Patients with undiagnosed diseases have a unique and difficult journey. Many patients and families undergoing a diagnostic odyssey experience anxiety, depression, and, at the very least, much uncertainty. A complete understanding of all aspects of the patient’s condition (comprehensive phenotyping) is vital to make a diagnosis and to understand the condition. In many patients, the undiagnosed disease is genetic—either a known but difficult-to-diagnose condition or a new genetic disease, and comprehensive and accurate phenotyping (semantic or deep phenotyping) is critical to inform the genetic analysis that may lead to the answer. Comprehensive phenotyping is also vital to building cohorts of phenotypically similar patients to complement genomic data for gene discovery and to improve diagnostic capabilities.
Although the cost and ease of collecting and analyzing genomic data has improved rapidly, collecting the phenotypic data has not become more standardized, convenient, or less expensive, limiting such algorithmic approaches. Thus, a major challenge in clinical care and research aimed at understanding rare and undiagnosed diseases is phenotyping patients accurately, yet efficiently. One approach to deep phenotyping is to bring the patient to a medical center for a coordinated visit where comprehensive phenotyping takes place. On the other hand, patients may be the most comprehensive source of information about their condition, and two recently developed approaches have the potential to allow patients to provide information about their condition without going through expert phenotyping, and could provide a low-cost information gathering system to guide genetic analyses and lead to a diagnosis.
The overall premise of the proposal is that not only may “self-phenotyping,” using either the GenomeConnect survey or the layperson version of the Human Phenotype Ontology (HPO), be an accurate and comprehensive source of data on patients, it also empowers patients, and may be particularly beneficial to patients with rare diseases and the undiagnosed disease population.
The outcomes the project team hopes to achieve with this project are 1) computationally validate both the GenomeConnect survey and HPO layperson application and 2) assess the relative performance of these approaches in terms of their utility for self-phenotyping by patients. The project’s long-term goal is to develop a best-of-breed self-phenotyping instrument that is backed by a much greater understanding of patient and clinical diagnostic needs using informatics and statistical approaches. This type of phenotype data has never before been collected by patients for use clinically, and the researchers anticipate that neither tool in its original untested state will be perfectly suitable for robust clinical use as is.
Patients and other stakeholders will be an integral part of the research team to ensure the success of this project. To ensure the patient-centeredness of the approaches, the researchers will convene a patient advisory board of patient stakeholders who have had a family member with an undiagnosed disease and have dedicated themselves to helping other families. The board and other stakeholders will meet with the team regularly and guide researchers on all aspects of the project.