Public Health – Seattle & King County (PHSKC) serves 2.2 million King County (KC) residents, including people in 39 cities. The county spans urban, suburban and rural areas, and has a diverse population and social and economic landscape. In population health assessments, KC often performs better than the US average, but that hides stark disparities by race, place, and income. The COVID-19 response is no exception; cases, hospitalizations and deaths are occurring disproportionately in communities of color. 

COVID-19 community mitigation measures have economic and social impacts. PHSKC is examining these impacts over time, by race/ethnicity, geography, age, gender, LGBTQ, occupation, and disability status where data allow. These indicators, in conjunction with an overlay of the shifting policy landscape at the federal, state, and local level related to COVID-19 mitigation measures, can inform communities and policymakers’ response planning efforts by providing useful information to help them address impacts and/or adjust community mitigation strategies. Using some traditional and non-traditional public health data sources, this monitoring and evaluation framework can be replicated by other jurisdictions. 

Since April, we implemented an online dashboard updated weekly, monthly or quarterly (depending on the data source), worked with existing and new partners (see section 2 below) to disseminate relevant, high-quality and novel data, released infographics and technical briefs to assist in interpretation, and are receiving more than 500 website hits weekly (ranging from 500 to 1,500). 

Unemployment rose drastically following the implementation of stay-at-home-measures; since March 1st, over 554,000 KC residents (over 1 in 4 workers) filed claims for unemployment benefits, tripling unemployment.  Employees in the accommodation and food services, health care and social assistance, and retail trade industries were most heavily affected, as were workers who identified as American Indian/Alaska Native, Native Hawaiian/Pacific Islander, or Black. Housing and food-related needs increased significantly, with food insufficiency almost doubling since March and remaining high. Over 67,000 (7.6%) households did not have internet access, meaning approximately 21,200 children lacked a computer in the household or broadband internet access, which are essential for remote schooling. It is estimated that the percent of working age adults without health insurance have doubled, and enrollment among working age adults increased by 12%. Disparities continue in mental and behavioral health, domestic violence, and child abuse/neglect as well. 

Factors that led to success include existing partnerships, the ability to pivot staff to this project, and existing data sharing agreement, data analysis, and data visualization infrastructure and skills. 

Indicators for this project work to reduce health disparities by considering needs of vulnerable communities (e.g. must be income-eligible for Medicaid) and all data are disaggregated to the extent possible by race/ethnicity, age, place, gender identity, sexual orientation, disability status, occupation, educational attainment. 

Public health decisionmakers need timely, reliable data in the midst of the pandemic. These monitoring and evaluation data have been used to inform the development of an eviction prevention program, inform recovery planning, guide decision making in distribution of millions in food assistance funding and inform the county’s budget decision making process.

Problem: Community mitigation measures such as stay-at-home orders or business closure are effective at reducing the spread of COVID-19 in the community, but can have economic, social, or health impacts on individuals. There is the potential for inequities to increase based on who closes, what services or resources are available, and whether an individual is able to stay-at-home. Policymakers and other decisionmakers need high quality, timely information to make evidence-based decisions on how to mitigate those economic and social impacts and reduce disparities.

Innovation: This innovative monitoring and evaluation framework is a creative and responsive practice that involves pivoting to new, timely, relevant data sources, new dissemination channels, and a frequent update of data combined with policy to inform decisions during the COVID-19 pandemic. The interactive data dashboard and briefs are tailored to follow policy changes around mitigation strategies and have improved on the timeliness and relevance of the data. These are new and innovative practices to improve the context and understanding of the many impacts and the disparities that are being experienced across the population.

This population-level surveillance tool monitors and evaluates the impacts of the COVID-19 non-pharmaceutical interventions across economic, social, and non-COVID-19 health impacts. Examples of topics include unemployment, food security, transportation and mobility data, computer/internet availability, mental and behavioral health, drug overdose, suicide, and family violence.  To address inequities that might be occurring, we disaggregated the data by race/ethnicity, age, place, gender, disability status, and occupation. Not all data collected contained all demographic characteristics of interest; in those cases, we often partnered with community-based organizations or the emergency response’s Community Mitigation Team to include qualitative stories and context.

