OR WAIT 15 SECS
Julian Upton is Pharmaceutical Executive's Online and European Editor. He can be reached at firstname.lastname@example.org
Pharm Exec speaks to Prognos CEO Sundeep Bhan about how the company’s unique offering will reach beyond life sciences as the long-promised benefits of AI become reality.
Launched in 2017 by eHealth entrepreneur Sundeep Bhan and family physician Jason Bhan after six years of development, Prognos is a healthcare AI company focused on informing mission-critical clinical and operational decisions the earliest for health plans, life sciences, and diagnostics companies. It is the only company specializing in applying AI and advanced analytics to clinical diagnostics; the Prognos Registry consists of 13 billion clinical diagnostics records for 180 million patients across 35 disease areas.
Pharm Exec spoke to Prognos co-founder and CEO Sundeep Bhan about how the company’s unique offering is being applied to the healthcare space, and how it will reach beyond life sciences as the long-promised benefits of AI become reality.
Sundeep Bhan: When Jason Bhan, MD, co-founder and CMO, and I started the company seven years ago, we were thinking about big trends in healthcare, about what was going to have the biggest impact on patients and the industry. There were a couple of themes we kept going back to. One was precision medicine, the whole idea of scaling medicine down to sub-segments and maybe one day down to individuals, and the role that data analytics was going to play in helping to understand all that and make it a reality.
A lot of precision medicine is based on understanding patients’ needs better, which is how physicians practice medicine: more than 70 per cent of the decisions physicians make about a patient are based on lab and diagnostic information. But when it comes to the healthcare industry, most of the information used in decision making is based on medical claims and prescription information, which by their nature are after the fact. If you want to understand what’s happening earlier, it is difficult if you’re using data that are retrospective. So that was our “a-ha” moment-if we can bring better decision making to the healthcare industry, it would mean better outcomes for patients and more efficiency for the industry.
What makes us unique is our clinical focus, our access to very large data in the clinical diagnostics space. We just crossed over 13 billion medical diagnostics records and we’re tracking about 175 million patients in the US in around 35 disease areas. Having worked with diagnostic data over the last seven years, we now have data from more than 200 labs and have been able to integrate and normalize these across all of the different labs.
It’s still early days as far as the potential of the work we are doing and what can really be done with AI but I feel that in the last couple of years, especially, there has been a shift. When we first went and talked to people about what we were doing, there was a lot of education required. We were educating the diagnostic companies, the life sciences companies. Today we are working with 25 pharmaceutical companies; that in itself is a validation of how far we’ve come.
From a data perspective, 90-plus percent of all the data in healthcare in the US was created in the last two years. Computational power and the ability to store vast amounts of data were very difficult even five, six years ago, but companies like Amazon and Google have leveled the playing field. The technology existed for decades, but the application of it was not really possible until recently. Now we have access to these tools, a lot of the technology is open source, and storage is pretty cheap. The investments in data and healthcare that have happened over the last decade have created the basic infrastructure to collect a lot of this data. This is really exciting for a company like Prognos, because we are able to pull all of that together. In the last two years, for example, our database has doubled twice.
There are three areas where we can showcase success. First, in the new biomarker-driven world, one of the biggest challenges that pharma companies have is understanding what tests are being conducted, who is doing the tests, and how are they are being used in decision making. These tests are the drivers for prescribing. We are in a unique position to help pharma understand those trends and testing, and also to predict the adoption of certain tests. When new tests, such as the PDL-1, came out, we could predict for our clients what the adoption was going to be.
Second is in identifying patients that are right for therapy. If you look at prescribing history, you’re looking at high-decile prescribing trends. That’s not really telling you where the patients are, it just tells you about doctors who wrote X amount of prescriptions in the last few weeks. A much better way is to look at the clinical information of a physician’s patients, especially their diagnostic information. Then you can start to understand where the market needs are. We have helped with 30Â–40 brands, providing them with weekly reports to help them understand where the patients are. We can also predict if a patient will need to go on a specific therapy in, say, the next 90 days. This is something that was not possible before.
The third area is around measuring and predicting outcome, and really understanding how patients are doing on a company’s drug or their competitor’s drug.
We have about 100 people in the company now. It has definitely been a challenge. There is a huge demand for this kind of talent and it will continue to grow. One big advantage we have had over other industries, however, is that we have been able to attract people from different industries, such as financial services, and we’ve been able to do that because we are a mission-based organization. If you are doing work that you know is impacting patients’ lives or is saving lives, it is a lot more fulfilling and exciting. Everybody can connect to that.
One goal is to expand our current relationship on the sales and marketing side. There are increasing amounts of pre-commercial and pre-launch areas where we are being called in to help with an opportunity, so that will be a big area of growth. And next year we plan to launch on the R&D side, which could be bigger for us than sales and marketing.
Also, pharma companies are sitting on a ton of data from all their clinical trials, whether successful or unsuccessful. If all that data can be mined and combined with real-world data sets, I think it is an opportunity for us to discover new uses for certain molecules, an opportunity to leverage technology to perform analysis in a more efficient way.
Ultimately, the applications of the algorithms we are developing and have developed will go beyond life sciences. So we also working with payers and are looking to work with both providers and patients in the future.