BACKGROUND: The Colorado Immunization Information System (CIIS) is a system created to improve patient care. Customized local analyses of population-based vaccine trends have been limited. Assessing vaccine dose validity is challenging and CIIS is unable to remove patients who have left the state or moved to a non-reporting practice, necessitating exclusion of inactive records. A system was built to produce population-based immunization coverage reports by integrating CIIS with open-source software to meet the needs of Denver metropolitan counties working to improve human papillomavirus (HPV) vaccine rates.
METHODS: CIIS data were loaded into a data warehouse which retains time-stamped transactions, including patient record updates and the date patients moved in or out of participating counties. An application was built to allow users to choose an evaluation date and patients and immunizations in the warehouse as of that date. The application includes an option to exclude patients with no CIIS activity five or ten years prior to the evaluation date. Batches are processed through the Immunization Calculation Engine (ICE) software (HLN Consulting, CA) which assesses patients’ age and vaccine doses as of the evaluation date and returns a list of valid/invalid doses and vaccines needed presently or in the future. In the final phase, ICE and CIIS data were joined to calculate patients’ age, address, and medical home as of the evaluation date. Interactive dashboards display vaccine coverage estimates by area of residence, age group, gender, medical home, number of valid doses per antigen, and recommended vaccines.
RESULTS: As of November 2016, the data warehouse contained 2,277,105 patients (1,950,113 of whom were active in the past ten years) and 22,247,399 vaccines. The largest ICE batch to date (2,194,320 patients) processed in 29 hours. Among 11-12 year olds, an estimated 63% had not received any valid HPV vaccine doses. Local pediatrician practices have been recruited to strengthen their vaccination efforts and receive practice-level reports to direct these efforts. Five practices have received reports. At one, HPV vaccine initiation among 11-12 year old patients increased from 22% in June to 29% in October 2016.
CONCLUSIONS: This novel application of ICE made it possible to estimate population-based vaccination rates based on evaluation dates. Challenges of CIIS data (e.g., voluntary provider participation and difficulty differentiating inactive and vaccine-refusing patients) may influence these estimates. Immunization information system managers and analysts may want to consider such a system to facilitate population-based assessments and targeted outreach efforts.