Project crowdsources specialists’ diagnoses for safety-net care

Timothy M. Smith , Contributing News Writer

Lack of access to specialty care is a notorious driver of health inequity in the U.S., but until now there have been few scalable options for connecting primary care physicians in underserved areas with their specialist counterparts to make major clinical decisions. A new open project—inspired by Wikipedia, Linux and others that leverage collaborative content creation—has already brought together thousands of volunteer physicians to provide online curbside consults to front-line physicians.

Roughly 29 million uninsured patients in the U.S. rely on the nation’s safety net public hospitals, community health centers and free clinics, but wait times for specialist care can be measured in months for uninsured patients, many of whom cannot afford to pay out of pocket for medical services and who have nowhere else to turn for health care.

The Human Diagnosis Project, or Human Dx, is a free online system, available on mobile and desktop, that enables primary care physicians working with underserved patients to access the insights of multiple physicians on a single case. It has more than 5,000 contributors from more than 60 countries across 40-plus specialties.

Similar to how users contribute articles to Wikipedia or how engineers contribute software code to Linux, the global medical community shares and solves clinical cases on Human Dx.

For example, when an attending physician receives a challenging case, rather than paging colleagues or searching for answers online, he or she can create a new case on Human Dx. From a mobile phone or a desktop computer, the physician enters a suspected diagnosis, along with relevant findings, such as the patient’s chief complaint and its frequency, duration and related symptoms, as well as physical exam findings, and laboratory and radiographic results.

Once the case is finalized, it is visible to the entire Human Dx community, as well as to specific contributors the physician wants input from. The physician is notified when a contributor gives input on the case and can see how the case was approached, step by step, and all the submitted differentials are compiled and sorted by relevance. The attending chooses from the diagnoses and later feeds details on the outcome of the case back into the system. Solved cases are then available to other users to test their knowledge.

Human Dx aims to verify the quality and accuracy of insights using the collective intelligence of the medical community and comparing this against stored data and algorithms. Over time, by automatically encoding the insights from similar prior cases—a process known as machine learning—the hope is that Human Dx will also provide software-based decision support to physicians everywhere.

Human Dx already has support from four of the largest medical organizations: the American Board of Internal Medicine, the American Board of Medical Specialties, the American College of Physicians and the AMA.

As lack of access to specialty care disproportionately affects racial and ethnic minorities, the AMA supports Human Dx as a way to mitigate health disparities. In addition, Human Dx’s approach of engaging community health centers and physicians in a collaborative effort matches the AMA’s vision of building a healthier nation by enabling health care teams to partner with patients to achieve better health for all.

The Association voiced its support for the project as part of the application Human Dx submitted to the John D. and Catherine T. MacArthur Foundation’s 100&Change competition. The winner of the competition will receive a $100 million grant to fund a single proposal that, according to the foundation’s website, “promises real and measurable progress in solving a critical problem of our time.” Human Dx has since been announced as one of eight semifinalists from a pool of nearly 1,900 applicants.

Staff at Human Dx, a Washington, D.C.- and San Francisco-based nonprofit and public benefit corporation, has included alumni from the World Health Organization, Facebook, Amazon and other leading health and technology organizations, as well as academic institutions such as Harvard, the Massachusetts Institute of Technology, Stanford and Yale.

Human Dx’s goal is to engage more than 100,000 physicians and to help millions of patients access specialist expertise over the next five years.

Human Dx is accessible on the web or through the Human Dx app for iPhone, available free from the App Store. After creating a brief profile, users may immediately post cases for help or solve cases from other physicians.