Wa

Ka

b

c

a

ARRAA

KVVHHIH2

1

ditwhanIti(o

0d

Vaccine 29 (2011) 2978–2985

Contents lists available at ScienceDirect

Vaccine

journa l homepage: www.e lsev ier .com/ locate /vacc ine

orkplace efforts to promote influenza vaccination among healthcare personnelnd their association with uptake during the 2009 pandemic influenza A (H1N1)

atherine Harrisa,∗, Jürgen Maurera, Carla Blackb, Gary Eulerb, Srikanth Kadiyalac

RAND Corporation, 1200 South Hayes Street, Arlington, VA, United StatesU.S. Centers for Disease Control and Prevention, Atlanta, GA, United StatesRAND Corporation, Santa Monica, CA, United States

r t i c l e i n f o

rticle history:eceived 7 October 2010eceived in revised form 22 January 2011ccepted 31 January 2011vailable online 18 February 2011

eywords:accination requirementaccination mandateealthcare personnelealthcare worker

nfluenza vaccination1N1009 pandemic

a b s t r a c t

Background: Survey data suggest that, in a typical year, less than half U.S. healthcare personnel (HCP) arevaccinated for influenza. We measured workplace efforts to promote influenza vaccination among HCPin the U.S. and their association with seasonal and pandemic vaccination during the 2009–10 influenzaseason.Methods: Self-reported survey data collected in June 2010 from eligible HCP (n = 1714) participatingin a nationally representative, online research panel. HCP eligible for participation in the survey werethose reporting as patient care providers and/or working in a healthcare setting. The survey measuredworkplace exposure to vaccination recommendations, vaccination requirements, on-site vaccination,reminders, and/or rewards, and being vaccinated for seasonal or H1N1 influenza.Results: At least two-thirds of HCP were offered worksite influenza vaccination; about one half receivedreminders; and 10% were required to be vaccinated. Compared to HCP in other work settings, hospitalemployees were most (p < 0.001) likely to be the subject to efforts to promote vaccination. Vaccinationrequirements were associated with increases in seasonal and pandemic vaccination rates of between 31and 49% points (p < 0.005). On-site vaccination was associated with increases in seasonal and pandemic

vaccination of between 13 and 29% points (p < 0.05). Reminders and incentives were not associated withvaccination.Conclusions: Our findings provide empirical support for vaccination requirements as a strategy for increas-ing influenza vaccination among HCP. Our findings also suggest that making influenza vaccination

couong

available to HCP at workpromoting vaccination am

. Introduction

Influenza is a leading cause of illness, death, and lost pro-uctivity in the United States [1–4]. During a typical season,

nfluenza-related morbidity and mortality is concentrated amonghe elderly, children less than two years of age, and others witheakened immune systems [1,2,5–7]. During the recent pandemic,owever, the burden of influenza was heaviest among childrennd young and middle-aged adults who lacked immunity to theewly circulating 2009 influenza A (H1N1) (2009 H1N1) virus [8].

nfluenza vaccination is the most effective way to prevent the

ransmission of influenza [9,10]. Public health officials stress themportance of high vaccination levels among healthcare personnelHCP) [11]. Vaccinating HCP can reduce the iatrogenic transmissionf influenza virus to patients who are at elevated risk of influenza

∗ Corresponding author. Tel.: +1 703 413 1100x5466.E-mail address: kharris@rand.org (K. Harris).

264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2011.01.112

ld increase uptake and highlight the need to reach beyond hospitals inHCP.

© 2011 Elsevier Ltd. All rights reserved.

and influenza complications [7,12]. In addition, vaccinating HCP canalso reduce influenza-related absenteeism, ensuring the capacityof the healthcare system to meet elevated demand for healthcareduring influenza outbreaks [13]. These considerations motivatelong-standing federal recommendations that all HCP be vaccinated[14], accreditation standards adopted by the Joint Commission onAccreditation of Healthcare Organizations requiring hospitals andlong-term care facilities to offer influenza vaccine to their staff [15],and the high priority placed on vaccinating HCP as early as possibleduring the 2009 H1N1 pandemic [16].

