Precision medical company BaseHealth today released a genomic API for health, fitness, well-being and medical application use. The API provides access to five services that can combine an individual’s genomic information with evidence-based research to give more precise, personal recommendations for illness prevention, injury rehabilitation, health and lifestyle maintenance, and chronic disease management.
BaseHealth is a Stanford University-alum software-as-a-service startup that adds individual genomic profiles to its knowledge base of medical evidence in order to help medical centers and physicians provide personal health plans to their patients.
“Most common diseases have multiple causes, so we need to understand all of those causes, not just one or two of them,” says Prakash Menon, CEO of BaseHealth. “So we built a model that measures someone’s disease risk based on all risk factors of a particular disease. For example, for Type 2 diabetes, your genes, lifestyle, physical activities, family history and medical history all play a factor, so we have built a model that includes all of these when calculating risk.”
Menon says that BaseHealth has mapped the causal factors for the 40 most common diseases. Through its SaaS product, it enables physicians to enter the genomic profile of their patients in order to produce a health plan suited to each patient’s particular needs. In many cases, explains Menon, a full genomic profile is not needed; relevant genetic markers that may have been mapped in other diagnostic tests are sufficient for a partial analysis and an optimized health plan.
As a cutting-edge technology in an emerging medical domain — that of genomic assessment of an individual in order to optimize healthcare — BaseHealth realizes its reach can extend much further if it opens up its API products to a larger developer pool. “This new offering will empower us to reach out to a lot more people and scale up a lot more efficiently,” says Hossein Fakhrai-Rad, chief scientific officer and founder of BaseHealth.
The BaseHealth API enables five services via API calls:
- Curated Content: By adding a query parameter for the name of a disease, BaseHealth can return evidence-based information about a disease, including associated genes/variants, risk factors ranked by their impact on the disease, preventive health actions that may reduce the disease risk, and general statistics about the disease and its prevalence.
- Anonymous Assessment: The API can be used to input an individual’s genetic, dietary, lifestyle and environmental information so that the BaseHealth API can return a risk profile based on a specific disease, including probabilities of risk and the key risk factors that may contribute to an individual’s personal probability of getting that disease.
- Drug Response: Based on a person’s genomic profile, the BaseHealth API can provide information on how effective any medication will be for the individual.
- Food Response: Based on a person’s genomic profile, the BaseHealth API can provide an analysis of the impact of any food on that individual's health and disease.
- Personalized Assessment: For service providers where end users authorize use of their personalized health information, the BaseHealth API can maintain an ongoing lifestyle and health action plan.
The BaseHealth API is free up to 1,000 calls per month. For more than 1,000 calls, the cost begins at 10 cents per call and decreases based on volume.
Menon explains why the APIs were launched:
As part of the commercialization process, we saw that clients wanted to use our assessment engines in their own applications. People were wanting to build applications around our assessments. Since our architecture was API based anyway, we have wrapped APIs around our assessments and now made them available.
Menon says this was made easier because of the application architecture model BaseHealth had constructed. Its SaaS product is basically a single page app, with the front end written in Backbone and making service calls to the back end via APIs already. So opening up the APIs to enable a larger ecosystem of developers to create new products beyond BaseHealth’s key focus areas was much easier to instigate.
"We believed in our architecture," Menon says. "We wanted to use the same back end for our Web and mobile apps. So when the market told us we needed APIs, we were very happy to do it. It turned out to be reasonably easy."
BaseHealth was then able to select an API provider (3scale) to manage developer accounts and track rate limiting and payments and commercialize its API offering.
Health application developers can also make use of some of the standard information in BaseHealth’s knowledge library of evidence-based health and medical data without knowing any genetic information about an individual user. The curated content service provides metadata on the diseases that BaseHealth is cataloging, so calling this information details of all the risk factors, ranked by their level of impact.
“It is like a very specialized WebMD call,” says Menon. “So you could display that to someone, or you could use it to set up the assessment call and then add their personal profile, then you could pass that in as a JSON structure. Then it would return it as an assessment.” This means that even without the genetic profile, application integrations could create personalized health plans for users, then enable users to add more detailed information such as their analytics results from a 23andMe DNA test or any specific diagnostic markers they have received from their healthcare professionals.
Among the use cases Menon and Fakhrai-Rad imagine are applications for athletes that determine a specific nutritional and exercise regime based on personal fitness and genetic recovery factors in order to optimize sports training. They are also working with some startups in stealth mode that are applying the BaseHealth APIs to their specific “disease verticals” — businesses that are focusing on improving management for just one type of disease among the BaseHealth knowledge base.
They hope to see other imaginative use cases being built while not detracting from their focus of working with their own principal target group, which, for now, is predominantly virtual medical clinics. Menon says:
Our real target is what we call building virtual health clinics. People are trying to build a low-touch way to deliver health and information through online physician mediation. We are also looking at a whole heap of health assessment tools that are addressing the enterprise health market, where we believe we can add value there. Also, there is the personal health record space where a lot of activity is taking place to try and build sites where people can own their own health information.
While there are multiple use case opportunities for developers, the overall approach is driven by a strong biomedical model of health. This fits neatly with wearables, medical device technology and pharmaceutical industries but takes less advantage of the growing body of evidence that shows that social determinants — workplace stress, the cities we live in, or access to food, water and housing — play a part in our health and well-being.
For example, recent research shows that stressful working conditions impact physical and mental health, and the impacts are worse for those on lower incomes. Other research shows that Type 2 diabetes is more likely among households that have limited access to healthy food. Meanwhile, new online services like GeoHealth are emerging to analyze environmental conditions such as pollutants on health and well-being. How these sorts of wider social factors will be drawn into the BaseHealth model is something Fakhrai-Rad is assessing one peer-reviewed paper at a time:
The actual flow of how we added a disease starts with scientists: What disease do we have to add based on risk categories, genetic information and lifestyle? Then we look at environmental data. We would be more than happy to add that information into the model, as long as the scientific data behind that data is validated. Environmental factors that we currently cover are direct, like passive smoking and rayon exposure, but I do expect that social determinants of health data will be available over time.
As for meeting regulatory requirements for the storage of individual health information, BaseHealth says that its SaaS platform is completely compliant, while its API does not store any personal information on any user and is completely anonymous. Menon explains: “In terms of the API, we don’t store any of the information, except we have the personalized service, which is similar to the Blue Button initiative, which is essentially an OAuth token that an existing customer on our platform provides to share their information. So that is entirely within the ambit of their information and shared through SSL. In all other cases, it is anonymous information.”
Fakhrai-Rad is excited to see how developers make use of the new API. "We think a lot of diseases in the world are preventable, and our idea is to make this really accessible to people," he says. "I hope we get a whole bunch of application developers with the same mindset who are going to use their skills to engage with health."
Developers can sign up for the freemium access to prototype an application at BaseHealth’s developer portal.