ORIGINAL CONTRIBUTION JAMA-EXPRESS Effectiveness and Cost-Benefit of Influenza Vaccination of Healthy Working Adults A Randomized Controlled Trial Carolyn Buxton Bridges, MD Context Although the cost-effectiveness and cost-benefit of influenza vaccination William W. Thompson, PhD are well established for persons aged 65 years or older, the benefits for healthy adults Martin I. Meltzer, PhD younger than 65 years are less clear. Objective To evaluate the effectiveness and cost-benefit of influenza vaccine in pre- Gordon R. Reeve, PhD venting influenzalike illness (ILI) and reducing societal costs of ILI among healthy work- Walter J. Talamonti, MD, MPH ing adults. Nancy J. Cox, PhD Design Double-blind, randomized, placebo-controlled trial conducted during 2 in- fluenza seasons. Heather A. Lilac, RN Setting and Participants Healthy adults aged 18 to 64 years and employed full- Henrietta Hall, BS time by a US manufacturing company (for 1997-1998 season, n=1184; for 1998- Alexander Klimov, PhD 1999 season, n=1191). Keiji Fukuda, MD, MPH Interventions For each season, participants were randomly assigned to receive ei- ther trivalent inactivated influenza vaccine (n=595 in 1997-1998 and n=587 in 1998- HE COST-EFFECTIVENESS OF IN- 1999) or sterile saline injection (placebo; n=589 in 1997-1998 and n=604 in 1998-1999). activated influenza vaccina- Participants in 1997-1998 were rerandomized if they participated in 1998-1999. tion in reducing influenza ill- Main Outcome Measures Influenzalike illnesses and associated physician visits and ness, hospitalization, and death work absenteeism reported in biweekly questionnaires by all participants, and serologi- Tis well established in persons aged 65 cally confirmed influenza illness among 23% of participants in each year (n=275 in 1997- years or older, a group that is at in- 1998; n=278 in 1998-1999); societal cost of ILI per vaccinated vs unvaccinated person. creased risk of severe influenza-related Results For 1997-1998 and 1998-1999, respectively, 95% (1130/1184) and 99% complications.1-5 However, the ben- (1178/1191) of participants had complete follow-up, and 23% in each year had se- efits of annual influenza vaccination of rologic testing. In 1997-1998, when the vaccine virus differed from the predominant healthy adults younger than 65 years are circulating viruses, vaccine efficacy against serologically confirmed influenza illness was less clear.5-12 Between 1% and 26% of 50% (P=.33). In this season, vaccination did not reduce ILI, physician visits, or lost persons aged 18 to 64 years may be in- workdays; the net societal cost was $65.59 per person compared with no vaccination. 13-18 In 1998-1999, the vaccine and predominant circulating viruses were well matched. fected with influenza annually, and Vaccine efficacy was 86% (P=.001), and vaccination reduced ILI, physician visits, and the associated work absenteeism can re- lost workdays by 34%, 42%, and 32%, respectively. However, vaccination resulted 5,6,9,11,19-21 sult in substantial societal costs. in a net societal cost of $11.17 per person compared with no vaccination. To date, only 1 randomized, placebo- Conclusion Influenza vaccination of healthy working adults younger than 65 years controlled cost-effectiveness study can reduce the rates of ILI, lost workdays, and physician visits during years when the among healthy working adults has been vaccine and circulating viruses are similar, but vaccination may not provide overall eco- published.6 That study, conducted from nomic benefits in most years. the societal perspective, reported a net JAMA. 2000;284:1655-1663 www.jama.com savings of $46.85 per healthy adult risk adults have not shown similar eco- Author Affiliations are listed at the end of this article. worker vaccinated against influenza. Corresponding Author and Reprints: Carolyn Bux- However, other studies of non–high- nomic benefits or similarly high attack ton Bridges, MD, Influenza Branch, MS A-32, Divi- rates of influenza-attributable ill- sion of Viral and Rickettsial Diseases, National Center ness.5,12-18 Most influenza vaccine stud- for Infectious Diseases, Centers for Disease Control For editorial comment see p 1699. and Prevention, 1600 Clifton Rd, Atlanta, GA 30333 ies of healthy working adults have been (e-mail: [email protected]). ©2000 American Medical Association. All rights reserved. (Reprinted) JAMA, October 4, 2000—Vol 284, No. 13 1655 Downloaded From: https://jamanetwork.com/ on 09/26/2021 INFLUENZA VACCINATION OF HEALTHY WORKING ADULTS conducted during a single influenza sea- instructed to record temperatures and had been ill for 4 days or less. Speci- son,6,10,11,19-22 limiting their generaliz- symptoms during any respiratory ill- mens were refrigerated at 4°C (39°F) un- ability because influenza illness rates and ness episodes in a study log book as an til they were sent by overnight mail to vaccine efficacy may differ substan- aid to completing subsequent surveys. either the Kaiser Permanente Labora- tially