Imagine the number of cancer researchers and drug companies and medical institutions clamoring for case studies of patients to help them develop better medicines and treatments while doctors around the world see thousands of patients every day and produce hundreds of thousands of notes which they routinely stash away in file cabinets and personal computers rarely to be seen by others.
By Dimitri C. Michalakis
“Doctors dictate these notes for legal and clinical purposes as a summary of what they did for the patient, or what the patient’s experience was in the hospital or while undergoing certain procedures, but from a research perspective this data has not been leveraged: it’s been stored away in doctors’ computers or the file server of a large health care system,” says Dr. Gerasimos Petratos, at 37 already a former global head of Healthcare Data Analytics at the Roche Group and now a founder of Hiteks, a firm that mines and refines the avalanche of clinical data produced every day on the spectrum of medical conditions and makes it readily available to researchers and medical institutions.
“The focus of our business,” he says, “is to apply sophisticated tools to search and find relevant medical information from all the digital information being collected now. Enough of it has already become digital and life sciences companies--specifically pharmaceutical, medical device, and biotech--are trying to get a better understanding of that data to help inform and develop new products: innovative products that will meet the needs in the markets.”
Working with these companies, Hiteks uses language technology to “take the unstructured information and automatically structure it into a base they can report from and they can use to meet governmental reporting requirements. And at the same time we’re taking their legacy data, which amounts to hundreds of millions of records, and we’re organizing that into a clinical research database.”
As Petratos describes it, Hiteks puts information through a multi-step process. “We first strip any identifiable information from the notes and records; then we organize it by disease areas. All the diabetic diagnosis is put together and restructured so when we want to go down to an individual patient, even though we don’t know who that patient is, we can get a good understanding of what medication they are on, how long they were diagnosed for, what medications they were given. Imagine trying to get that information manually when it’s time-sensitive? An average institution sees 25,000 patients a year and it’s very difficult for them to comb through every note they have on file. We use software and software technology that allows us to provide a complete and accurate record and pull out meaningful data.”
Hiteks, while new, already has contracts with some of the biggest institutions in the field, including Roche, which was among the first to join up. “My work over there was directly a steppingstone to what I’m doing now,” says Petratos. “We have Roche as a customer because we know their business pretty well. We can help accelerate the identification of patients that meet the criteria of their research studies and we are also decreasing the barrier for those institutions so they can feel comfortable referring patients to these studies without having necessarily the people or the facilities to run these studies as part of their organization.”
Seeing the larger picture is not only the goal of Hiteks, but the personal vision of Petratos, who majored in biology at Cornell and got a medical degree from Howard, but saw his mission to diagnose whole populations rather than individual patients.
“I realized I personally wasn’t suited to seeing tens or hundreds of patients a day,” he says. “I wanted to be involved in more population-based treatment and in generating new knowledge for the medical field. So for me at the time the best option was to join a drug company that was doing the work and developing drugs and treatment for a large number of people and that allowed me to get to know the industry and the approaches, both the scientific and the analytic and business processes.”
At Roche, he was the wunderkind who in a few years rose in the ranks from designing ways to assess the effectiveness of medicines for oncology and neurological patients, to directing the sweep of clinical data needed to discover, approve, and promote a new drug, to managing a team supporting the development of Phase 2-4 drugs.
But he was eager for more.
“There’s a saying that if a pharmaceutical company were only made up of marketing people everybody would be in jail; and if it was only composed of science people then they would go out of business, so there’s really a delicate balance that’s needed,” he says. “I learned a lot while I was there, but after being with the industry for eight years, I was able to make out what was efficient and what was inefficient and what were better way to do things.”
So he started Hiteks with a business partner, Lalit Bakhi, who he worked with at Roche.
“He was involved in the professional services area and did a lot of recruitment and staffing of people that we ended up using in one way or another, so we had a good relationship and I was always interested in his approach to business,” he says of his partner. “I knew if I was going to do anything in health care I needed to partner with somebody who had access to a group of investors as well as the tech thinking that allowed us to create a capability fairly quickly in the market and then grow it from there in such a way that we knew we could compete.”
Of course, his company is not the only one aiming to provide clinical data in a high tech way. “It’s been tried at larger institutions that have good clinical databases, but it’s still early on. The idea we had is let’s get on this now, because it’s going to be in such high demand in a couple of years and having us well established now will assist us in the near future.”
The company is based in the medical center at Columbia University, not far from the Washington Heights neighborhood where Petratos was born, before his family moved to Queens and then Port Washington, Long Island. (His dad Vasilis immigrated from Cephalonia, his mother Katerina’s family is both from Cephalonia and Sparta, his sister Theodora is an ophthalmologist, and he often works now with his brother Nikos, a lawyer.)
After college he worked at a Native American non-profit “both applying my knowledge from the university to specific groups and also getting involved with health care issues. I knew I was going to be applying for medical school and I knew the non-profit was working with HIV as a disease within the American Indian population and its different nations. One of the big capabilities that we worked on is to allow a more efficient way of identifying HIV patients that could be enrolled in clinical trials of HIV drugs which at the time in the late ‘90s was a huge need.”
He’s also a member of the Hellenic Medical Society of New York and the Global Hellenic Medical Network.
He hopes the success of his business will drive the science and the success of the science will drive the business.
“I’m trying to achieve that combination,” he says. “I’m trying to make it work for my new business because I think you do need that combination of things to be successful and I don’t think you can do one and not the other.”
