Objectives: To evaluate the performance of rapid influenza diagnostic tests (RIDT) and influenza vaccines' effectiveness (VE) during an outbreak setting. Methods: We compared the performance of a RIDT with RT-PCR for influenza virus detection in influenzalike illness (ILI) patients enrolled during the 2016/17 season in Mexico City. Using the test-negative design, we estimated influenza VE in all participants and stratified by age, virus subtype, and vaccine type (trivalent vs quadrivalent inactivated vaccines). The protective value of some clinical variables was evaluated by regression analyses. Results: We enrolled 592 patients. RT-PCR detected 93 cases of influenza A(H1N1)pdm09, 55 of AH3N2, 141 of B, and 13 A/B virus infections. RIDT showed 90.7% sensitivity and 95.7% specificity for influenza A virus detection, and 91.5% sensitivity and 95.3% specificity for influenza B virus detection. Overall VE was 33.2% (95% CI: 3.0-54.0; p = 0.02) against any laboratory-confirmed influenza infection. VE estimates against influenza B were higher for the quadrivalent vaccine. Immunization and occupational exposure were protective factors against influenza. Conclusions: The RIDT was useful to detect influenza cases during an outbreak setting. Effectiveness of 2016/17 influenza vaccines administered in Mexico was low but significant. Our data should be considered for future local epidemiological policies.

No abstract