ClientConnect 2023: ICF and the Principles of Data Governance

kdavidson114@gmail.com

ClientConnect 2023, our inaugural peer-to-peer conference for human services organizations, was a resounding success! But if you missed out, don’t let that stop you from catching the next one. In the follow-up to this article, we’ll explore the ClientConnect presentation from ICF and their take on data governance.

ICF is a global consulting firm with many independent teams that provide consulting, technical assistance, and more for the intersection of social and health services. At ClientConnect 2023, their representatives impressed upon our partner communities the importance of building a solid foundation for data governance, especially when it comes to homelessness with other service provisions.

icf-cc-logos.jpg

ICF’s senior director of Homeless Services, Mike Lindsay, introduced us to ICF as a prime HUD SNAPS TA provider for homeless management information systems (HMIS), and he went on to explain the importance of data governance: Data governance can be both an asset and a pain, as we all know.

“The majority of communities we work in that have strong governance structures in place, whether that’s from their CoC or HMIS, have larger impact on their communities’ and their homeless response system,” says Mike. “They can do things more efficiently without wasting time, going back and forth, rethinking decisions, or lacking clarity on how decisions can be made.”

The Importance of Data Governance in Social Services

Not having a proper governance structure can lead to unforeseen crises. In ICF’s experience, inadequate or nonexistent data governance structures have resulted in:

  • lack of vision or clarity around priorities,
  • loss of investments,
  • delays in advancements and strategic planning,
  • setbacks in data quality,
  • unclear decision-making authority,
  • organizational frustration, and
  • the inability to be agile.

Mike further explains, “As TA providers, we’re often asked to come in to communities because they are dealing with a situation or problem . . . We often hear, ‘APRs [annual performance reports] can’t be produced; therefore, this community needs a new HMIS lead,’ or, ‘APRs can’t be produced; therefore, this community needs a new HMIS vendor.’“ These situations, he says, are often symptoms of a larger problem—a lack of clear data governance.

importance-in-social-services.jpg

Data Governance to the Rescue

Mike cites the City of Philadelphia as an example of overcoming challenges with data governance. At first, their HMIS system was overly customized, and all users struggled to engage with it. They were confronted with a hard choice between two options. They could:

  • power through with their current HMIS software, which would require extra resources and training on an already-complex system, or
  • create an RFP for a new system.

After about 5 or 6 years, they eventually decided to switch vendors and adopt ClientTrack by Eccovia for their HMIS, but the RFP process on Philadelphia’s end was a long road with many barriers. It was later discovered that lack of governance, primarily, led to the consternation they experienced with decision making.

ICF worked with Philadelphia on its software system, but the processes of governance became the main focus, because without a solid governance structure, it was unclear who was supposed to do what and when, when it came to the HMIS.

good-vs-bad-image.jpg

The Life Cycle of Data Governance

Structure, process, performance, and policy are the main four pillars or data governance, Mike emphasizes.

  1. A formal governance structure creates the foundation for strong, successful data utilization within a community (i.e., how decisions are made, who has authority to start an RFP).
  2. Clearly defined process upholds data governance and how priorities are set. For example, ICF has worked with many HMIS leads who are unsure of their annual or semiannual priorities, and the CoC leads sometimes conflict with those HMIS leads because they have a differing vision of those priorities.
  3. Monitored performance with transparency and clear accountability creates more thorough communication.
  4. Clearly written—and implemented—policies help identify roles and responsibilities, as well as set expectations.
The-Life-Cycle-of-Data-Governance-Graphic.png

Who Should Be Involved in Data Governance Decisions

Alissa Parrish, senior manager in ICF’s Homeless Services division, makes clear that the governance structure required will depend on the data governance needed, whether it be HMIS, data warehousing, or community information systems: “One thing we’re seeing is that communities really are looking beyond HMIS, now looking at other data that helps inform their homeless response system, because we know the people we’re serving through those systems are not only experiencing homelessness: they’re accessing other services too, so we need to serve them holistically and effectively.”

