Public Health in the Rockies 2026

Background/Purpose: Following its establishment in 2023, the Adams County Health Department (ACHD) inherited fragmented systems, siloed workflows, and inconsistent data practices that limited coordination, efficiency, and timely decision-making. To build a more resilient and connected public health infrastructure, ACHD launched a Data and Technology Strategic Plan focused on modernizing how data are collected, managed, analyzed, and used across all six divisions of the department.

Methods: ACHD conducted a department-wide assessment to inventory systems, map workflows, and identify opportunities for integration. Through collaboration among internal program staff, community partners, and leadership, ACHD implemented an Integrated Data System, enterprise GIS, standardized governance, automated data pipelines, analytics solutions, and user-centered web applications and data collection tools designed to support diverse public health functions.

Results: Since implementation began, ACHD has developed and modernized more than 1,000 web-based applications and data collection tools supporting operations, epidemiology and data science, nursing, nutrition and family health, strategic health initiatives, and environmental health. Since 2024, the department has fulfilled more than 300 data requests and expanded systems that strengthen surveillance, service coordination, referral management, field operations, reporting, program evaluation, and community health assessment. These efforts have improved operational efficiency, increased access to timely information, strengthened cross-division collaboration, and supported equity-focused, data-driven decision-making.

Implications: ACHD's experience demonstrates how data modernization can strengthen public health by connecting people, programs, and information. Building shared infrastructure, governance, and workforce capacity creates a foundation for stronger collaboration, improved service delivery, and greater organizational resilience.

60-Minute Session Justification: Additional time will support demonstrations of modernization strategies and facilitated discussion on governance, change management, workforce development, and implementation lessons. Participants will identify practical approaches for advancing data modernization and cross-program collaboration within their own organizations.

Background/Purpose:
Local health departments are increasingly expected to make rapid, data-informed decisions, yet adoption of artificial intelligence (AI) tools remains limited due to workforce training gaps and technical barriers. This study examined whether structured AI education improves public health professionals’ confidence and readiness to use no-code forecasting tools. The research addressed two questions: (1) How does AI-focused training influence knowledge and attitudes toward AI in disease forecasting? and (2) Can a no-code platform (Akkio) generate feasible county-level influenza forecasts using weather variables?

Methods:
A quantitative pre–post design was implemented with public health professionals from Douglas and Jefferson Counties, Colorado (n=8). Participants completed a 20-item Likert-scale survey assessing AI literacy, readiness, and trust before and after two structured training sessions. During the sessions, attendees applied historical influenza and weather data within the Akkio no-code AI platform to generate 90-day forecasts. Paired t-tests were used to evaluate changes in survey responses, and forecasting outputs were reviewed for practical interpretability.

Results:
Post-training scores increased across all survey domains. Familiarity with AI-driven forecasting tools increased from a mean of 1.25 (strongly disagree) to 3.75 (agree). Participants reported greater confidence in interpreting AI-generated forecasts and stronger agreement that public health professionals should receive AI training. Forecasting outputs demonstrated feasible short-term influenza projections using temperature-based variables, with higher predicted case trends observed in the 65+ population.

Implications:
No-code AI platforms can reduce technical barriers while enhancing workforce readiness for predictive surveillance. Structured AI education improves confidence and supports responsible integration of forecasting tools into routine public health practice. Expanding AI literacy initiatives may strengthen local preparedness and data-driven decision-making nationwide.

Background/Purpose: Colorado’s newborn hearing screening program works with birthing facilities, audiologists, and other providers to collect data on infant hearing screening and follow-up. The Colorado Department of Public Health and Environment (CDPHE) sends follow-up letters for infants who missed screening or did not pass and require rescreening or additional evaluation. Before this dashboard was completed, quarterly reporting required manual calculation of facility-level and statewide measures, making it time-intensive to present screening performance, follow-up outcomes, timeliness and benchmarking results to both birth facility staff and the Colorado Infant Hearing Advisory Committee (CIHAC). This project aimed to create a more user-friendly system for monitoring newborn hearing screening performance across Colorado.

Methods: As part of CDPHE’s Informatics work, we began the dashboard by writing a stored procedure, designing dashboard tabs, and improving how measures were organized and displayed. Although CIHAC initially requested detailed numeric tables, we recognized that tables alone would not provide a complete picture of statewide performance. The final reporting structure combined tables with visualizations to better communicate trends, outcomes, timeliness, and facility comparisons. This enables users to access the detailed numbers they needed while also observing clearer trends and outcomes. A map-based tool is also being developed to help families identify local hearing screening sites.

Results: The dashboard reduced reliance on manual quarterly preparation, enabled secure access to state and facility-level reports, and improved transparency around facility categorization and comparison to statewide coverages. Denver Health highlighted the dashboard’s accuracy and its value for real-time monitoring. It generated productive conversations with professionals from other states at the 2026 EHDI conference.

Implications: This work shows that user-centered dashboards can reduce reporting burden while making data more actionable. Key takeaways are that combining required tables with visualizations strengthens oversight, supports quality improvement, and improves communication with facilities, program leaders, and advisory stakeholders.

Background: Cytomegalovirus (CMV) is a ubiquitous virus that generally causes minor disease. Congenital CMV (cCMV) causes birth defects in about 20% of infected infants & is the leading non-genetic cause of pediatric hearing loss. Colorado added cCMV as a reportable condition in January 2025.




Methods: Retrospective review of cCMV cases reported to Colorado public health between January 2025-January 2026. We collected data through electronic lab reporting & provider reports, then matched records to vital statistics. We excluded out-of-state records.


Results: 60 reported cases met definition for either confirmed cCMV infection (n=54) or disease (n=6). Females were overrepresented (61.7%). The ethnic distribution of cases mirrored the Colorado population (23% Hispanic, 73% not Hispanic, and 3% unknown). Case racial distribution was 61.7% White, 20% other, 13.3% Black, 3% unknown, & 1.7% Asian.



Discussion: Ten percent of cases presented with signs visible at birth, aligning with reported U.S. rates. While race and ethnicity largely reflected DOLA 2024 population data, we observed a slight under-representation of non-Hispanic White (61.7% vs. 65%) & Asian (1.7% vs. 3.9%) infants, alongside over-representation of Black infants (13.3% vs. 4.3%). Surveillance case rates were 0.94 per 100,000 in the five-county Denver metro area, & 1.08 per 100,000 outside the metro; both figures remain significantly below the national estimate of 4.5 per 1,000 live births, in screening data.




Implications: Colorado’s early cCMV epidemiology for race, ethnicity, & rate of disease at birth is similar to national trends. Our rates overall and female preponderance are different from national estimates.These findings likely indicate a combination of variable testing practices, variation related to the first year of a newly reportable condition, & variation due to a small sample size. Regional epidemiologic differences or changing epidemiology over time may also be factors.