Not many healthcare startups hire a chief data officer in their early years, but the leaders of Cityblock Health, a value-based provider for Medicaid, dually eligible and lower-income Medicare beneficiaries, knew that managing data would be key to their success. Ankush Khona, Cityblock’s chief data officer, recently spoke with Healthcare Innovation about the company’s data strategy and infrastructure.
Cityblock came in at No. 29 on Fast Company’s list of the World’s 50 Most Innovative Companies for 2023. It has operations in Indiana; New York; Connecticut; Massachusetts; Washington, D.C.; North Carolina; and Ohio, and delivers medical care, behavioral health, and social services to individuals from historically underserved and marginalized communities, in particular those with complex needs. This care includes 24/7 access to multi-modal care delivery, with doctors and other clinical and non-clinical caregivers treating patients in their homes and community settings, via telehealth or at local Cityblock hubs.
Khona, who previously worked at Deloitte and PwC, said he was attracted to Cityblock because the leadership was very serious about data. He said that even many large organizations still don't have a chief data officer. “But here is a 5-year-old startup that really wants to bet on data as a core capability of its growth engine.”
He said he was also drawn to the company’s core mission. Although we often speak of healthcare in terms of serving populations, there were few organizations “that were really focused on serving marginalized, underserved populations,” he said. Khona also was drawn by the diversity of the company’s team.
“We are trying to solve a multi-faceted, complex problem. We are talking about a population that is underserved. We're talking about people who get onto a program, but they could drop out at any point in time,” he said. “Cityblock is a value-based care organization. We focus on behavioral health, social care, primary care. Everything requires a personalization capability, and hence you have to look at technology. Everything is driven by data.”
He said Cityblock has adopted a hybrid approach — looked at buying off-the-shelf tech solutions, partnering with certain vendors, and building their own solutions. “We recently moved from Elation to athenahealth for our common EHR system,” he said. “We also are working with Innovaccer/Redox as our partners to do processing of the data. They are our data-processing engine. They help us ingest data, standardize it into a certain format because the healthcare data comes across in different formats, different shapes, and sizes every time, so they have helped us setting up this initial processing engine. And they gave us the data in the standard format, which then we ingested into our broader data platform and that is used for downstream consumption.”
Where Cityblock is building its own tools is in data engineering, data science, and artificial intelligence models. “You can't buy those,” Khona, said. “You're not going to get what you're looking for because our business is so different. For us, that's kind of our secret sauce. The data science/AI capabilities my team is building right now will allow us to differentiate from the market and we will probably look more often to the build mode for those capabilities.”
He shared one example of what they’re trying to achieve. “We have all these medical notes that we capture in the care plans. There are conditions that we believe could be helpful for the care managers. We built an NLP engine on John Snow Labs tools to identify these conditions and then allowing care managers to become more efficient in their interactions with members.”
Another use case involves improving member engagement. “There are seven or eight use cases my teams are currently working on, at different levels of maturity, but they focus first on the outcomes, starting with what we are trying to achieve and then building it vs. just using AI and not knowing what we are using it for.”
He shared one example of what they’re trying to achieve. “We have all these medical notes that we capture in the care plans. There are conditions that we believe could be helpful for the care managers. We built an NLP engine on John Snow Labs tools to identify these conditions and then allowing care managers to become more efficient in their interactions with members.”
Another use case involves improving member engagement. “There are seven or eight use cases my teams are currently working on, at different levels of maturity, but they focus first on the outcomes, starting with what we are trying to achieve and then building it vs. just using AI and not knowing what we are using it for.”
Cityblock Health works with Medicaid managed care organizations such as MDWise in Indiana. I asked Khona if there were challenges around sharing data with managed care organizations and other value-based care partners.
He said that's a fundamental problem of the entire healthcare segment. “I don't think Cityblock or any of its partners is different in that regard. I think that's been one of the reasons why healthcare has not moved beyond a certain place,” Khona, added. “We are all aware that not every healthcare organization has spent money to upgrade their infrastructure and their data capabilities. They haven't really focused on data quality. We continue to foresee that challenge as an organization that is heavily reliant on data. However, with CMS and the ONC working on things like price transparency, the payers and providers are starting to exchange data, so we should be able to get access to some of this data that was previously not accessible, supporting the larger vision of whole-person care.”
Cityblock has created some governance processes that allows it to talk to partners on a continuous basis in order to get access to the data it needs. So, the problem still exists, Khona said, “but we also have figured out ways to resolve those issues.”