In the past, most of our population level surveillance used administrative files that have been fully processed and finalized but lag by 1-2 years. For example, our Community Health Needs Assessment/Community Health Indicators project has tracked over 160 indicators across birth, death, hospitalization, communicable disease, phone and school surveys, but the 2021 report relies upon data from 2018-2019. However, to be responsive to the immediate changes that were occurring based on policy implementations, we needed to pivot to more real-time information. This involved using syndromic surveillance data to monitor for behavioral health, physical health, and maltreatment; provisional death files to examine the impact on deaths; developing new partnerships with other data providers such as the Emergency Medical Services, the Prosecuting Attorney’s Office, our Department of Community and Human Services, and community organizations. in addition, the data are a mix of new data sources not routinely used by local public health (such as Safegraph mobility data, measures of access to technology, or unemployment data).

Indicators were chosen to be

• Timely and valid—consistently available on a regular basis (daily, weekly, or monthly) with no more than a month lag in reporting; high quality data
• Responsive-- likely to change quickly during and after implementation of strategies to slow the spread of COVID-19
• Representative—relevant to a topic area and directly linked to community mitigation efforts
• Disaggregated -- able to be disaggregated by COVID-19 risk group, race, place, gender, health and socioeconomic status to support equity analyses.

Adaptation: We adapted our Tableau dashboard experience to reflect a new landing page and to allow people to delve more deeply into the data. Our main landing page provides a broad overview of the major topics; each tile leads to a specific dashboard with more detailed information. We also partnered with our communications department and subject matter experts to release blogs, infographics, and technical brief reports to disseminate the information. Existing staff pivoted to this project, re-tooling analytic frameworks, checklists for analysis, data processing and data visualization to make the process unfold more smoothly. We leveraged existing relationships for data sharing agreements (DSAs); processes of DSAs; and were able to develop new relationships and DSA processes.

Local context: The first identified COVID-19 case in the US came through the international airport located within the County, and the first known US death due to COVID-19 occurred in our county. COVID infections and deaths have been higher in the Black, Indigenous, and People of Color (BIPOC) community and South King County. Mitigation measures put into place by the Governor in March 2020 correlate immediately with rising unemployment and related to that, food security. Even though unemployment has decreased since the peak, food insecurity remains high. This project’s food insecurity data was used for County decisions on placement and disbursement of millions in resources. Cities have used the metrics to adapt policies around financial security, inform recovery planning and make decisions around eviction moratoria. These data have also been used to inform the county’s budget decision making process.  In addition, many people are concerned about a potential increase in suicide attempts and ideation; following the Substance Abuse and Mental Health Services Administration (SAMHSA) pandemic response curve, the anticipated trajectory of increases would occur delay from the onset of the pandemic. By working to research the patterns in mental health, including mental health call line information, use of the healthcare system and emergency response for ideations and attempts, we hope to provide advance notice of any worsening situation so messaging and resources can be deployed.

Evidence base:  While COVID-19 and mitigation measures are new, the practice is evidence-based, as it was designed to capture relevant data points suggested from previous pandemic and disaster recovery literature. It also follows the CDC monitoring and evaluation framework for COVID-19 and includes suggestions from the Community Guide to Preventive Services about building evidence from analysis to action. 

Planning and Implementation: This is a collaboration between the Centers for Disease Control (CDC) Incident Management Team (IMT), the Washington State Department of Health (DOH) and PHSKC. Other county partners that have become included in this process include the Department of Community and Human Services (DCHS), who provide behavioral health and recovery treatment, and data on homelessness and affordable housing; the Prosecuting Attorney’s Office (PAO), and the county emergency response. We also have collaborated with the statewide Department of Transportation (DOT) for traffic patterns, the Employment Security Department (ESD) for unemployment data, our internal Emergency Medical System (EMS) for tracking calls, the ESSENCE platform through DOH for the syndromic surveillance and Rapid Health Information Network (RHINO) program. We get regular data feeds from the Department of Social and Health Services, the Washington Poison Center, the Health Care Authority, and Washington State’s 211 to look at changing social needs. This project works closely with a community advisory group to review findings.

The planning process involved developing the framework, the indicators to be measured, and performing literature reviews about pandemic recovery. Community was not involved in the research background as the opportunity evolved quickly, but the project, its goals, and the indicators to be measured were shared with the Pandemic and Racism Community Advisory Board (PARCAG), a local group that included members of community, non-profits, community based organizations, community allies, healthcare, business, academia, and other subject matter experts. While the project inception was mostly well-received, prior to project kick-off, we adapted the project and indicators based on their feedback to include other data elements or sources that were important to the group. 