As we bring data together for more effective community care coordination, data leaders and community leaders ought to include people with lived experience in the process.

data-governance-who-is-involved.jpg

MOUs, SOWs, and Charters: Keys to Success

Memorandums of understanding (MOUs), statements of work (SOWs), and charters are really key to success. “They’re not ‘cool,’ they’re not fun, and they’re not sexy, but they’re really important to write down. Once you have them established, they become the framework and foundation from which you do all your other work,” says Alissa.

She tells the story of a community that ICF engaged with that decided to go through a vendor transition without going through any community process or RFP, seemingly making the decision overnight. A couple months into the transition, this community realized they were asking questions of the vendor they should’ve asked much earlier in the process—questions that had no clear decision maker:

  • “How customized is your new system?”
  • “How much data will you be able to bring over?”
  • “How are you going to prioritize which custom reports you’re going to rebuild in the new system?”

This community still hadn’t completed this vendor transition at the time of this presentation.

So what are the main keys to MOUs and SOWs and charters?

  • Write everything down.
  • Don’t assume anything.
  • Make what’s written down clear, actionable, and specific.
  • Make sure everyone shares the same definitions of terms.
  • Align expectations among all stakeholders.

Data Governance Structures

Every community is really different, so there’s no one-size-fits-all structure to prescribe. For example, when it comes to data sharing, you should ask yourself a few questions:

  • In what ways is data shared, and through what platforms?
  • Who is responsible for what?
  • What data sharing is necessary, with whom, to what degree, and why?

Alissa recounted an experience as an HMIS lead for a community where there was a new permanent supportive housing (PSH) project recently built, serving about 47 people. They needed to show the value of that housing project so funder, partners, and community leaders could see the impact in terms of dollars. Multiple entities had a stake in seeing this project succeed, so the HMIS needed to show how access or usage of other emergency services had shifted once access to housing was provided, for a specific period of time (i.e., 6-month, 1-year, 5-year reviews).

Many of these community partners included emergency services, crisis services, the county jail, and a few other entities. They wanted to be able to access the housing project and HMIS data to determine whether costs had changed over time. By answering the questions listed above, the HMIS team was able to provide client data to stakeholders who then provided their cost data, which was imported into the HMIS platform so comprehensive reports could be run at will. Doing so helped them prove the value of the PSH project.

Data Governance Processes and Communication Strategy

Data priorities need to be set and communicated across the community, and this can be a complex topic, because every stakeholder’s priorities can be different. How you speak to end users about data quality is going to differ from how you talk to funders, so you need to know what’s important to each audience. That means you need a communication strategy.

In other words, you need to know your audience. A direct service provider will care more about data quality because it impacts the way in which clients are served (i.e., eligibility, coordinated entry), but a funder will care about population data and outcomes, as well as data quality.

So when establishing processes, communities should ask themselves:

  • How will data priorities be set and communicated?
  • Who will communicate those priorities to the managing entity of the data, to the public, to elected officials, and others?
  • How often does the governing body meet, and who’s part of that governing body?
  • Who will review documents created to support the data environment?
Data-Sharing-Case-Study-Graphic.png

Monitoring Data and Service Performance

Performance is best supported by accountability with transparent monitoring and communication, but that requires shared understanding of the roles and responsibilities of each entity in the CoC. When Alissa served as an HMIS lead, the CoC was experiencing turnover in its leadership. Her team came in to help them with reporting and compliance, but there was little clarity around roles and responsibilities when it came to decision-making power: “We were getting into scenarios where there wasn’t anyone on the CoC side to make the decisions, because of the turnover they were going through, and by not making a decision, they were inadvertently making worse decisions.”

A strong data governance structure should be transparent, inclusive, and ensure accountability for all entities involved. The structure ought to establish, approve, and implement monitoring processes for the managing entity of the data environment. That means you have to monitor against an MOU with clearly established expectations, with a work plan, and so on.

