The EHR or other clinical software system has the ability to re-evaluate and update the immunization forecast using a patient’s newly updated immunization history, where the updated forecast results from the reconciliation of immunization data contained in the public health immunization registry with immunization data contained in the EHR. Processing the new forecast can be internal to the EHR or it can use an external forecasting service.
The EHR record for Juana Maria Gonzales, age 6, includes an inactivated polio vaccine that was administered in Dr. Smith’s office 3 months prior to the date of Juana’s office visit. Due to an EHR system upgrade, the polio vaccine was not reported to the public health registry history. Therefore, the original forecast received from the registry had recommended a fourth dose of inactivated polio vaccine. The EHR system sends the updated vaccine history to a third-party forecaster and receives a new forecast that does not include the recommendation for a fourth dose of inactivated polio vaccine, as it was already given.
The American Immunization Registries of Association defines immunization forecasting as “the result of the process of applying rules to determine dates for the next vaccine dose(s) to be administered to a Patient. Vaccination Forecast may also include a reason why the next dose in a series is or is not recommended and recommended Trade Name for the next vaccine in a series.[i]
Discussion about immunization forecasting includes two areas of concern. First is addressing the accuracy of forecasting logic; i.e., that the recommendation is correct. Second is the provider’s view of the forecast presentation.
- Immunization Forecasting Logic. The logic used to develop a vaccine forecast originates with recommendations about how to use vaccines to control diseases in the United States from the Advisory Committee on Immunization Practices, an Advisory Committee to the CDC Director.[ii] CDC publishes the recommendations regularly as public health advice.[iii] CDC also provides the Clinical Decision Support for Immunization (CDSi) logic specification as an authoritative, implementation-neutral foundation for technical and non-technical immunization-related clinical decision support.[iv] CDSi includes business rules logic, test cases, supportive data, workflow descriptions, and describes methods to determine if the vaccine doses a patient received are appropriate (valid) when compared to the ACIP recommended schedule. Based on the logic, a clinical decision support engine can recommend the earliest and latest acceptable dates for providing each vaccine as well as the appropriate intervals between individual vaccines. The clinical decision support engine can also indicate if individual doses already given are not valid because they were given ahead of the prescribed vaccination schedule. A clinical decision support engine that incorporates CDSi content requires a mechanism to capture patient information and send results to a clinician for review and reconciliation with all known patient information. Some EHR vendors indicate they have started to evaluate use of CDSi content within CDS engines in their software. Many vendors express a preference for an external service to provide such decision support for their products. The requirement is purposely not prescriptive about how the software should process the forecasting logic. The post-reconciliation forecast may be processed by logic provided by the vendor, by accessing an external web service, or by requesting a new forecast from the registry (IIS) after the EHR sends an immunization history update to the registry. Refer to capability 2.6 for discussion about historical updates submitted to registries.
- Immunization Forecast Presentation. This capability addresses display of an immunization forecast after reconciliation of the EHR data with updated historical information from the registry or from other sources, including the patient. The intent of the forecast is to help providers make real-time decisions. Display of both the pre- and post-reconciliation forecasts may cause confusion and affect system usability. Since usability issues can inhibit the system’s effectiveness vendors should consider forecasting an important area to address using UCD processes. Attachment D provides an example of a model UCD process addressing vaccine forecasting.
- Vendor Perspective. Accurate and timely vaccine recommendations are essential to address efficient, effective, and safe use of clinical software. Usability is integral to encourage appropriate interpretation and use of the information provided.
- Provider and Implementer Perspective. The logic used for immunization forecasting is complex. Users should understand the logic used to present vaccine recommendations and provide feedback to software vendors regarding questions about forecasting logic and the presentation.
Click here for a test script scenario that includes the capability, View Reconciled Immunization Forecast. The script also indicates successful performance for each of the test sections. The data included in the test is limited to vaccine due and timing, i.e., it addresses vaccines given too early or those that are administered late and thus require recalculation of the next dose in a series. Further work should address administration of specific antigens in proximity to others (e.g., acceptable interval between live virus and inactivated vaccines).
There is no specific standard set of data elements to evaluate immunization forecast presentation.
[i] American Immunization Registry Association. Data Quality Assurance in Immunization Information Systems: Selected Aspects. May 17, 2013. Available at: http://www.immregistries.org/resources/aira-mirow_dqa_selected_aspects_best_practice_guide_05-17-2013.pdf. Accessed 22 September 2015.
[iv] Centers for Disease Control and Prevention. Clinical Decision Support for Immunization (CDSi). 11 February 2014. Web. 4 March 2014. Available at: http://www.cdc.gov/vaccines/programs/iis/interop-proj/cds.html. Accessed 9 December 2014.