Evaluation of the Use of Alternative Data Sources for Influenza Virologic Surveillance, Maine 2013-14

BACKGROUND:  The CDC-APHL Influenza Virologic Surveillance Right Size project was launched in 2010 to systematically define the rationale, vital capabilities, and optimal “right size” for influenza virologic surveillance.  Using the Right Size calculator, states can determine the number of influenza samples they need to test weekly in order to meet the optimal surveillance marker.  Maine participated in this federal project to evaluate using alternative data (data other than that generated at our Health and Environmental Testing Laboratory (HETL)) sources in order to meet this number. 

METHODS:  Maine used the Right Size calculator to determine the optimal size of samples for influenza surveillance.  This included comparing the advantages and disadvantages of using molecular assays and/or rapid diagnostic tests for influenza surveillance at various times of the year.  Data was also stratified by individual type of submitter; HETL, clinical laboratories, and primary care office.  We also evaluated our contributions to national virologic surveillance by sample type, geography and age group.

RESULTS:  Maine needs to test 117 influenza samples per week to meet the optimum virologic surveillance goal.  HETL tested an average of 33 samples per week (range 0 to 88 samples).  Two in state reference laboratories that run polymerase chain reaction testing tested an average of 147 samples per week (range 36 to 294 samples).  HETL tested samples from 13/16 counties, reference laboratories tested samples from 15/16 counties and rapid positive test results were reported from all 16 counties.  Maine submitted 52 samples to contribute to national virologic surveillance (39 A/H1, 5 A/H3, and 8 B).  Maine completed pyrosequencing on 20.8% of the influenza A/H1 positive samples, and 64.2% of the influenza A/H3 samples.

CONCLUSIONS:   We determined that it is unfeasible in Maine to achieve the optimum sample size using only data from tests performed at HETL, but that we could meet the sample size by incorporating alternative data.  We also determined that influenza testing type varies by geography, and including all test types provides the best state wide data.  Our contributions to national surveillance were representative of type, and age group, but slightly limited in geography.