Effectiveness of Influenza Vaccine for Preventing Laboratory-Confirmed Influenza Hospitalizations in Immunocompromised Adults

• Published in: Clinical infectious diseases Vol. 73; no. 11; pp. e4353 - e4360
• Main Authors: Hughes, Kailey, Middleton, Donald B, Nowalk, Mary Patricia, Balasubramani, Goundappa K, Martin, Emily T, Gaglani, Manjusha, Talbot, H Keipp, Patel, Manish M, Ferdinands, Jill M, Zimmerman, Richard K, Silveira, Fernanda P
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 06.12.2021
• Subjects: Adult; Case-Control Studies; Hospitalization; Humans; Influenza Vaccines; Influenza, Human - epidemiology; Influenza, Human - prevention & control; Laboratories; Middle Aged; Online Only; Seasons; United States - epidemiology; Vaccination; United States; influenza; vaccine effectiveness; vaccine; immunocompromised
• ISSN: 1058-4838; 1537-6591
• DOI: 10.1093/cid/ciaa1927

Abstract Background Yearly influenza immunization is recommended for immunocompromised (IC) individuals, although immune responses are lower than that for the nonimmunocompromised and the data on vaccine effectiveness (VE) in the IC is scarce. We evaluated VE against influenza-associated hospitalization among IC adults. Methods We analyzed data from adults ≥ 18 years hospitalized with acute respiratory illness (ARI) during the 2017–2018 influenza season at 10 hospitals in the United States. IC adults were identified using prespecified case definitions using electronic medical record data. VE was evaluated with a test-negative case-control design using multivariable logistic regression with polymerase chain reaction–confirmed influenza as the outcome and vaccination status as the exposure, adjusting for age, enrolling site, illness onset date, race, days from onset to specimen collection, self-reported health, and self-reported hospitalizations. Results Of 3524 adults hospitalized with ARI, 1210 (34.3%) had an immunocompromising condition. IC adults were more likely to be vaccinated than non-IC (69.5% vs 65.2%) and less likely to have influenza (22% vs 27.8%). The mean age did not differ among IC and non-IC (61.4 vs 60.8 years of age). The overall VE against influenza hospitalization, including immunocompetent adults, was 33% (95% confidence interval [CI], 21–44). VE among IC vs non-IC adults was lower at 5% (95% CI, –29% to 31%) vs 41% (95% CI, 27–52) (P < .05 for interaction term). Conclusions VE in 1 influenza season was very low among IC individuals. Future efforts should include evaluation of VE among the different immunocompromising conditions and whether enhanced vaccines improve the suboptimal effectiveness among the immunocompromised. Annual influenza vaccination is recommended for immunocompromised (IC), but its effectiveness is not well known. Using prespecified case definitions of immunocompromise, we demonstrated that in the 2017–2018 influenza season, influenza vaccine effectiveness was not significant among hospitalized IC adults.