Spotlight on a User: Improving Agency Data Quality
In 2019, several agencies were awarded Data Quality Incentives funds from the HMIS Governing Board to help them address their agency’s data quality challenges and create lasting improvements. In this edition of Spotlight on a User, Sarah Broich, Single Adult Program Manager at Simpson Housing Services, shares her insights from the Adult Shelter Connect’s data quality improvement project and provides advice for other agencies that are looking to improve their data quality.
What does your agency do and how do you use HMIS?
Adult Shelter Connect is one program of Simpson Housing Services, which operates on behalf of the single adult shelter collaborative. The ASC is the centralized access point for single adult shelter in Hennepin County. We use HMIS daily to see available space at the six Single Adult Shelters in Hennepin County and connect individuals in need of shelter to those open spots. We conduct the initial intake for individuals new to the shelter system and issue a community card which is used to make shelter reservations. Being responsible for those initial intakes means ASC staff create hundreds of new HMIS profiles each year. The data collected by the ASC is used by the community to assess the needs and identify gaps in the system. The data drives decisions about priorities and improvements needed.
What data quality challenges has your agency had?
Being a new program in the community inherently comes with challenges, so there were learning curves in getting a new team up to speed. The ASC works with a very high volume of individuals coming in crisis, which can make data quality details difficult. Focusing on handling crises and de-escalating situations often takes priority. There are many staff involved in running the ASC, which operates days and evenings 365 days a year, so there are lots of different staff learning not only HMIS, but the details and nuances of each individual shelter. It’s a lot to learn and mistakes are inevitable. We also have different protocols in the evening, because we only operate over the phone. Our policy during that time is to create a partial profile when we have someone new to shelter contact us. We then ask them to come in-person the following day to complete their full intake. One thing we learned through our data quality project is that many of those folks do not come back to complete the intake, making it impossible to complete their HMIS information.
What were you hoping to achieve with your Data Quality Incentives project?
Our goal was to improve our data completeness in the new HMIS profiles we create at the ASC. When we started the project, our team data completeness average was 87.52% with a range of 78%-98%. We set a goal of achieving 95% as a team average. We used the HMIS “ASC New Clients by User” report to measure our progress. The goal was also to get everyone on the team a solid understanding of our data expectations, so that new profiles going forward are as complete as possible from the start and fewer corrections are needed. We were not able to achieve our goal of 95%, ending our project with a team average of 89% and improving our range of scores to having everyone on the team above 80%. However, a valuable outcome of this endeavor was gaining a thorough understanding of why we’re not able achieve 95%. While the documented improvements to our data may seem minimal, the ASC has made dramatic improvements in data quality since we became aware of the issue near the end of 2018.
How did you use the Data Quality Incentives funds?
We used the funds to do a team dinner at a restaurant that the team voted on. The team members shared they enjoyed the time with each other, as it’s a rare treat to be able to gather as a team.
Have you continued to make improvements to your data quality processes since the Data Quality Incentives project?
We have continued to pull data quality reports monthly, and each team member has worked on their own corrections. We have identified a need for an instruction manual that ASC staff can easily reference, and are working to put that together.
What strategies have helped your agency improve your data quality?
One new piece we’ve implemented is that the ASC manager pulls the New Client by User report monthly, works with staff to make all possible corrections, and follows up quickly with any team member struggling with their data completeness. We also put together clear, written instructions for expectations of partial profiles for staff to reference. We’ve set clearer boundaries around when creating a partial profile is necessary, which has resulted in fewer being created overall. The ASC manager also reminds staff often to do careful searches before creating a new profile, to prevent duplicate profiles.
What strategies have you used to support your team in improving their data quality? Do you have any recommendations for supervisors that want to help their staff improve data quality?
Monitor data frequently to catch issues early. Checking in with staff 1:1 is important to address individual questions. It’s also really important for staff to understand the reasons behind the data we’re being asked to collect and what it is used for. The full ASC team meets quarterly, and we’re trying to incorporate this into our team meetings, especially sharing examples of successes and improvements that have resulted from our data collection. I encourage agency supervisors to solicit input from staff doing the data entry. Ask for feedback from staff often and use their ideas!