In the implementation process, this project has worked with both PARCAG, subject matter experts, and community to review findings prior to releasing reports to address gaps that the community may identify. This project always disaggregates data by demographic characteristics to include issues that we have heard from community are of concern. Disaggregated data by disability status is one concern that has been raised by community and we disaggregate wherever possible; since many of the data sources are external to us, we don’t have sway on what information is collected. Community also asked for more digestible forms of data and so the team created infographics and adjusted some dashboards to make them more user-friendly.

Strategic partnerships: In addition to community advisory group review, this project relied upon strategic partnerships with the countywide Equity and Social Justice Office review to ensure health equity is considered at every step of the process. In co-developing a local survey around needs with the WA DOH, we brought the survey questions and instrument to the large community guidance group and included questions suggested by the group that were areas of interest. While it was a statewide survey, we paid for translation to ensure the survey was available in 6 languages and incentives for completion. During the review process with community, particularly in places where quantitative data or disaggregated data are missing, we try to include community stories that speak to the impact in the brief or blog.

Some unexpected partnerships came through the Employment Security Department with sharing unemployment data, the National Domestic Violence Hotline, and the Prosecuting Attorney’s Office. Our state DOH released weekly provisional death data that allow us to analyze impacts on deaths; prior to the pandemic, quarterly provisional data was available but rarely used. All of these new partnerships allow for a broader look at the social, economic, and overall health measures rather than only having a piece of information, which is important as mental, physical, and economic health are all intertwined. Due to these new partnerships, we have begun to see other new collaborations as off-shoots of these initial partnerships for similar but non-COVID related work that are expected to be sustained after the COVID evaluation has completed.

Adapt and respond: This project has worked closely with the community in implementation and dissemination. As briefs and analyses were presented to community advisory group and subject matter experts prior to release, we asked community if the quantitative results were matching their lived experience. We were able to re-tool briefs to include some elements that may have been left out, and on occasion, go back to external data partners to ask for different data elements to include. We have been able to iterate and improve upon data visualizations and share out with community on a regular basis. In order to reduce inequitable health outcomes, we have presented data by an extensive list of demographic characteristics to meet the desires of the community. 

Resources: This project built upon existing staff, data sharing agreements, and software infrastructure, which were adapted and innovated. Additional partnerships have led to novel data being shared in the project, including several non-traditional public health datasets.  We have been able to maintain and update the dashboard while many staff have been deployed to help with the direct COVID-19 response activities and case/contact tracing.

Lessons learned: In order to limit challenges, we developed our timelines for briefs, created checklists to help with analyses and data visualizations, identified key points of contact for each data as well as back-up analysts, implemented data visualization standards so that users can easily switch from topic to topic and still be able to follow the charts, and worked to automate data analysis and visualizations to reduce update times.  These automations include programming interface (API) pulls as well as internal statistical processing code that make data updates less prone to error. There are also manual and automated quality assurance steps included in the checklists and updates.

Recognizing that disaster recovery work continues long after the disaster (or pandemic) has subsided, this project is designed for a minimum of 2 years from April 2020. We will continue to maintain updates on our existing data sources and work with the community for interpretation, feedback on content, and dashboards. It’s expected that some of the regular data feeds will continue beyond the 2-year period as they also provide important context for the social determinants of health even outside of a pandemic response. We also have already leveraged one of the relationships to develop a project that will to look at the impact of employment on housing. As we share updates with data providers, the relationships are strengthened and they can use the health outcome data for education around the importance of their data as it relates to health and health outcomes.

The public health impact has been seen through the use of the data and dashboards by community to obtain CARES dollars, by cities and county to distribute and plan for resources and recovery, by cities to measure the impact on their populations and to evaluate their policy changes; by community mitigation groups to discuss what’s going on in the community and for elevating issues of high visibility to community when the data is available. Food resources, CARES Act dollars, and community educators have been deployed to areas and populations that have been shown at high risk or who have large disparities. For example, these data helped inform a new County food program distribute funds specifically to organizations that could provide culturally relevant foods and in the most impacted areas of the county. These monitoring and evaluation data have also been used to inform the development of an eviction prevention program, inform recovery planning, and inform the county’s budget decision making process.  This monitoring and evaluation framework can be replicated by other jurisdictions.