National survey data suggest that for most of the past decade,less than half of all HCP received annual influenza vaccination[17]. Even with unprecedented levels of public outreach and mediaattention to influenza-related issues during the recent pandemic,

only 62% of HCP were vaccinated for seasonal influenza and 37 forpandemic influenza as of January 2010 [18].

The perception that voluntary vaccination efforts are insuf-ficient to generate substantial increases in uptake of influenzavaccination among HCP has fueled calls for making influenza

ine 29 (2011) 2978–2985 2979

vhavfoec

tHkvvsi

2

2

iCoTcab4teipatpPwtdg

sKtotP(stooes

qtposr(t

Table 1Unweighted sample sizes and weighted prevalence of U.S. healthcarepersonnel—Knowledge Networks, June 2010 (n = 1714).

Characteristics Unweighted n Weighted % (95%-CI)

Combined sample 1714 100.0Work settingHospital 715 36.3 (32.4,40.2)Ambulatory, outpatient, clinic 455 23.8 (20.5,27.0)Long term care 322 25.7 (21.7,29.6)Other 222 14.2 (11.2,17.2)OccupationMD, NP, PA, or dentist 114 5.6 (4.0,7.1)Nurse 460 21.4 (18.1,24.7)Allied health professional 587 34.9 (31.0,38.7)Administration, management 302 16.6 (13.6,19.5)Non-clinical support or other 251 21.6 (17.7,25.5)

K. Harris et al. / Vacc

accination of HCP mandatory [19–22]. As of 2010, at least 58ealth systems, free-standing hospitals and medical practicescross the U.S. have voluntarily required that their employees beaccinated annually for influenza or face consequences, such asace-mask requirements, reassignment to non-patient care duties,r dismissal [6,23]. Case studies and testimonials based on thexperiences of these institutions suggest that such requirementsan be highly effective in increasing uptake [6,23].

To date, however, there exist no broadly generalizable data onhe prevalence of efforts to promote influenza vaccination amongCP and their effectiveness in increasing vaccination. To improvenowledge in this area, we present data from a recent national sur-ey of HCP describing a broad range of efforts to promote influenzaaccination and their cross-sectional associations with uptake ofeasonal and H1N1 pandemic influenza vaccine during the 2009–10nfluenza season.

. Methods

.1. Data source

This study presents self-reported survey data collected dur-ng the June wave of a monthly survey effort funded by theenters for Disease Control and Prevention to monitor uptakef influenza vaccines by HCP during the 2009 H1N1 pandemic.he data were collected between June 1 and June 30, 2010 andomprise a sample of self-identified HCP drawn from a nation-lly representative, online research panel developed and operatedy Knowledge Networks (KN), Inc. The KN panel contains about0,000 U.S. households who are randomly selected using bothelephone- and address-based sampling methods, in order tonsure coverage of cell-phone only households in the U.S. Adultsn selected households are invited by telephone or mail to partici-ate in the panel. KN provides households lacking Internet accesst the time of recruitment with required hardware and trainingo access the Internet and an Internet service connection. Othersarticipate using their own computers and Internet connections.anelists access surveys online using a unique username and pass-ord, and receive emails three to four times a month inviting

hem to participate in surveys. More information about the panelesign can be obtained at http://www.knowledgenetworks.com/anp/docs/Knowledge%20Networks%20Methodology.pdf.

Eligibility for the HCP survey was based on responses to acreening questionnaire administered upon recruitment into theN panel. The screening questionnaire asks panelists to describe

heir current occupational characteristics and work setting basedn the Standard Occupational Classification (SOC) System [24] andhe North American Industry Classification System (NAICS) [25].anelists were eligible for inclusion in our study if they reported1) working as a medical doctor, health technologist, healthcareupport staff, or other health practitioner or (2) working in a hospi-al, ambulatory care setting, nursing home, residential care facilityr other health-related setting based on corresponding occupationr industry codes. A total of 2001 or 73.1% of sampled KN pan-lists eligible for participation in the survey responded to the Juneurvey.

