INSIGHT has nurtured a diverse research portfolio that reflects the wide range of our capabilities. Over time and through crucial partnerships we have developed the capacity to work on pragmatic clinical trials, data linkages, health systems research, predictive analytics, and social determinants of health. Our research partners have queried our database for a range of common chronic diseases as well as rare diseases. Below is a snapshot of key projects from our portfolio and the particular INSIGHT data and services that helped bring them to fruition.

Pragmatic Clinical Trials

Every year 720,000 Americans have a heart attack, and nearly 380,000 die of atherosclerotic cardiovascular disease (ASCVD). Aspirin is a mainstay therapy for these patients. Despite dozens of clinical trials, the optimal dose of aspirin – the dose that most effectively reduces ischemic events while minimizing adverse events such as gastrointestinal (GI) bleeding – has not been determined.

The ADAPTABLE trial attempts to identify this optimal dose through an INSIGHT-supported pragmatic clinical trial using data from 15,000 enrolled patients who are at a high risk for ischemic events. The INSIGHT team and participating sites facilitated the enrollment of close to 700 patients. INSIGHT used low-touch, effective engagement techniques to recruit eligible participants and provided assistance in navigating the ADAPTABLE patient portal for enrollment. 

INSIGHT’s ROLE : The study team used INSIGHT’s clinical data – along with clinical data from 9 CRNs around the country – to supplement patient-reported outcomes documented periodically in an online portal by study participants. This provided the research team with comprehensive clinical and patient-reported data from a large, nationally-representative cohort of clinical trial enrollees. 

Data Linkages

Using Predictive Models to Improve Care for Hospitalized Patients with Novel Coronavirus Disease

PI: Rainu Kaushal, M.D., MPH (Weill Cornell Medicine)

Predictive modeling has been at the center of informing clinical decision making. Using a combination of machine learning and data analytics, it can provide in-depth insight into patient outcome trajectory. With the surge of hospitalization rates for COVID-19 patients in the U.S., researchers are once again turning to modeling for support in predicting hospitalization, diagnosis, and prognosis to deliver patient-centered care and improve patient outcomes.

Working at the epicenter of the pandemic, Weill Cornell Medicine investigators are leveraging robust data infrastructure from INSIGHT Clinical Research Network to develop their models for hospitalized COVID-19 patients. By accessing comprehensive COVID-19 patient data across multiple health systems in New York City, they aim to predict three important clinical measures: the ICU need, the risk of mortality, and the course and outcome of intubation.

The project is an extension of an ongoing research project, “Identifying and Predicting Patients with Preventable High Utilization,” led by principal investigator Dr. Rainu Kaushal, senior associate dean for clinical research and chair of population health sciences at Weill Cornell Medicine. Developed initially to identify high preventable healthcare utilization among Medicare and dual-eligible patients, the project applies traditional regression methods and advanced machine learning techniques to implement prediction models in the clinical process.

Health systems, clinicians and patients have identified High-Need, High-Cost (HNHC) patients as a priority focus for population health programs. HNHC patients account for 5 percent of the nation’s population, but a staggering 50 percent of healthcare spending.

Dr. Rainu Kaushal and her team identified ten computable phenotypes to characterize HNHC patients to support intervention strategies that better serve these patients’ complex health needs.

INSIGHT’s ROLE : INSIGHT provided technical assistance to Dr. Flory and his team for his Medicare Reuse application to the Centers for Medicare and Medicaid Services (CMS) to link his clinical data to a pre-existing Medicare dataset. By leveraging INSIGHT’s data linkage services, the research team obtained a comprehensive integrated dataset containing Medicare claims and clinical data to support their patient cohort. In addition, the INSIGHT team partnered with OneFlorida to both pioneer this conceptual model across CRN’s, and to make the results replicable across the country.

In order to improve health outcomes, reduce costs, and address health disparities related to diabetes, New York State implemented a Health Homes initiative, providing shelter and care management services to over 60,000 New York City Medicaid-eligible patients.

To evaluate the efficacy of this initiative, Dr. Victoria Mayer and her team leveraged NYS Medicaid claims data and INSIGHT clinical data to identify and compare two groups of patients: one group with diabetes who are enrolled in Health Homes, and a control group of patients with diabetes who are not a part of Health Homes. Our findings enable and support health systems demonstration research as it relates to Affordable Care Act initiatives.

INSIGHT’s ROLE : Using INSIGHT’s data, Dr. Mayer and her team created a computable phenotype to identify a diabetes patient cohort and leveraged the linkage capabilities of INSIGHT to merge their clinical data with the NYS Medicaid claims dataset.

Rare Diseases

Utilization of an ICD-Coded Big Database to Characterize the Epidemiology of Prosopagnosia
PI: Christina Pressl, M.D. (The Rockefeller University)

Prosopagnosia, or face blindness, is a rare disease characterized by the inability to perceive, recognize, or memorize faces. Through an observational retrospective study, Dr. Pressl aimed to reveal new epidemiologic information about prosopagnosia, identifying primary and surrogate diagnoses to develop a computable phenotype.

INSIGHT’s ROLE : Dr. Pressl used INSIGHT’s clinical dataset to significantly increase her patient count to a staggering 129,549 undiagnosed patients, in addition to the 871 patients with a listed clinical diagnosis. The use of our data enabled the study team to gain unprecedented insight into the epidemiological characteristics of prosopagnosia.

Rare epilepsies are a devastating group of diseases that begin in childhood, and often cause profound neurologic, medical, and psychiatric disabilities. Using regular expression keywords and INSIGHT’s data repository, Dr. Grinspan sought to develop reliable epidemiological estimates of rare epilepsies.

His improved surveillance estimates have guided clinical care, prioritized research initiatives, and helped caregivers better understand these devastating diseases. 

INSIGHT’s ROLE : Utilizing INSIGHT data, Dr. Grinspan was able to increase his patient cohort 7 fold, thereby facilitating efficient collection of data across the fragmented healthcare landscape of New York City. He leveraged the results of his Pediatric Epilepsy study as well as the relationships he secured through the INSIGHT network to create momentum for the development of the Pediatric Epilepsy Learning Healthcare System. He launched this learning health system with 18 sites that are affiliated with a CRN in order to share lessons learned and insights gleaned from his initial results, his work with INSIGHT data, and the CRN infrastructure.

Medicare Reuse Data

Comparative Effectiveness of Metformin for Type 2 Diabetes with Chronic Kidney Disease
Co-PIs: James Flory, M.D. (Memorial Sloan Kettering Cancer Center) and Alvin Mushlin, M.D., Sc.M. (Weill Cornell Medicine)

Metformin is the first-line treatment for most patients with type 2 diabetes mellitus (T2DM), but its use has historically been discouraged in patients with chronic kidney disease (CKD). This research utilizes INSIGHT Network’s database to test the hypothesis that Metformin is safer and more effective than commonly-used alternatives for patients with T2DM and CKD. 

INSIGHT’s ROLE : INSIGHT provided technical assistance to Dr. Flory and his team for his Medicare Reuse application to the Centers for Medicare and Medicaid Services (CMS) to link his clinical data to a pre-existing Medicare dataset. By leveraging INSIGHT’s data linkage services, the research team obtained a comprehensive integrated dataset containing Medicare claims and clinical data to support their patient cohort.