Eva is a web-based data analysis tool created by HMIS technical assistance groups that work closely with HUD. This tool can be used to assess the accuracy and completeness of the data collected within an HMIS.
As Housing Alliance Delaware continues to follow updates made to CMIS by our vendor, we have come to a conclusion that Eva is a more beneficial tool overall to review data quality than many of the built-in reports made available by WellSky. Therefore, the Delaware Continuum of Care and the CMIS Program will be introducting Eva to CMIS end users and transition some of our processes to incorporate Eva more into the work that we do.
This article provides more information about Eva, a replica of Eva training provided by Housing Alliance Delaware, and answers to questions received by CMIS users about Eva.
WHY USE EVA?
There are four (4) major reasons that we recommend transitioning to Eva for data quality review as opposed to reports previously used, like CMIS's Data Quality Framework:
1. The software will be updated by TA providers that work closely with HUD. Eva should automatically respond to updates provided by HUD relating to the HMIS Data Standards, and will actually be the quickest of sources to adapt to these updates. This will be better than relying on the DQF, which has a delay as we wait for WellSky to updated affected reports; or a BObj report, which would have a delay as Housing Alliance Delaware updates those reports.
2. Eva does a better job identifying what about a specific data issue/error needs to be changed. Eva has a more detailed list of errors that are returned for each client, and includes a glossary that allows users to look up the specific remedy for each type of data issue/error.
3. Eva prioritizes types of errors by breaking errors into three categories – High Priority error, Standard Error, and Warning. This is useful for situations where users have a limited amount of time and need to prioritize fixing only the most impactful errors.
4. Eva includes a wider variety of possible error types, allowing for a more detailed analysis of data quality problems.
DATA SECURITY WARNING.
While Eva utilizes a report from CMIS (specifically, an Hashed HMIS CSV) that does not export all data from CMIS, a Hashed CSV is still considered to have Personally Identifiable Information (PII).
Names and Social Security Numbers are not explicitly provided in a Hashed CSV, reducing the risk of sharing client data with people who should not have it, but there are still examples (especially in a smaller CoC like ours) where a client can still be identified.
COMMUNITY EXAMPLE.
You are assisting a client who identified their race as Asian and provided their full Date of Birth. They also reported having a 10-year old daughter.
Being in a small CoC, these details are enough to uniquely identify someone within all clients in CMIS - an Asian client born on MM/DD/YYYY with a 10-year old daughter.
Since Race, Date of Birth, Gender, and Household information are all included in the Hashed HMIS CSV, someone may be able to figure out exactly who the client is based on the codes/numbers in the report.
WHAT DOES THIS MEAN WHEN USING EVA?
It is important to be vigilant when using reports that contain large amounts of PII. Therefore, Housing Alliance Delaware requires that:
Hashed HMIS CSVs must be sent through email using an encrypted email.
Hashed HMIS CSVs cannot be shared with those who are not authorized CMIS end users.
Hashed HMIS CSVs cannotbe uploaded to any website except for HUD's Eva tool.
TRAINING VIDEO.
The video below is a replication of the training provided to CMIS end users on January 9, 2025. In this training, Housing Alliance Delaware covers:
Running a Hashed HMIS CSV
How to upload a Hashed HMIS CSV to Eva
How to review and assess data quality
How to read a downloaded errors report
A brief introduction to the System Performance module of Eva
To make the video larger, use the "Full Screen" option in the bottom-right corner of the video, or click the "Watch on YouTube" button to be redirected to the video in another window.
FAQs.
Q: Who can acess Eva?
A: CMIS end users are the only people who should be reviewing Eva, as a Hashed HMIS CSV with PII is needed to use Eva. If an Agency Administrator who oversees CMIS end users need to access Eva, they must request an Hashed HMIS CSV that will be sent to them through encrypted email.
Those who do not work with CMIS data cannot access this software. Should anyone want/need to review the System Performance module of Eva, they must contact Housing Alliance Delaware.
Q: Can I run a Hashed HMIS CSV for any of my projects?
A: A Hashed HMIS CSV can be run for any specific project or a group of projects that an end user has access to, via Reporting Group. Some older/closed out projects may lead to the Hashed HMIS CSV not running, due to HMIS Data Standards of the past not aligning with the reporting requirements today. However, current projects and all projects moving forward should allow for a CSV to run as long as the Provider was set up correctly.
Q: How do I make sure my data is removed from Eva?
A: Eva will only access your data during your session on Eva, no data is being retained or viewed by anyone besides the person on Eva. That means that inactivity (of 20 or 30 minutes) or closing your web brower with Eva open will lead to your data not being "saved" on the site for the next time you revisit it.
RESOURCES.
Eva - access the web-based data quality tool using this link
HUD Exchange | Eva - access HUD's main webpage about Eva, which includes additional Resources
Link to Eva Training on Youtube - access Housing Alliance Delaware's Eva Training video if the embedded video above does not work
INTRODUCTION The Data Quality Framework (DQF) is a canned report available in CMIS. The Data Quality Framework provides details on HUD-required data elements and those clients whose data relating to those data elements are flagged as data quality ...
This article will provide information on how to complete entry assessments and give detail on ways that data quality issues could appear and be addressed. ENTRY ASSESSMENTS: HUD Data Elements The following assessment questions must be answered for ...
This article is to provide guidance to Delaware CMIS users to accurately and consistently enter data relating to clients’ length of time homeless (LOT), according to and in compliance with the HUD Data Standards Manual. The attached document ...
INTRODUCTION This article's purpose is to provide guidance to DE-CMIS users to utilize the Household Data Sharing capabilities that are available in the system. Adding Household Data will allow for data relating to client's CoC Enrollment, location, ...
See attached for the most recent version of the Clients Module Data Entry PowerPoint. This PowerPoint provides step-by-step directions on how to enter data using the Clients Module, which is used for the following Project Types: Emergency Shelter ...