To increase the specificity of our analyses, we asked detaileduestions about respondents’ current occupation and work set-ing and whether respondents’ work involved “hands-on care ofatients” during the last twelve months. Consistent with definitions

f HCP in other national data sources [18], the restricted analyticample included only those who report work in a healthcare-elated setting or work involving “hands-on care of patients”n = 1798). We further dropped respondents with missing data onheir occupation or work setting (33 deletions) or any of the other

Patient contactProvides hands-on patient care 1126 65.5 (61.5,69.5)Work around seriously ill patients 496 25.5 (22.1,29.0)Contact with influenza patients 890 51.6 (47.5,55.8)

survey items used in our analysis (51 deletions), which resulted ina final analytic sample of 1714 HCP.

2.2. Measures

The survey measured being subject to employer efforts to (1)promote influenza vaccination through voluntary means includ-ing advice and information about vaccination, recommendationsto be vaccinated issued via letters, emails, and phone calls, andrewards for being vaccinated, (2) make vaccination more conve-nient by offering it in the workplace, and (3) require vaccinationwith and without penalties for non-compliance, such as reassign-ment to a different work location or being terminated. The surveymeasured self-reported uptake of seasonal influenza vaccinationfrom August 2009 through the interview date and uptake of pan-demic influenza vaccine from October 2009 through the interviewdate. Subgroups of HCP were formed based on questions about theiroccupation, work setting, involvement in hands-on patient care,contact with seriously ill patients, and contact with flu patientsand patients with flu-like symptoms.

2.3. Analysis

Our final analytic sample (see Table 1) is described in termsof work setting, occupation and patient care as well as selectedrespondent characteristics such as race and ethnicity, gender andexistence of personal health risk factors or household contactsthat imply an influenza vaccination recommendation for reasonsother than the respondents’ status as a health care worker [14].Unadjusted estimates of the (1) prevalence of employer-based vac-cination policies and programs for seasonal and H1N1 influenza and(2) uptake of seasonal and H1N1 influenza vaccines are presentedboth in aggregate as well as by vaccination policy, patient con-tact, work setting, and occupation. Estimates of the strength of theassociation between employer-based policies and programs andthe uptake of influenza vaccines were derived from multivariablelinear probability models controlling for observable factors influ-encing the uptake of both pandemic and seasonal vaccine such asoccupation, work setting, basic demographics, and being recom-mended for influenza vaccination for reasons other than being anHCP (i.e., on the basis of age, health status, and personal contacts).

All estimates were weighted to reflect selected demographic

and geographic characteristics of HCP in the U.S. population asobtained from the most current monthly Current Population Sur-vey (CPS) [26] and occupational characteristics measured in theKN screening questionnaire. Statistical analyses were conductedusing STATA SE 10.1 (StataCorp, College Station, TX) and adjusted

2 ine 29

Wda

3

sanssmospRog

ttrsocLccsvsbpimm

br2oawbHptwc

ovacinptptRia(

980 K. Harris et al. / Vacc

ald tests were used to assess statistical significance of groupifferences. The RAND Corporation’s Institutional Reviewed Boardpproved the study design and survey protocols.

. Results

Overall, 10.5% of HCP reported that their employer requiredeasonal vaccination during the 2009–10 season (Table 2a). Inbout half of these instances, HCP were subject to a penalty foron-compliance, including requirements to wear face masks, reas-ignments to different work locations or job termination. At theame time, 63.5% of HCP reported that their employer recom-ended seasonal influenza vaccination and 71.5% of HCP were

ffered seasonal vaccination onsite during the 2009–10 influenzaeason. In addition, 47.6% of HCP received a postcard, letter, email orhone call reminding them to get vaccinated for seasonal influenza.eceipts of vaccination rewards appear to be fairly uncommon;nly 4.7% of HCP worked for an employer offering an incentive foretting a seasonal influenza vaccination.

