How Probably Genetic's Dashboard Benefits Patient Finding Campaigns
February 13, 2022
Cheyanne Perry Suarez
Drug developers and patient advocacy groups use Probably Genetic’s machine learning platform to find undiagnosed rare disease patients. This occurs in two steps: (1) To begin, undiagnosed patients complete the phenotyping algorithm known as the Symptom Checker, and (2) then patients predicted to have the target disease are tested using Probably Genetic’s telemedicine testing service.
You’ve already learned all about Navigating the New Features of Probably Genetic's Symptom Checker. The data collected by the Symptom Checker displayed in Probably Genetic’s dashboard. The dashboard provides real-time insights on the progress of patient finding campaigns.
The Dashboard is divided into three main sections (1) Overview, (2) Campaign, and (3) Sequencing and answers several key questions. This post will focus on an Overview.
#1. How many patients are dropping off in the funnel?
The overview below shows the campaign’s status. Probably Genetic’s drop-off rates are generally below the industry standard.
- Unique Submissions— The number of people that completed the symptom checker
- Genetic Tests Offered—The number people that were offered testing
- Genetic Tests Ordered—The number of people that claimed their tests
- Results Ready—The number of completed tests
#2. How fast are patients signing up?
The curve represents sign-up rates. Probably Genetic’s programs scale rapidly because the technology is designed to provide patients with the simplest-possible access. Most programs receive submissions from more than 10 countries because patient communities share the links broadly.
The dashboard also features a live count of how many patients with the target disease are found during a testing program, which includes patients with related conditions.
#3. Where are patient populations located?
The map displays the locations in which patients submit to Probably Genetic’s symptom checker. Zooming out offers a visual representation of submissions from across the globe. The map becomes more specific upon zooming in. Commercial teams, medical science liaisons, and patient advocacy teams use this feature to learn which regions the campaign covers and to make future targeting decisions.
All locations are always anonymized within the first three digits of the actual zip code to protect the confidential information of patients. Our partners never have access to a patient’s name, private address, or contact information such as email and telephone number.
#5. What is the phenotypic profile of patients participating in the program?
An interactive pie chart breaks down unique symptoms patients exhibit by number and percentage. Example phenotypes include unique patients with mitochondrial myopathy, cardiomyopathy, and ragged-red muscle fibers. Phenotypic data is ICD10, SNOMED, and Human Phenotype Ontology (HPO) compatible for partners to use in future programs and research.
#6. Which patient sourcing channels yield the most patients with specific diseases?
[Insert submissions by source with map]
Patients with rare disease often struggle to obtain an accurate diagnosis, so they turn to internet research in hopes of answers. A large part of Probably Genetic‘s patient outreach takes place online through social media websites such as Facebook and Instagram or by influencers and direct traffic.
The dashboard shows partners which social sources have been most productive at finding the target patients participating in their campaign. These sources may differ depending on the region, which is why Probably Genetic incorporates a map next to the pie graph to portray the sources and areas that drive the most traffic to Symptom Checker submissions. As previously noted, locations have been anonymized to Zip 3 level to protect patient privacy.
#7. Does Probably Genetic’s technology reach patients of all ages?
Probably Genetic’s programs work reliable for patients of different ages, regardless of whether or not the patient or their care-taker are the focus of the campaign., Patients are also encouraged to share testing programs with their family members that have symptoms of disease, friends, or other patients they know. That ensures a range of patients from neonates to the elderly.
Additionally, the most frequented social media outreach sources vary by age. For example, Facebook attracts the older patients, while other social media platforms like Instagram are geared towards adolescent patients and young adults.
#8. How can the physicians treating patients with the target disease be contacted?
Probably Genetic asks patients to submit the names of their treating physician(s). Drug developers need to know which physicians are treating patients with specific conditions in order to let those physicians know if clinical trials or approved treatments become available for their patients. In the dashboard, pharma partners and advocacy groups have access to the contact information, medical center affiliation, and specialty of physicians treating patients participating in the programs.
#9. Which genetic variants were found in patients that were tested?
The dashboard displays anonymized test result information, including the gene, associated disease, cDNA protein change, and the variant’s ACMG 2015 classification.
Cheyanne is the patient community manager at Probably Genetic. She uses her own experience of living with a rare genetic condition in her patient advocacy work. After graduating with her degree in psychology, she is especially interested in the biological processes that connect physical illness and mental health.