When you visit the doctor’s office or hospital, you’ll likely be asked about your race, ethnicity, or language preference (REL). While the questions may seem routine, your answers play an important role in identifying health care disparities and improving the care you receive.
In Connecticut, the state’s health information exchange is called Connie. It is a secure way for providers to share health information electronically, so they can access it quickly, and across systems. All health care providers who are connected to Connie are required to ask patients for their race, ethnicity, and language preference (patients are not required to provide it). Providers can use this information to see if all patients are receiving the highest-quality care and if not, if there are patterns that need to be addressed.
Collecting the data is just the first step. For the information to be useful, it has to be complete and consistent.
In 2025, the Connecticut Health Foundation awarded Connie a $100,000 grant to assess and improve the collection of REL data by health care organizations across Connecticut.
“The funding from the Connecticut Health Foundation allowed Connie to invest in a data quality platform that allows us to not only look at REL data compared to state standards, but also to look at the wide range of data in general and assess quality,” said Russ Dexter, director of analytics and data quality for Connie.
Connecticut’s strengths
The tool found there are areas in which Connecticut does exceptionally well at collecting this data.
Race and language data is relayed to Connie nearly universally. Language data collection in particular stands out as one of the state’s strengths.
Knowing this information can help health systems better serve their patients. Research shows that language barriers in a doctor’s office or hospital room can lead to miscommunications that can impact the quality of health care and patient safety.
What the numbers tell us
The tool looked at data from 35 hospitals (this figure includes acute care hospitals and some rehabilitation and hospice facilities) in Connecticut and examined more than 1 million electronic messages that hospitals use to communicate information about patients, including their demographic and REL information.
It found that of those patient records:
- 98.2% included race data
- 73.7% included ethnicity data
- 84.8% included language data
It also looked at what data was in alignment with Connecticut standards and found that:
- 51.5% of race data was fully compliant
- 9.7% of ethnicity data was fully compliant
- 83.4% of language data was fully compliant
Though some of the numbers may seem low, Dexter said it’s not because hospitals aren’t collecting the data. Rather, most instances of non-compliance were because of inconsistencies with wording.
“What we found a lot of times in the data is that the hospitals were collecting the data that intended to comply with Connecticut standards, but it wasn’t exactly the values that Connecticut wanted,” Dexter said.
For example, some hospitals listed ‘white or Caucasian’ as a category, while Connecticut’s standards list the category as ‘white’. The question was being asked, but the small difference made it appear non-compliant.
Why data quality matters
Collecting good REL data doesn’t improve someone’s health on its own. But it can help reveal problems with care that may otherwise not be found.
Better data may sound like a technical goal, but in practice, it means making sure our state and health care providers have the information they need to understand where disparities are occurring and what can be done to address them.
Knowing that the data is accurate and consistent across the board is a good first step to making that possible.
The Connecticut Health Foundation supports a network of health care providers who are working on these issues to convene monthly and has awarded grants to some participants. To learn more about the network, click here.
This blog post is based off of a presentation at one of the REL monthly convenings.