Being subject to employer-sponsored policies and programso promote seasonal vaccination varied widely across work set-ings. Being subject to both vaccination recommendations andequirements, with and without penalties for non-compliance, wasignificantly more prevalent in hospital settings. Moreover, 91.7%f HCP working in hospitals were offered seasonal vaccine at work,ompared to roughly 60% of HCP working in non-hospital settings.ikewise, the probability of receiving a reminder for seasonal vac-ination was twice as high among HCP working in hospital settingsompared to those working in non-hospital settings. While expo-ure to policies and programs to promote seasonal vaccination alsoaried by occupation, supplementary multivariable analysis (nothown) suggests that this variation is explained, in large measure,y work setting. The prevalence of workplace efforts to promoteandemic vaccination was somewhat lower than that for seasonal

nfluenza, but otherwise displayed a very similar pattern (i.e., theost substantive efforts to promote pandemic vaccination wereade by hospitals) (Table 2b).Sixty-one percent of HCP were vaccinated for seasonal influenza

y June 2010, compared to only 42.9% vaccinated for H1N1, likelyeflecting the relatively late arrival of H1N1 vaccine in October of009 after the peak of the pandemic had passed (Table 3) [27]. Ratesf both seasonal and pandemic influenza vaccination were stronglyssociated (p < 0.001) with employer-based policies and programs,ith the exception of vaccination rewards. Vaccination rates for

oth types of influenza were significantly higher (p < 0.001) amongCP who cared for seriously ill patients or had regular contact withatients with influenza. Vaccination rates varied within work set-ing and occupational categories, with highest rates among thoseorking in hospitals and among the combined category of physi-

ians, physician’s assistants, nurse practitioners, and dentists.Table 4 shows the percentage point change in the probability

f vaccine uptake associated with the presence of a workplaceaccination effort or occupational characteristic relative to itsbsence, holding other variables constant. Of the six employer poli-ies considered, vaccination requirements displayed the strongestndependent association with both seasonal and pandemic vacci-ation. Requirements with penalties were associated with a 39.7%oint increase (p < 0.001) in the probability of seasonal vaccina-ion and a 49.0% point increase in the probability of vaccination forandemic influenza (p = <0.001) compared to HCP who were nei-

her required to be vaccinated or recommended for vaccination.equirements without penalties were associated with a 31.5% point

ncrease (p < 0.001) in the probability of seasonal vaccination and30.5% point increase in the probability of pandemic vaccination

p < 0.05). Vaccination recommendations, by contrast, were asso-

(2011) 2978–2985

ciated with an increase in the probability of seasonal vaccinationof only 9.2% points (not significant) and 19.5% points (p < 0.001) forpandemic vaccination. Offering vaccination at work was associatedwith a 28.8% point increase (p < 0.001) in the probability of seasonalvaccination and a 13.4% point increase (p < 0.05) in pandemic vac-cination. Receipt of vaccination reminders and rewards was notsignificantly associated with increased vaccination for either sea-sonal or pandemic influenza in a multivariable analysis.

Working around seriously ill patients and having contactwith influenza patients were associated with 8.6 and 7.2% pointincreases (p < 0.05) in the probability of seasonal vaccination,respectively, even after controlling for policies and programs topromote vaccination. Patient contact was not associated with pan-demic vaccination when controlling for other characteristics ofHCP.

After controlling for vaccination policies and programs, workingin an ambulatory care settings was associated with higher seasonalvaccination rates compared to working in a hospital (the referentcategory) (p < 0.01). Otherwise, work setting was not associatedwith either seasonal or pandemic vaccination in the presence ofcontrol variables.

Occupation was not independently associated with seasonalvaccination after controlling for specific policies, programs, andHCP characteristics. However, support staff and other types ofHCP were significantly less likely to be vaccinated for pandemicinfluenza compared to front-line personnel, including physicians,physicians assistants, nurse practitioners, and dentists (the referentcategory) (p < 0.05).

