As we move deeper into a digital world, health and human services leaders have more and more data at their fingertips. When this data can be shared across agencies in real-time and is structured in a user-friendly, objective, and informative way, it can play a critical role in seeding breakthrough innovations. In particular, health and human services leaders can identify new opportunities to improve the client experience, reduce costs, enable population health management, enhance care, provide greater equity, and focus on interventions for clients who touch multiple health and human services systems.
However, sharing data and applying predictive analytics comes with challenges. Leaders must establish governance structures; incorporate new technologies; and navigate a maze of legislation, regulations, and relationships at different levels of government. They also have to ensure that they have fiscal support, manage cultural change, and assuage concerns about and protect against encroachments on privacy.
To help leaders navigate this challenging process, The 2016 Health and Human Services Summit featured a session that spotlighted two agencies that have made dramatic progress in leveraging data sharing and predictive analytics. In the Allegheny County (PA) Department of Human Services, Marc Cherna has led a 20-year effort to create an integrated data system and use predictive analytics to improve decision making and practice. Similarly, in the Montgomery County (MD) Department of Health and Human Services, Uma Ahluwalia has spearheaded an initiative to create an integrated case management system while addressing questions about confidentiality laws and regulations. This section details these organizations’ efforts to leverage data and analytics and identifies valuable takeaways for other leaders hoping to replicate their progress.
In 1997, when Marc Cherna became the Director of the Allegheny County Department of Human Services (DHS), he took on a situation that was both extremely challenging and ripe with opportunity. The prior year, in response to the community’s desire to modernize and streamline county government, and amid concerns over the quality of its human services offerings, county officials had decided to merge several formerly discrete departments and create DHS. Thus, Cherna faced the challenge of leading a brand-new mega department. As daunting as that was, he saw the opportunity to create an agency that combined integrated services, data, and predictive analytics to reach the highest levels of the Human Services Value Curve. “Creating this synergy from scratch,” Cherna said, “was a tremendous opportunity.”
Cherna began by creating a foundation for reform. This included engaging in a community-driven process that established a vision for an “integrated human services system” with a focus on the utilization of cutting-edge technologies. He recruited staff with diverse perspectives and skill sets to help the agency realize this vision, and sought input from local residents and other stakeholders about what the agency should prioritize. This was critical, Cherna emphasized, because it helped DHS to gauge public sentiment and served as a reminder that “we work for and are responsible to the people we serve.”
With this foundation in place, Cherna and his staff began to initiate reform. An early priority was creating a data warehouse that would integrate DHS data; the vision was to add data sets from as wide a range of related people-serving systems as possible. To accomplish this, Cherna sought advice from the head of the Greater Pittsburgh Chamber of Commerce, who was part of a “kitchen cabinet” of advisors that he had recruited. That official connected Cherna with a team of Chief Information Officers from local financial firms who, on a pro bono basis, helped DHS write a Request for Proposals for the data warehouse. In part because of the way in which the community had been involved in creating the vision, local foundations provided extensive financial support for the creation of the data warehouse. What’s more, the community’s stake in the data warehouse made diverse stakeholders—including universities and health and human services providers—enthusiastic about sharing data. “It’s a community asset, paid for by the community,” Cherna explained, “so everybody gets to benefit from it.”
DHS next focused on developing technologies that could help them to leverage the integrated data. This included establishing “a 360-degree client” view, developing a coordinated intake mechanism, and moving toward a more-uniform Information Technology (IT) system for case management. Building systems to enhance decision making was another critical part of the reform because, while other innovations provide valuable data, there is no guarantee that staff will know how to interpret that information to improve decision-making. By contrast, as Erin Dalton, Deputy Director for the Office of Data Analysis, Research and Evaluation, explained, “Predictive analytics can provide a score, which analyzes and summarizes data in a way that is easier for the workforce to interpret.”
DHS issued a competitive solicitation to enhance the delivery of services to its clients by using data to improve decision making, and received proposals from around the country and around the world. A group led by a Principal Investigator from Auckland University of Technology, partnering with the University of Southern California, was selected. This research partnership stood out because of their prior implementation experience (actually delivering a model to the field) and because of their shared interest, with DHS, in the ethics of predictive modeling. The research team supported Allegheny County’s efforts to improve child welfare front-line decision making by modeling the likelihood of an adverse event in the two years following a child welfare referral. Historic data showed there was room for improvement: an analysis revealed that DHS was screening out 27 percent of the highest-risk cases and screening in 48 percent of the lowest-risk cases.
