ImmuKnow: Automating Immunization Notice Generation

Introduction #

Childhood immunization is one of the most effective tools available to protect community health. Maintaining high vaccination coverage among school-age children reduces the risk of outbreaks, safeguards students who cannot be immunized for medical reasons, and supports the broader goal of preventing the resurgence of diseases that vaccination has brought under control. In Ontario, the Immunization of School Pupils Act (ISPA) makes this goal a legal requirement - and places public health units (PHUs) at the center of its enforcement.

For PHUs, that means regularly identifying students whose immunization records show them as overdue for required vaccines and ensuring families receive the information they need to act. The immunization notice is the primary vehicle for that outreach: a personalized, legally mandated communication that tells a family where their child stands and what is needed to bring them into compliance.

For years, producing these notices was a manual process, with each one individually prepared and mailed at considerable cost to staff time and turnaround speed. As notice volumes grew, so did the need for a better approach.

This article describes ImmuKnow, the automated pipeline our data team developed to address that challenge: a process that takes immunization data, preprocesses it, and generates fully personalized notices for over a thousand students in under five minutes.

Background #

The immunization notice is the formal mechanism through which ISPA follow-up begins. When a student is identified as overdue, a notice must be sent to their parent or guardian (or to the student, if they are over 16). This notice documents that student’s outstanding vaccines, their known immunization history, and the personal and contact information needed for the communication to reach the right household. Instructions are also provided on how to update public health with new vaccine records. Beyond its practical function, the notice initiates a defined compliance process that, if unresolved, can ultimately result in a student’s suspension from school under ISPA.

Getting notices out accurately and promptly is therefore not just an administrative task, but the starting point for every step that follows in the compliance process.

Previous Workflow #

The process of identifying and notifying overdue students begins with a query run against Panorama (provincial immunization database) to identify students whose records show them as overdue for one or more ISPA-required vaccines. This produces a list of students requiring follow-up, along with the immunization data needed to communicate with them.

Panorama includes a built‑in notification letter function that can generate PDF notices for these students. However, it lacks support for key customizations, such as languages, tailored disease lists, or more dynamic content.

As a result, what followed was still labour‑intensive. Staff would modify these generated PDFs individually using PDF editing tools to ensure each notice accurately reflected the student’s immunization history and required messaging. In some cases, PHUs implemented mail merge processes to streamline this work, but these approaches were limited in their ability to incorporate more complex elements, such as detailed immunization history tables. Completed notices were then printed and mailed to families.

While this process reliably produced compliant notices, it did not scale well. Although Panorama’s built-in letter generation and, in some cases, mail merge approaches provided partial automation, there was still a need for manual customization. The volume of overdue students each cycle meant significant staff time was dedicated to notice production, and turnaround time from identification to mailing was constrained by these manual steps.

New Automated Workflow #

The data identification step remains unchanged - a query against the provincial database continues to be the starting point, producing the overdue student list and associated immunization records that drive the rest of the process.

What is new is the automated pipeline that handles the rest:

Preprocessing. The output of the query becomes the input to the ImmuKnow pipeline, where the first preprocessing stage cleans and standardizes the data before it is used to generate notices. This involves normalizing fields and validating records to ensure that downstream notice generation runs against consistent, reliable input.

Notice generation. The preprocessed data is passed into the next pipeline stage that generates a fully personalized immunization notice for each overdue student. Each notice contains the student’s immunization history, their list of overdue diseases, and the personal information required to route the letter to the correct household - all drawn directly from the data and assembled automatically. The pipeline produces over a thousand notices in under five minutes.

Personalized QR codes. New this year for WDGPH, each notice includes a QR code unique to that student. When scanned, it directs the family to a vaccine record submission portal - and automatically pre-fills demographic information that was already included in the notice. This reduces friction for families reporting vaccines their child has received and improves the quality of data returned to the unit.

Generated Immunization Notice #

Each notice is generated based on a Typst template. This template contains the overall structure and styling of a notice, such as the format of the vaccination history table and the placement of the PHU logo. Student data is used to populate template fields with their individual demographic and immunization information.

Below is an example immunization notice from WDG. The first page contains the list of overdue vaccines with guidance on next steps, including a personalized QR code directly to a vaccine submission portal (the example QR code does not lead to this portal). On the second page is an immunization history table and privacy statement.

Impact #

This automated pipeline has meaningfully improved turnaround time. Notices that once required painstaking individual preparation can now be produced in bulk within minutes of the identification query being run - helping ensure families receive timely communication and the compliance process can proceed without unnecessary delay.

Automation also brings consistency. Because every notice is generated from the same validated, preprocessed data, the risk of fields being missed or incorrectly entered is reduced compared to manual assembly. With such high volumes involved, that reliability matters.

This is primarily a communications quality improvement, and one that still requires staff involvement. Quality checks are in place, and the data team continues to run and oversee the process, with the pipeline itself receiving ongoing improvements. What has changed is the speed and quality of what gets sent and (with the addition of personalized QR codes) the ease with which families can respond.

External Pilots #

WDG has been fortunate to pilot ImmuKnow’s automated notice generation process with several other Ontario PHUs. These pilots have been an invaluable opportunity not only to validate the pipeline in different operational contexts, but also to learn from the ways these PHUs implement ISPA processes.

Across participating units, variations in data structures, notice requirements, and compliance workflows highlighted both the flexibility needed in the pipeline and the shared challenges faced across the sector. Feedback from partners directly informed improvements to all pipeline stages, helping ensure that ImmuKnow can adapt to varying local needs while maintaining consistency in output.

Developing shared custom report templates for PEAR helped to standardize core data outputs while still supporting local needs. This approach alleviated many of the data structure challenges and laid the groundwork for supporting more advanced features that are currently under development.

We are grateful to our pilot partners for their time and openness in sharing their processes. This work lays the foundation for a more mature, shareable version of ImmuKnow, with the long-term goal of enabling wider adoption and ongoing collaboration across PHUs.

Conclusion #

The automated immunization notice pipeline represents a meaningful step forward in how our PHU fulfills its obligations under ISPA. By replacing a manual, notice-by-notice process with an automated one capable of quickly producing thousands of personalized communications, we have improved the speed and consistency of a workflow that directly supports school immunization compliance.

The addition of personalized QR codes further demonstrates what is possible when notice generation is treated not just as a document production task, but as an opportunity to reduce barriers for families. Pre-filling demographic information in the REDCap portal is a small change from the recipient’s perspective, but it is the kind of thoughtful design that makes compliance easier and complete vaccine record submission more likely.

More broadly, this work reflects a growing recognition within public health that technological innovation has a role to play in strengthening core program delivery. Immunization follow-up is not a new challenge, but the tools available to address it are evolving. Pipelines like this one show that investment in data infrastructure and automation can translate directly into better public health communications - without compromising the accuracy or legal integrity that notices require.

Part 2 of this series will examine the next step in that evolution: moving from mailing immunization notices to delivering them by email, and what that transition has meant for reach, response, and the compliance process as a whole.