4. Comment

Overall, our results indicate that employers used a variety ofapproaches to promote influenza vaccination among HCP duringthe 2009–10 influenza season. Efforts included worksite vaccina-tion, vaccination reminders, recommendations, and requirementsto be vaccinated, with or without penalties for non-compliance. Atleast two-thirds of HCP worked for employers who offered work-site vaccination, and about one half of HCP received remindersfrom their employer to be vaccinated. By contrast, only 10% ofHCP worked for an employer who required vaccination, withor without penalties for noncompliance. Compared to HCP inother work settings, hospital employees were most likely to besubject to employer vaccination policies such as on-site vacci-nation programs, vaccination recommendations, and vaccinationrequirements.

Among the worksite policies studied, vaccination requirementswere most strongly associated with vaccination for both seasonaland pandemic influenza. We found that employer requirementsbacked by penalties were more strongly associated with vaccina-tion than requirements without penalties. This finding is consistentwith case studies reporting universal, or close to universal, vacci-nation rates among hospital employees who face the prospect oftermination or other strict penalties [28]. Published case reportsare less conclusive about the effect of requirements paired withless strict penalties, such as having to sign a declination form[29–31].

Our findings suggest that even without penalties for non-compliance, a policy of requiring vaccination is associated withsharply higher rates of compliance than that associated with morecommonly used recommendations, such as recommendations and

reminders alone or in combination. Our findings also suggest thatthe convenience of on-site vaccination is very important. We foundthat HCP offered vaccination at work were nearly 30% more likelyto be vaccinated for seasonal influenza and almost 15% more likelyto be vaccinated for pandemic influenza compared to those not

K.H

arriset

al./Vaccine

29 (2011) 2978–29852981

Table 2aSelf-reported employer-based seasonal influenza vaccination requirements, recommendations and other workplace programs and policies by work setting, occupation, and vaccination type during the 2009–10 influenza season(n = 1714).

Employee characteristic Requirements and recommendations Other workplace policies and programs

Vaccination requiredwith penalty

Vaccination requiredwithout penalty

Vaccination recommended Vaccination offered atworksite

Vaccination reminders issued Vaccination rewards offered

Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea

Overall 5.8 (4.1,7.6) 4.7 (2.8,6.6) 63.5 (59.3,67.6) 71.5 (67.6,75.4) 47.6 (43.5,51.7) 4.7 (2.8,6.6)Patient contactProvides hands-on care 5.9 (4.0,7.9) 0.880 5.3 (2.7,7.8) 0.379 63.3 (58.2,68.4) 0.909 70.5 (65.7,75.3) 0.468 47.8 (42.8,52.8) 0.908 4.8 (2.4,7.2) 0.935Cares for seriously ill 14.0 (9.1,18.9) <0.001 6.3 (2.5,10.2)b 0.299 69.8 (62.3,77.2) 0.071 91.5 (87.3,95.7) <0.001 63.9 (56.3,71.5) <0.001 5.0 (2.6,7.4) 0.827Contact with influenza

patients7.1 (4.6,9.6) 0.177 5.3 (2.7,7.8) 0.553 69.7 (64.3,75.2) 0.002 80.9 (76.5,85.2) <0.001 50.8 (45.0,56.5) 0.120 4.5 (2.2,6.7) 0.781

SettingHospital 12.9 (8.4,17.3) <0.001 4.0 (2.3,5.7) 0.784 72.4 (65.8,79.1) 0.004 91.7 (87.1,96.4) <0.001 68.9 (62.2,75.6) <0.001 6.9 (3.2,10.5) <0.041Ambulatory, outpatient 2.3 (1.0,3.6) 4.1 (1.4,6.9)b 63.3 (56.2,70.3) 61.2 (54.1,68.4) 37.5 (30.0,45.1) 2.6 (0.3,4.9)b

Long term care 2.0 (0.4,3.5)b 6.2 (0.7,11.8)b 58.3 (49.0,67.6) 60.9 (51.8,70.1) 35.8 (27.1,44.5) 5.9 (1.3,10.5)b