from year to year. In addition, other Participants were then randomly as- tory (Los Angeles, Calif; 1997-1998) or studies of influenza vaccination of signed, using a random-numbers table, the Michigan State Department of Health healthy adults have not included labo- to receive either trivalent inactivated Laboratory (Lansing; 1998-1999), for vi- ratory confirmation of influenza ill- influenza vaccine (FluShield, Wyeth- ral culture. Influenza isolates from study ness.6,9-11 Laboratory testing to support Lederle, Paoli, Pa) or sterile saline in- participants were sent to the CDC and epidemiologic findings is important be- jection as a placebo. Participants in antigenically characterized.28 Isolates cause the specificity of clinical case defi- 1997-1998 were rerandomized if they were used only to characterize sea- nitions for influenza can be low and can participated in 1998-1999. Vaccine and sonal strains and were not used to de- vary depending on the cocirculation of saline were drawn up in identical sy- fine clinical illness. other respiratory pathogens.23-26 ringes by 1 nurse and were adminis- Blood samples were collected prior To address these issues, we studied tered by a different nurse who was to injection, 3 weeks after injection, and the effectiveness and societal cost- blinded to participant randomization. at the end of the season from approxi- benefit of vaccinating healthy work- Blinding was maintained until data col- mately the first 300 participants en- ing adults against influenza during the lection was complete. rolled each year. These samples were 1997-1998 and 1998-1999 influenza From November through March in tested to provide laboratory-con- seasons. each study year, participants were sent firmed estimates of influenza infec- follow-up surveys by e-mail twice tion rates. Not all participants could be METHODS monthly that collected information on tested because of resource limitations. Study Design and Population respiratory illnesses and related phy- In 1997-1998, a total of 298 persons We conducted a double-blind, random- sician visits, medications, hospitaliza- provided preinjection blood samples ized, placebo-controlled trial of inac- tions, and lost workdays. Responses and 275 (92%) returned for the end- tivated influenza vaccine among healthy were returned electronically and the of-season blood sample collection. In working adults during the 1997-1998 data were entered directly into a se- 1998-1999, a total of 278 (94%) of 296 and 1998-1999 influenza seasons. Per- cure database. Participants also were persons had complete blood sample col- sons eligible to participate were aged 18 sent by e-mail a questionnaire regard- lection. to 64 years, were full-time employees ing adverse effects that occurred in the Serum was separated from blood and of Ford Motor Co, Dearborn, Mich, did first 7 days after receiving the injec- stored at −20°C until it was tested us- not have any medical conditions for tion. A questionnaire sent at the end of ing hemagglutination inhibition (HI) at which influenza vaccine was recom- the study asked participants if they had the CDC.28 For 1997-1998, the HI test mended by the US Advisory Commit- received any influenza vaccine other antigens were vaccine strains A/Johan- tee on Immunization Practices, and did than the study injection since enroll- nesburg/82/96(H1N1), A/Nanchang/ not have any contraindications to vac- ment and asked them to guess whether 933/95(H3N2), and B/Harbin/7/94; cination.27 Participants were recruited they received vaccine or placebo. reference outbreak strain A/Sydney/ through e-mail notices and study pre- Reminder e-mails were sent if com- 5/97(H3N2); and outbreak strain sentations at the work site. Written in- pleted surveys were not received after A/Michigan/8/98(H3N2). For 1998- formed consent was obtained from all 1 week. Participants were telephoned 1999, HI test antigens were vaccine participants. The study was approved a minimum of 2 times if electronic re- strains A/Beijing/262/95(H1N1), A/Syd- by the institutional review board at the sponses were not received by 1 week ney/5/97(H3N2), and B/Harbin/7/94; Centers for Disease Control and Pre- after the reminder e-mail was sent. and outbreak strain A/Michigan/ vention (CDC), Atlanta, Ga. 15/99(H3N2). An HI antibody titer of Virologic Surveillance less than 10 was assigned a value of 5. Protocol and Serologic Studies A 4-fold or greater rise in antibody ti- During enrollment in October of each The influenza period was defined as the ter against either a vaccine strain or an study year, eligibility was determined, period during which clinical speci- outbreak strain between the 3-week- informed consent was obtained, and in- mens collected from ill study partici- postinjection and end-of-season se- formation on demographics, house- pants yielded influenza viruses. Dur- rum samples was considered evidence hold composition and income, prior in- ing November through April of each of influenza infection.