate Cancer: Though prostate cancer is commonly diagnosed among men in the United States, the optimal treatment course for the disease is unclear. The current treatment options and active surveillance are associated with varying degrees of mortality and impact on quality of life. LMC investigators are creating models that will estimate the effects of different treatment options on patient outcomes, taking into account disease severity and patient preferences. The end-product will be a support tool - designed for clinician and patient use – that will facilitate treatment decisions based on individualized risk prediction.
Adaptive Trial Design: Randomized clinical trials give important information on treatment effectiveness, but in many cases are not optimized to examine how treatment outcome may vary in patient subgroups. The LMC is designing and testing novel adaptive trial designs to determine the benefits of treatments in groups defined by age, sex, and disease severity. Simulations will be conducted to answer clinical questions relevant to stroke interventions, cardiac resynchronization therapy, Alzheimer’s disease, and HIV prophylaxis.
Open-Source Learning Environment for Research on Individualized Health (OSLER inHealth): Hopkins inHealth is dedicated to the rapid dissemination of the statistical tools and methodologies developed by the LMC. As such, Hopkins inHealth is developing OSLER inHealth, an R-based statistical package that will allow researchers to adapt the statistical tools developed by LMC investigators for their own research purposes, enabling the faster adoption of innovative inHealth solutions by the wider healthcare community.
Patient Consent: To establish a learning health system (LHS), tools are being developed by Hopkins inHealth investigators to extract patient information from electronic medical records. Recognizing the importance of building patient trust to successfully implement a LHS, LMC investigators are developing procedures to engage patients in the research consenting process.