Carolyn Greene: The Data Miner
Epidemiologist Carolyn Greene aims to use electronic health records to track chronic disease trends.
Denise Bosco for Proto
A typical electronic health record (EHR) includes places to enter hundreds of pieces of data, ranging from vital signs to allergies to radiology images to immunization dates. Keeping track of all of that information for hundreds of thousands of patients for several years generates millions of points of data that, taken together, might yield important findings. That’s the idea behind the NYC Macroscope, one of the nation’s first EHR surveillance systems. Carolyn Greene, co-principal investigator of the initiative and deputy commissioner of the New York City Department of Health and Mental Hygiene’s Division of Epidemiology, hopes this system and others like it will provide information that can bridge the gap between public health and clinical medicine.
Q: How did you choose which patients and conditions to track with the NYC Macroscope?
A: We chose those health indicators of chronic conditions that we believe we could influence through public health initiatives and that are likely to be tracked in EHRs, such as obesity, diabetes, hypertension and high cholesterol, as well as health indicators such as smoking status and flu vaccination. Our pilot study will include about 760,000 patients in primary care practices all over the city. In 2015, we plan to release a report with our first year’s data and a discussion of the lessons we’ve learned.
Q: How might health officials use EHR data to improve their response to public health needs?
A: We know, for example, from earlier studies that a large number of adults with diabetes weren’t being diagnosed. We could tell from glucose test results that they had diabetes, but when they were interviewed, they didn’t report having that diagnosis. We found similar results for cholesterol, with nearly 30% of those with high cholesterol unaware of their condition. With an EHR surveillance system, we can look at whether people are being appropriately tested for a condition based on their risk factors. If they’re not, we can give feedback to providers, make recommendations about who should be tested, and conduct public education campaigns.
Q: What are the advantages of using EHRs to monitor the health of large populations?
A: EHR systems collect vast amounts of information on large numbers of patients. This means that in addition to tracking the prevalence of diseases or risk factors within the population as a whole, we’ll be able to track how well we’re treating and controlling chronic diseases over time and by characteristics such as age, sex and neighborhood poverty level. If you already have an EHR infrastructure in place, you’re already collecting data, so using a system like this for surveillance is potentially less resource-intensive and costly than traditional forms of surveillance such as telephone surveys and disease registries. There is also the potential to get data much more quickly. When we send a digital query to all of the primary care practices in our system—for example, if we asked how many men age 60 and older who visited a practice in 2012 had a body mass index of 30 or greater—within 24 hours, we would have our answer: numbers (without identifying information) returned to a secure, centralized data site.
Q: Will this system mean you no longer need other surveillance tools?
A: EHRs should not serve as the only population health surveillance tool. In New York City, we conduct a telephone survey every year called the Community Health Survey that includes questions about sexual behavior and alcohol and drug use. I think it will be a long time before that kind of behavioral health information is consistently recorded in EHRs. The data from EHRs will be invaluable, but we will need to supplement it to get the complete picture of the population’s health.