Other 0.8 (0.0,1.8)b 4.7 (0.0,10.3)b 50.1 (38.6,61.6) 56.2 (44.8,67.6) 31.6 (21.6,41.7) 0.5 (0.0,1.1)b

OccupationMD, NP, PA, or dentist 3.4 (0.2,6.6)b 0.163 12.1 (0.0,24.4)b 0.214 60.3 (46.1,74.5) 0.048 72.1 (59.8,84.4) <0.001 38.3 (25.5,51.2) 0.060 –c 0.207Nurse 8.3 (4.2,12.4) 3.0 (0.8,5.2)b 73.3 (65.0,81.6) 88.4 (82.7,94.2) 54.8 (46.0,63.6) 3.5 (1.3,5.8)b

Allied health professional 5.4 (2.9,7.8) 4.0 (1.4,6.7)b 61.1 (54.6,67.7) 61.6 (55.1,68.1) 46.1 (39.5,52.7) 3.5 (1.7,5.4)Admin and management 8.3 (1.5,15.1)b 3.7 (1.3,6.0)b 67.6 (58.1,77.0) 80.2 (72.5,87.9) 54.7 (45.1,64.3) 6.1 (0.5,11.7)b

Support staff and other 2.9 (0.5,5.2)b 6.3 (0.1,12.4)b 55.1 (44.2,65.9) 63.9 (53.3,74.5) 40.0 (29.9,50.0) 8.0 (1.5,14.4)b

a p-Values indicate statistical significance of differences in vaccination rates across groups defined by employee characteristics.b Relative standard error larger than 30%.c Sample too small to calculate.

2982K

.Harris

etal./V

accine29 (2011) 2978–2985

Table 2bSelf-reported employer-based H1N1 pandemic influenza vaccination requirements, recommendations and other workplace programs and policies by work setting, occupation, and vaccination type during the 2009–10 influenzaseason (n = 1714).

Employee characteristic Requirements and recommendations Other workplace policies and programs

Vaccination requiredwith penalty

Vaccination requiredwithout penalty

Vaccination recommended Vaccination offered atworksite

Vaccination reminders issued Vaccination rewards offered

Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea Weighted %(95-CI)

p-Valuea

Overall 4.9 (3.0,6.7) 2.8 (1.4,4.3) 62.5 (58.3,66.7) 63.2 (59.1,67.3) 44.6 (40.5,48.6) 3.6 (1.8,5.4)Patient contactProvides hands-on care 4.1 (2.5,5.7) 0.305 3.0 (1.1,4.9)b 0.758 64.7 (59.7,69.6) 0.166 63.7 (58.7,68.6) 0.761 45.3 (40.4,50.3) 0.600 2.8 (1.1,4.5) 0.215Cares for seriously ill 10.5 (6.4,14.6) <0.001 2.8 (1.0,4.5)b 0.924 73.9 (66.9,80.8) 0.001 87.0 (82.2,91.8) <0.001 60.2 (52.6,67.8) <0.001 3.5 (1.7,5.3) 0.919In contact with flu patients 6.0 (3.3,8.7) 0.231 2.8 (1.3,4.3) 0.959 68.4 (62.9,73.9) 0.004 69.2 (63.7,74.7) 0.003 45.0 (39.4,50.7) 0.824 3.2 (1.2,5.3)b 0.664SettingHospital 9.2 (5.3,13.1) <0.001 2.5 (1.1,3.9) 0.628 78.6 (72.6,84.6) <0.001 89.8 (84.6,95.0) <0.001 68.4 (61.7,75.1) <0.001 6.3 (2.7,9.9) 0.034Ambulatory, outpatient 5.1 (0.5,9.8)b 2.0 (0.8,3.3)b 54.6 (47.0,62.1) 49.6 (42.1,57.1) 33.5 (26.3,40.6) 1.3 (0.0,3.1)b

