BACKGROUND: With reports of human cases of novel influenza A (H7N9) virus in 2013, the Centers for Disease Control and Prevention (CDC) encouraged state public health agencies to review and strengthen their pandemic influenza preparedness plans. The Tennessee Department of Health (TDH) Immunization Program (TIP) evaluated their process for vaccine allocation. In 2009, TDH had created a process that distributed pandemic influenza vaccine broadly and swiftly to over 1,500 public and private facilities. This helped Tennessee immunize 7.13% of its population in the first critical weeks of vaccine availability, compared to the national average of 4.87%. Although successful, the cumbersome process required Microsoft Access queries and Excel macros. In 2013, TIP refined its procedure to increase efficiency without losing essential flexibility.
METHODS: A 3-step vaccine allocation process was developed using SAS version 9.3. Available vaccines are first divided by population among 13 public health regions. Within each region, doses are allocated among facilities by the priority category of the facility (1 [highest] through 5 [ineligible]) assigned by TIP based on the priority tier of patients typical of that type of facility. A preliminary analysis of vaccine requests identifies potential errors in provider-submitted orders and gives data to the user setting the proportion of each available vaccine to allot to each priority category. The second step allocates each type of available vaccine. Providers place orders by brand name and age indication; however, if a specific brand is unavailable, one with the same age indication and packaging is substituted. Providers specify their preference for prefilled syringes (PFS), multi-dose vials (MDV), or first available: substitutions are made only if “first available” is specified. Once the user reviews, modifies and approves the system-generated allocations, the final automated step formats orders for batch uploading into the Vaccine Tracking System (VTrckS), CDC’s application for ordering federal vaccine.
RESULTS: A largely automated process for vaccine order decision support allows the user to efficiently and equitably process vaccine requests from all interested facilities and place small, frequent orders for hundreds of facilities each day, to achieve the objective of easily accessible vaccine for priority recipients during a pandemic. Calculated order recommendations are displayed in editable color-coded XML files to simplify review and manual order adjustments.
CONCLUSIONS: TIP streamlined daily pandemic vaccine allocation and ordering with an automated, yet flexible, process using SAS. The editable output can be manipulated to respond to changing needs or issues with individual facilities.