“One thing about The Human Diagnosis Project -- It’s not thinking small. Its goal is to build an open diagnostic system for patients, doctors and caregivers using clinical experience and other information contributed by physicians and researchers around the world. So, not small.”
A recent Forbes article Building A 'Collective Superintelligence' For Doctors And Patients Around The World dives into the vision and promise of the Human Diagnosis Project, a Moore Foundation grantee. Read more here.
The Moore Foundation is supporting the Human Diagnosis Project in applying its novel software platform to improve patient safety by understanding how it can be implemented in an ambulatory, or outpatient, setting. When it comes to patient safety-related issues, diagnostic errors tops the list. In primary care settings, an estimated one in 20 U.S. adults experience an error in diagnosis – defined as missed, delayed, or wrong – and about half of these errors can lead to serious preventable harm. Few interventions have been developed and tested to reduce diagnostic errors in real-world primary care settings. In current usual practice, providers commonly diagnose patients independently, without the benefit of collaboration or consultation with other health professionals or the use of health information technology, leading to increased risk of diagnostic errors. In its recent report on diagnostic errors, the National Academy of Medicine highlighted inter-provider collaboration and the development and utilization of health IT innovations in the diagnostic process as two key strategies essential to reducing diagnostic error in the ambulatory care setting.
Here is where the Human Diagnosis Project (Human Dx) comes into play. It is a web-based, mobile software that allows health care providers the combined insight of multiple doctors and machines on any given clinical case. Using the Human Dx software application, providers can ‘check’, or electronically consult on a case with multiple peers and receive a single, collective opinion within 12-24 hours. Preliminary data suggests that Human Dx outperforms 97 percent of individual physicians (92 percent diagnostic accuracy of Human Dx vs. 63 percent for the median physician). Although Human Dx has been shown to improve diagnostic accuracy, applying this innovative tool to improve patient safety will require a scientific understanding of how it can be implemented in a real-world ambulatory care setting that provides care for high-risk populations, evidence of its effectiveness in practice and a business case for providers and payers. The foundation will enable Human Dx to address these requirements.
Human Dx will partner with University of California, San Francisco and the San Francisco Health Network (an integrated network of 12 primary care safety-net clinics in the county) to conduct feasibility and effectiveness testing. Designing and adapting workflows in this type of multi-clinic setting, which represents a diverse, medically complex patient population, will inform potential broad-scale implementation efforts.