Data Improving Lives
Data does not make meaning, people do, and there is absolutely no use integrating data if they will not be used to drive decisions.
On a recent national survey of data sharing efforts, respondents commonly reported the following impacts:
- increased capacity for evidence-based policy-making
- improved ability to evaluate what works
- greater capacity to respond to federal and foundation grants
- cost savings and efficiencies
- improved relationships and trust across agencies
- increased public engagement in conversations about data and programs
Examples in Action
AISP helps sites start small for quick wins and steadily build a culture of data use and cross-agency learning to drive policy and improve lives long term. Below, we describe six use cases for integrated data that successfully built on existing momentum, established relationships, and strong governance to advance cross-agency policy and practice.
Early Childhood Iowa uses its integrated data system to better understand early childhood service utilization and the early childhood workforce. Data sources included public health, education, and human services data. Initial efforts focused on determining an unduplicated count of children in care from birth to age 5 across the state, and found that 73% of children had at least one center-based experience during the year before kindergarten entry. Importantly, the analysis revealed significant gaps for vulnerable children, particularly those in rural counties. Analysis also found shortages in both the quantity and quality of the early childhood workforce, with staffing challenges being particularly acute in rural counties, which comprise 89% of Iowa counties. This project was an important precursor to receiving a Preschool Development Grade, Birth-5 grant in 2019.
The Housing Advisory Board of Charlotte-Mecklenburg, with support from Mecklenburg County Support Services, used the county’s integrated data system to better understand families experiencing housing instability and homelessness. Analysis found a disconnect between students identified by schools as experiencing homelessness (using McKinney Vento records) and children and youth identified as literally homeless in an emergency shelter, transitional housing facility, or unsheltered location (using Housing Management Information Systems [HMIS] records). Some individuals were not identified as experiencing homelessness by the school system, but had, in fact, experienced homelessness in an emergency shelter. This discrepancy was communicated to the county, local providers, and the school district; as a result, additional social workers were placed within the emergency shelter system to identify children for and connect them to services.
The Miami-Dade IDEAS Consortium for Children used their integrated data system to map aggregate outcomes of early childhood education that better inform decision-making at local agencies, including resiliency mapping by census tract to identify areas of persistent need and areas where children are outperforming socioeconomic expectations. The Consortium determined that while 83% of children entering kindergarten had a preschool experience, countywide, there are significant opportunities to increase access to high-quality preschool programs, as only 31% of preschool programs are licensed. Using preschool attendance data and K-12 data, analyses also found that children who consistently attended preschool scored higher on math and reading assessments in preschool and in kindergarten, especially for children living in census tracts with higher concentrated disadvantage. These analyses have supported work to improve early childhood program attendance.
Caliornia’s Health and Human Services Agency (CHHS) partnered with the USC Children’s Data Network (CDN) to conduct a record reconciliation that linked and organized administrative, individual-level client records across eight major CHHS programs (31+ million records prior to deduplication) and generated an encrypted master client identifier for interagency use. This record reconciliation supports everyday operations as well as the development of longitudinal, cross-sector evaluation and research that facilitates a holistic view of client experiences. These efforts have spurred the development of a secure, cloud-based research enclave for hosting record-level research data sets and accompanying linkage keys. Once operational, this environment will provide carefully controlled, role-based access to analysts within CHHS.
In the longer term, the goal is to develop protocols that, with necessary approvals, will give external university-based and other research partners access to curated data sets and statistical resources within this analytic environment. It is anticipated that this secure platform will advance rigorous evaluation, improve the reproducibility of research, create efficiencies in data management, and further the engagement of university-based researchers with government. Additionally, this research data hub will enhance record security and client confidentiality through data access and security protocols that can be more carefully audited.
The City of Philadelphia, in partnership with the Penn Child Research Center at the University of Pennsylvania, used its CARES integrated data system to inform how Philadelphia could best leverage revenue from its newly established soda tax to expand universal pre-k offerings. This is an example of how data integrated at the individual level can be aggregated and mapped geographically to drive resource allocation to neighborhoods and families with complex needs. The city was also able to build on this initial success and execute a separate legal agreement allowing trusted practitioners access to information about families who might benefit from the new pre-K slots so that those most in need could be connected directly to services.
After the passage of Chapter 55 legislation in Massachusetts, multiple state agencies collaborated to use integrated data to better understand trends in opioid-related overdose and death. Their efforts resulted in a more holistic view of those affected by the public health crisis and improved their ability to target resources and interventions, leading to promising reductions in opioid-related deaths in 2019.
COVID-19 Research & Response
Secure data sharing for the public good is more urgent now than ever. State and local governments and their non-profit and university partners are working hard to adapt and respond to the evolving crisis, and existing data sharing capacity gives some jurisdictions a major advantage.
AISP Network sites are supporting COVID-19 research and response by:
- Creating public data visualizations and tools to communicate with the community about the virus and related program and policy changes
- Maintaining dashboards to track cases, capacity and resource use, and inform policy-making
- Forecasting impacts on Medicaid, agencies, and state budgets
- Studying patterns in population health and disparate impact of the virus, including by race and ethnicity
- Identifying individuals and neighborhoods most at risk for hospitalization and negative outcomes and deploying interventions
Based on recent Network conversations, we’ve started an informal spreadsheet to collect and share ideas among members, as well as to provide links to local efforts that might be replicated. We hope you’ll contribute! For additional resources related to COVID-19 and data sharing, visit https://www.aisp.upenn.edu/covid-19-data-sharing-for-research-response/