Data sharing doesn’t need to be a four-letter word. It is sometimes stigmatized within communities because of privacy and compliance concerns, but Alissa reminded us that people experiencing homelessness will often provide different data to various programs, like Medicaid, housing assistance, and so on. We need to match and share that data, in a secure way, between providers so clients can benefit from true care coordination.

Communities should ask themselves:

  • What accountability structures are in place to ensure data is shared appropriately with authorized entities and used for approved purposes?
  • How are data-sharing platforms monitored to ensure privacy and security standards are met, and who conducts that monitoring?
  • What processes are in place to address a data breach or misuse of data?

Creating Data Policy

Clearly-written procedures should identify roles, responsibilities, and expectations, Alissa explains: “Nobody is a mind-reader, so we need to agree on what the different entities are doing on behalf of the community as a whole.”

The CoC governing body needs to act as a liaison between key entities: for example, the creation of an HMIS advisory committee, a data committee of the CoC, and so on.

There should be pre-established roles and responsibilities too, to outline who has the authority to develop, review, and approve new policies and procedures. That creates a feedback loop to integrate policy feedback from the field, including people with lived experience, to see whether the policy met the needs of the community.

Communities should ask themselves:

  • Who is acting as the liaison between the different entities sharing data with each other? What is that liaison’s specific roles and responsibilities? As you incorporate more data from other sources outside of HMIS, figuring out the liaison might change. It could be the local government, even.
  • How are different entities supported in ensuring they are upholding their specific roles and responsibilities? This is related to the monitoring concept from before.
  • How are decisions made about what data is shared, how it is shared, and for what purposes? Some communities are incorporating predictive analytics, where they are attempting to prevent homelessness for people deemed at risk.
  • How is feedback solicited to update and refine policies and processes over time?
concluding-thoughts.jpg

Concluding Thoughts

Data governance can be frustrating and difficult, but that’s often because it’s not established up front. Communities need to build trust within their CoCs, which in turn will help them have shared understanding of what needs to happen to improve service duration, increase outputs and better outcomes, and improve utilization of community resources.

“The overarching theme I’ve noticed in communities, when it comes to governance, is it often has to do a lot with broken relationships and broken trust,” Alissa says.

So, to recap the main data-governance best practices:

  • Have a formalized governance structure
  • Make sure there is a unified decision-making body
  • Be transparent about potential conflicts of interest in the decision-making structure
  • Establish single policies and procedures
  • Ensure ongoing monitoring and evaluation
  • Incorporate transparent, diverse input on the development of policies
  • Increase your capacity to leverage data for decision making
  • Acquire financial support for data systems

ICF provides consulting, technical assistance, and much more for health and human service organizations. We at Eccovia are proud to associate with them and encourage HMIS and CoC leads to contact them for further information on how they can help improve data governance processes.

ClientConnect is our peer-to-peer, educational event series. If you are interested in learning more about how Eccovia’s solutions can help your community coordinate care more effectively, please schedule a demo. If you would like to pre-register for ClientConnect 2024, please do so here.

More Topics

In many communities, individuals who are chronically homeless—many of whom are Medicaid beneficiaries—are being served by coordination between community providers of healthcare, housing, and housing assistance. However, the “revolving-door” effect …

In the United States, Medicaid expansion is now being adopted by 41 states, which unlocks a new funding stream for social service organizations in whole person care. In this article, …

It’s often said that the plural of anecdotes is not data. Perhaps a more nuanced take on the phrase could be: conclusions should not be drawn from anecdotes alone. Because, it turns out, anecdotes (i.e., self-reported individual data) are the primary kind of data we receive at client intake, whether for HMIS, domestic violence shelter, refugee services, etc. So there’s a sense in which it’s possible to derive useful data from multiple anecdotes.

Contact Us

BE IN THE KNOW

Subscribe to our monthly newsletter for health and human service organizations. You'll learn from our experts and partners on best practices, industry news, and never-before-seen resources!