Long term care 1.0 (0.1,1.9)b 3.0 (0.0,7.1)b 54.9 (45.7,64.2) 52.8 (43.6,62.0) 32.6 (24.0,41.1) 3.8 (0.0,8.0)b

Other 0.4 (0.0,1.1)b 4.8 (0.0,10.3)b 48.4 (37.0,59.8) 37.1 (26.4,47.7) 23.9 (15.0,32.7) 0.2 (0.0,0.6)b

OccupationMD, PA, NP, or dentist 2.1 (0.0,4.8)b 0.016 5.1 (0.8,9.3)b 0.588 61.3 (47.9,74.8) <0.001 64.6 (51.7,77.5) <0.001 38.2 (25.4,51.0) 0.056 –c 0.150Nurse 3.7 (2.0,5.3) 1.7 (0.0,3.6)b 78.7 (71.5,85.9) 81.6 (75.3,88.0) 51.3 (42.6,59.9) 2.6 (0.8,4.4)b

Allied health professional 4.9 (2.3,7.5) 2.6 (0.5,4.6)b 60.2 (53.7,66.7) 53.2 (46.5,59.8) 42.6 (36.1,49.1) 2.2 (0.8,3.7)b

Admin and management 10.7 (2.1,19.3)b 2.4 (0.5,4.3)b 66.3 (56.7,75.9) 68.1 (58.8,77.3) 52.4 (42.8,62.0) 4.4 (0.0,9.5)b

Support staff and other 2.3 (0.4,4.2)b 4.1 (0.0,9.3)b 47.6 (37.0,58.1) 57.1 (46.4,67.9) 36.6 (27.0,46.3) 7.2 (0.8,13.5)b

a p-Values indicate statistical significance of differences in vaccination rates across groups defined by employee characteristics.b Relative standard error larger than 30%.c Sample too small to calculate.

K. Harris et al. / Vaccine 29 (2011) 2978–2985 2983

Table 3Uptake of influenza vaccine by vaccine type and worker characteristic (n = 1714).

Employee characteristic Seasonal influenza vaccination 2009 H1N1 influenza vaccination

Weighted % (95-CI) p-Valuea Weighted % (95-CI) p-Valuea

Overall 61.0 (56.9,65.1) 42.9 (38.9,46.9)Requirements and recommendationsRequirement with penalty 98.3 (96.0,100.0) <0.001 86.9 (68.4,100.0) <0.001Requirement without penalty 88.0 (68.7,100.0) 63.7 (35.0,92.3)Recommendation 66.2 (61.7,70.7) 50.7 (45.7,55.6)Neither 35.2 (26.5,44.0) 17.3 (11.7,22.9)Other workplace policies and programsOffered vaccination onsite

Yes 72.9 (68.7,77.1) <0.001 54.6 (49.5,59.6) <0.001No 31.3 (24.1,38.5) 22.8 (16.9,28.7)

Received vaccination remindersYes 74.2 (69.3,79.1) <0.001 56.8 (51.0,62.6) <0.001No 49.1 (43.0,55.1) 31.7 (26.3,37.1)

Offered rewards for vaccinationYes 62.0 (40.6,83.5) 0.925 43.0 (20.1,65.9) 0.992No 61.0 (56.8,65.1) 42.9 (20.1,65.9)

Patient contactProvides hands-on care

Yes 62.1 (57.1,67.1) 0.494 44.8 (39.8,49.7) 0.213No 59.1 (52.0,66.1) 39.3 (32.5,46.2)

Cares for seriously illYes 78.8 (72.8,84.8) <0.001 61.7 (54.5,69.0) <0.001No 54.9 (50.0,59.9) 36.4 (31.9,41.0)

In contact with flu patientsYes 70.3 (65.1,75.5) <0.001 50.3 (44.6,56.1) <0.001No 51.2 (45.2,57.1) 34.9 (29.5,40.4)

