If you’re just joining us, this is the third article of a four-part series covering the new HUD SNAPS initiative strategies. We’re discussing what it looks like for communities to achieve an advanced level of implementation of said strategies.
What does it look like for communities to gather data that will improve how they use their HMIS? And what are the characteristics that HUD is looking for? The table below offers a brief overview of the desired state of each required characteristic.
While individual CoCs know best how to bring together the homeless providers and non-homeless providers in their particular community, an HMIS vendor should know best how to use software to facilitate a safe, secure shared data environment. Communities need accurate data in order to maximize their HMIS.
Data Sharing in your HMIS should be straightforward, robust and reliable — allowing for integration between systems and organizations while maintaining steadfast compliance to sharing and privacy obligations. A combination of system-, project-, user-, and field-level controls is what it takes to empower system administrators to tailor access to the minimum amount necessary to get the job done. Here are some controls that we recommend looking for in an HMIS:
To get the maximum benefit from HMIS, it must generate quality data. HUD defines three hallmarks of data quality as they relate to HMIS software:
Your data must be accurate and comprehensive if you expect to have faith in your system. A big step in this direction is to enter information as soon as possible. The longer it takes to add your data, the more likely you are to misremember, forget, or lose it — even if you wrote it on a paper form.
There’s a growing need to provide meaningful access to HMIS and functionality in the field. Unfortunately, many legacy information systems do not translate well to use on a tablet or phone. Look for an HMIS that is mobile friendly from the get-go. The easier your HMIS is to use from a mobile device, the more quickly your data can be added and the better you’ll maintain its accuracy. Mobile functionality is particularly important as communities turn to data systems as the gateway to Coordinated Entry and other vital services.
Another important step towards improving data quality is to make sure you’re entering the correct information into your HMIS. Once communities identify their required (or desired) data elements, system administrators should have the power to designate the related fields as “required.” This type of data safeguard means that before a user can submit a record, they have to answer all indicated fields.
In contrast to required fields are irrelevant fields. If your HMIS prompts you to answer questions that are irrelevant to a client, it’s more likely that you will accidentally enter an incorrect value. If this happens too much, your data quality will deteriorate.
System administrators need to be able to reduce the likelihood of accidental data entry by using field constraints to control whether or not a specific field is displayed on a user’s screen. For example, if a client is documented as female, then a pregnancy status will automatically appear, but only for females. And if the pregnancy field is set to “yes,” then a prompt to enter the client’s term due-date will appear.
When using Clarity, for example, system administrators have access to all of these features and they can also configure display logic to hide or show specific fields based on a predefined workflow or how a user answers questions. This feature streamlines data entry and reduces the chance a client account will feature irrelevant information.
According to SNAPS, when it comes to quality data, “comprehensive” means that the data standards reflect elements of client experience that need to be known. In other words: your CoC is collecting the information that your CoC needs to know in order to resolve homelessness in your community. System administrators and agency managers need to be empowered with the tools required to customize and build upon the federal HMIS Data Standards to meet the requirements of virtually any funding source.
This requires custom fields — and a lot of them. In fact, it calls for a customizable HMIS software in general. Look for an HMIS that you can personalize to meet your needs. What are some of the tools you need to have in a customizable HMIS? Here are our recommendations:
Of course, no matter how cautiously or comprehensively users enter data, mistakes will happen. Your HMIS should feature various tools like canned reports and ad-hoc data visualization capabilities that make it easy to identify and correct such errors.
Canned reports are an excellent resource for tracking your data quality. In your HMIS report library, it’s a great asset to have canned reports that cover:
Ideally, all canned reports will allow for drill-down functionality, allowing users to view reports that identify missing/don’t know/refused responses on HMIS fields.
Look for an HMIS that enables you to do ad-hoc reporting on your data quality and its performance for all HMIS data elements. Custom data queries lead to a more thorough perspective on you current state of affairs and the progress you have made.
The Data Quality Model, which is available using the Clarity Human Services Data Analysis Tool, provides easy-to-use calculations of data quality performance for all HMIS data elements. This model immediately reflects changes in Clarity Human Services; any corrections made to the data will be automatically updated in Looker in real-time. Bitfocus’ communities use these tools daily with striking results. We believe that the ultimate purpose of HMIS is to empower organizations to make evidence-based decisions to fight homelessness. See the below example of how it helped one Bitfocus community clean up their data.
Recently, one CoC working with Bitfocus system administrators launched an initiative to better track community-wide outcomes. Using Clarity’s data quality reports and built-in data analysis tools, the team examined various metrics to identify opportunities for improvement.
“It quickly became clear where data quality was lacking,” says one system administrator. “Looking at, for example, utilization on a program-by-program level, we noticed that some shelters appeared to be partially utilized — maybe only 20 percent full. That’s an alarming statistic to publish to the community! In reality, of course, these shelters were full or nearly full every night.”
Further investigation showed that the problem was due to user error. “Sometimes users weren’t entering all the necessary data; some weren’t exiting clients or enrolling them or tracking their bed nights in the HMIS,” says the administrator.
The CoC used Clarity’s data analysis tools to identify shelters with artificially low occupancy rates. Then, the shelters addressed the problem by coaching HMIS users on the best practices of data entry, emphasizing timeliness and comprehensiveness. The team also changed the requirement settings on specific fields so that they had to be answered before a record was submitted.
Many organizations enhanced their training plans. According to the system administration team, focusing on data literacy was especially important: “Data literacy takes users beyond simply understanding what’s required and shows them what the requirements mean, why they’re required, and what the impact of good data is on the system. Most importantly, it emphasizes how to use data and benefit from it.”
Utilization rates for these shelters are now more accurate, showing close to full occupancy. Stakeholders are much happier, too.
As this example shows, having accurate and comprehensive data entered in a timely fashion should be a priority for anyone who uses HMIS. These two factors will make a significant impact on your reporting quality, and ultimately, your community’s funding, too.