Cicada Immunization Forecasting Engine

Cicada is a Dart immunization forecasting engine that evaluates past vaccine doses and recommends future immunizations. It implements the CDC's Clinical Decision Support for Immunization (CDSi) specification and WHO Expanded Programme on Immunization (EPI) recommendations.

What It Does

Given a patient's demographics, immunization history, and clinical conditions, Cicada:

1. Evaluates each administered dose against the immunization schedule (age validity, minimum intervals, live virus conflicts, vaccine type matching)
2. Forecasts future immunization needs with earliest, recommended, past-due, and latest dates
3. Recommends specific vaccines by vaccine group with a forecast status (Complete, Not Complete, Immune, Contraindicated, Not Recommended, Aged Out)
4. Serves forecasts over HTTP via $immds-forecast (CDC) and $immds-forecast-who (WHO) endpoints

FHIR R4 Interface

Cicada uses FHIR R4 resources for both input and output:

Input — A FHIR Parameters resource containing:

• Patient (birth date, sex)
• Immunization resources (past doses with CVX codes and administration dates)
• Condition resources (clinical conditions)
• AllergyIntolerance resources (allergies and contraindications)
• Observation, Procedure, MedicationStatement resources (additional clinical context)
• Assessment date (the date to evaluate against)

Output — A FHIR Parameters resource containing:

• ImmunizationEvaluation resources (one per evaluated dose per antigen, with valid/not valid status and reason)
• ImmunizationRecommendation resources (one per vaccine group with forecast dates and status)

Dual Schedule Support

Cicada supports two immunization schedules:

	CDC (CDSi)	WHO (EPI)
Schedule	U.S. immunization schedule	WHO global recommendations
Antigens	30+ (U.S. formulations)	22 (global formulations)
Typical vaccines	DTaP, IPV, MMR	DTP, OPV+IPV, MR, pentavalent
Dose timing	2/4/6 months	6/10/14 weeks
Spec version	CDSi v4.61-508	WHO position papers (2025)

Switch between modes at runtime with ForecastMode.cdc or ForecastMode.who.

Test Results

Cicada has been extensively validated:

• CDC Healthy Test Cases: 1010/1014 (99.6%)
• CDC Condition Test Cases: 747/777 (96.1%)
• NIST FITS (external): 167/169 (98.8%)

All remaining failures are verified as version mismatch issues between test data versions and engine version, not engine bugs.

Installation

dependencies:
  cicada:
    git:
      url: https://github.com/fhir-fli/cicada.git
      path: cicada

cd cicada && dart pub get

Source Code

• Repository: github.com/fhir-fli/cicada
• Implementation Guide: Cicada Vaccination IG