Last August, DHS implemented a predictive analytics tool that uses information collected by DHS and other partners to inform screening decisions. A Family Screening Score is calculated by integrating and analyzing hundreds of data elements on each person related to the referral: it predicts the long-term likelihood of re-referral, if the referral is screened out without an investigation; or home removal, if the referral is screened in for investigation. With the exception of a mandatory screen-in threshold (which requires that the referral be investigated), the score provides additional information to assist in call screening decisions and is not intended to replace clinical judgement or to be used in making investigative or other child welfare decisions.
Independent process and impact evaluations of the predictive analytics tool are underway. Yet while it is too soon to know the tool’s impact, it has reinforced a 20-year trend in which Cherna and his staff have pushed boundaries and sought to leverage technology to help people in need. To that end, they are exploring the possibility of creating an analytics tool to prevent child abuse and neglect. More broadly, they are adhering to a mantra that Cherna has embraced since the start of his tenure. “You’ve got to really get out there and not be afraid,” he said. “Do the right thing, and the rest kind of takes care of itself.”
In 2007, when Uma Ahluwalia became the Director of the Montgomery County Department of Health and Human Services (DHHS), she, too, faced the challenge and opportunity of leading an organization in transition.
For DHHS, the journey of reinvention and modernization had begun in 1994 when four departments had been merged into a single agency (DHHS) with the objective of producing “integrated, coordinated, and comprehensive service delivery.” Prior to Ahluwalia’s arrival, the different departments had formally integrated, collocated their services, and created a single administrative structure. Nonetheless, DHHS had not made significant headway on coordinating treatments and services. This was a problem, Ahluwalia explained, because many of DHHS’s programs served the same clients but were not in dialogue about how best to support them.
Early in her tenure, Ahluwalia therefore strove to produce more effective coordination across DHHS’s programs. This included establishing a uniform intake form for all services, which created a “no wrong door approach” and made it easier to determine if a client requires multiple services. DHHS also launched a single client database and record. This has helped to create more detailed information about who the department is serving and “promoted information sharing for service integration.”
Still, the most important of these coordination efforts involved clarifying DHHS’s standing under the Health Insurance Portability and Accountability Act (HIPAA), a 1996 law that established “national standards to protect individuals’ medical records and other personal health information....” This was critical because, if cross-agency information sharing violated HIPAA, different programs would not be able to share data and coordinate their efforts. After an extensive legal analysis, DHHS determined that the entire agency was covered by HIPAA. This meant that even social service and income support programs could share information with one another. What’s more, DHHS has established an extensive infrastructure and trainings to ensure compliance with all privacy and confidentiality regulations. This includes a six-week annual workshop in which DHHS brings together officials from different disciplines, who are accustomed to not sharing data (often because of industry standards and norms), and emphasizes the benefits of collaboration.
Since laying this foundation and enhancing coordination, DHHS has continued to leverage technology to fulfill its vision of creating an “integrated and interoperable HHS enterprise.” This includes creating an electronic content management system; electronic health records; and, most importantly, an electronic integrated case management system, which was launched in late January 2017. The hope is that this can help DHHS improve efficiency and outcomes, particularly for the 20 percent of cases that account for 80 percent of agency resources. “They’re a finite set of tools that we’re going to offer up to our staff,” said Ahluwalia, “but those tools are going to be very powerful, and we’re hopeful that [they] will change the way we do our work.”
With DHHS having just launched the integrated case management tool, it is too soon to quantify its impact. Nonetheless, ten years into her tenure, Ahluwalia can point to critical takeaways from the reform process. These include the importance of training, understanding confidentiality laws, and “asking the right questions.” Above all, she emphasizes taking an incremental approach to reform. “Start small and resolve issues,” she said. “Then expand.” This is a helpful reminder that scaling the Human Services Value Curve is a multi-decade progress, but if an organization sustains its vision, learns, and grows, it can make dramatic progress, just as DHHS has.