Photo: FG Trade/Getty Images
A lot of strides have been made enabling semantic interoperability across care settings over the past 15 years or so, but one area where data exchange remains challenging is behavioral health.
New York State's public mental healthcare system was one place where this challenge – lack of easy access to patient data, where and when it was needed – was hindering care delivery and harming the patient and provider experience.
"There was a fragmentation of care across New York State," said Nitin Gupta of the state's Office of Mental Health. "We worked with 22 different teams across New York State, and each of those teams were collecting data in their own EMRs – some were standard EMRs, some were non-standard EMRs."
Gupta, director of PSYCKES, a HIPAA-compliant, web-based app to support clinical decision-making, care coordination and quality improvement across New York State, is scheduled to speak next week at the 2026 HIMSS Global Health Conference & Exhibition in Las Vegas.
Alongside his colleague, Vincent Grande, project director at the New York State Office of Mental Health, Gupta will discuss some of the challenges the statewide network was facing with regard to basic data access across behavioral healthcare providers.
They'll also describe how New York OMH launched a program designed to improve critical time intervention, focused on connecting vulnerable unhoused individuals with recovery services and helping them transition from the street to stable housing.
Such an approach is just as complex as it sounds, and it depends on multidisciplinary care teams that comprise licensed clinicians, care managers, peer specialists and registered nurses – with IT maturity and data governance capabilities varying widely.
"Out of 22 different CTI teams, we worked with six or seven different EMRs," said Gupta. "Some of them had a high level of technological advancement, and some were really basic – where they could just give us CSV files extracted from their EMRs, and then we had to merge that data in. It was a varying landscape."
Critical time intervention is a care management and coordination service designed for adults with serious mental illness. As explained by New York State OMH, the approach "is used during a 'critical time' of transition in their lives. CTI Teams coordinate the services and supports that help people return to their community."
"When a client or patient with behavioral health issues is in a critical condition – planning to harm themselves or others – the time between providing that person a behavioral health service and having them stable in stable housing is very important," said Gupta.
As he explains it, CTI includes four phases – requiring care teams to have special skillsets – nonjudgmental, person-centered, culturally sensitive, strength-based – that can meet the complex and unique needs of these vulnerable patients.
Such care approaches also depend on robust data collection for timely, informed clinical decision-making.
"We did a lot of diagnosis-based interventions based on a patient's history," said Gupta. "There was a lot of evidence-based practice involved – creating a holistic profile of a patient as to exactly which services have helped them in the past and how to connect them to those services now.
"When we started the program, there were roughly 130 critical encounters per month per team across New York State. Within six months, we reduced that to 119 or 120 – about a 10% improvement in six months of data collection."
To accomplish that, New York OMH had developed and implemented a hybrid semantic interoperability framework, powered by HL7's FHIR standard, to structure flat-file formats, enabling cross-system data exchange across providers.
The framework, which also incorporates terminology mapping (SNOMED CT, ICD-10, Gravity Project), metadata tagging and consent controls, was a significant undertaking, as Gupta and Grande will discuss in Las Vegas.
"We had to clean up a lot of data," said Gupta. "We had to apply a lot of standardized formats to that data. We created a staging environment and then converted that data into FHIR data elements so that we could move data across multiple agencies.
"Some of the agencies and providers we worked with were already working on standard FHIR-based models, so we were able to map our FHIR data elements across their data elements and bring that data to an API," he added. "That was a good learning lesson – how to tailor your own data elements for standard and non-standard protocols across multiple platforms and EMRs."
At HIMSS26, the public health leaders say they hope attendees at their season will learn more about cleansing techniques for ingesting complex datasets from various clinical sources to help ensure data coherence for interoperability.
They'll offer practical insights into using standards such as FHIR and non-standard application programming interfaces to facilitate data ingestion and integration, and will discuss strategies for more effectively engaging and onboarding stakeholders for data exchange efforts.
Thanks to those efforts, New York State behavioral health professionals are now able to deliver better care for vulnerable patients in crisis, and "triage their healthcare needs much more efficiently and effectively," said Gupta.
"We can refer patients to lower-cost services instead of directly setting them up with higher-cost services. It also creates efficiency in the system – people who need high-end services get them, based on priority rather than a social worker's incomplete understanding of the patient's situation."
OMH's data-intensive efforts on critical time intervention have been beneficial for patients, providers and the public health system at large.
"New York State invested a billion dollars in mental health and behavioral health in the past year as a budgetary enhancement to the already existing system," said Gupta.
Now there's a plan to "expand beyond our initial 21 CTI teams to 47, because of the success of the program," he said, noting that the next priority is to reach people in more rural communities.
Gupta and Grande's presentation, "Unlocking the Power of Semantic Interoperability Across Complex Healthcare Settings," is scheduled for Thursday, March 12, from 11:45 a.m.-12:15 p.m. in Level 5, Palazzo M at HIMSS26.
Mike Miliard is executive editor of Healthcare IT News
Email the writer: mmiliard@himss.org
Healthcare IT News is a HIMSS publication.