SettingHospital 68.5 (62.2,74.9) 0.008 55.4 (48.9,62.0) <0.001Ambulatory, outpatient, clinic 63.6 (56.5,70.7) 35.4 (28.8,41.9)Long term care 54.5 (45.2,63.8) 36.1 (27.5,44.7)Other 49.3 (37.8,60.8) 35.8 (24.9,46.7)OccupationMD, PA, NP, or dentist 71.8 (59.8,83.9) 0.021 53.9 (40.0,67.8) 0.002Nurse 69.1 (61.1,77.0) 55.1 (46.6,63.5)Allied health professional 58.1 (51.5,64.6) 40.6 (34.3,46.9)

ss gro

over

navwcHpign

ovtssBathpav

Administration, management 65.8 (56.8,74.8)Support staff and other 51.5 (40.8,62.2)

a p-Values indicate statistical significance of differences in vaccination rates acro

ffered vaccination at work. This finding implies that offeringaccination in work settings where it is not currently offered tomployees may serve to “jump start” efforts to increase vaccinationates among HCP.

Although occupation appears unassociated with seasonal vacci-ation, we found that, even controlling for other factors, physiciansnd other front-line personnel were substantially more likely to beaccinated for pandemic influenza compared to their counterpartsho hold technical, operational, or administrative roles in health-

are delivery organizations. This finding may reflect the fact that1N1 vaccination was recommended for persons involved directatient care [16], but may also suggest that the skepticism regard-

ng the need for pandemic vaccination held by many adults in theeneral population [32,33] is also prevalent among HCPs who areot involved in direct patient care.

Our findings indicate that hospitals played a leading role inrganized efforts to vaccinate HCP during the 2009–10 influenzaaccination season. Compared to doctors’ offices and other care set-ings, hospitals were more likely to have implemented each of theix promotion strategies studied, including offering vaccination on-ite and requiring vaccination with penalties for non-compliance.oth seasonal and pandemic vaccination was substantially highermong hospital employees than among HCP working in other set-

ings. Once we controlled for employer-based vaccination policies,owever, neither working in a hospital nor direct involvement inatient care was independently associated with an increased prob-bility of vaccination for either type of influenza. This suggests thataccination policies and programs, more than the settings in which

43.1 (33.6,52.6)31.5 (22.3,40.7)

ups defined by employee characteristics.

they are implemented, are the determining factor in a healthcareworker’s decision to be vaccinated for influenza.

To our knowledge, our study is the first comprehensive effortto collect nationally representative data describing uptake ofinfluenza vaccine by HCP and their relationship between influenzavaccination and specific employer-based efforts to promote vac-cination and occupational characteristics. Our findings are subjectto three important limitations. First, the representativeness of thesample is uncertain. In the absence of a practical and affordablemethod of drawing a population-based sample of HCP, we recruiteda sample comprised primarily of self-identified HCP who agreed inadvance to participate in our surveys and received a small paymentin exchange for doing so. Although we weighted the respondentsto be nationally representative based demographic and occupa-tional characteristics of HCP as measured in the CPS, this may notnecessarily account for unmeasured differences between the U.S.population of HCP and those in our sample.

Second, our data provide limited insight into the causal rela-tionship between vaccination policies and programs and vaccineuptake. Our data suggest strong associations between the strictnessand comprehensiveness of policies and subsequent vaccination.However, it may be the case that HCPs who have a clear under-standing of influenza vaccine’s benefits in health care settings are

more likely to work for employers with more stringent policies.

Third, our study indicates that there is a clear relationshipbetween workplace vaccination requirements—with or withoutpenalties for non-compliance—and HCP vaccination rates. Yet, oursample was not large enough to provide reliable information about

2984 K. Harris et al. / Vaccine 29 (2011) 2978–2985

Table 4Estimateda percentage point change in the probably of influenza vaccination associated with changes in employee characteristics by vaccine type (n = 1714).

Employee characteristics Seasonal influenza vaccination 2009 H1N1 influenza vaccination

Percentage point change inprobability of uptake (95% CI)

p-Valueb Percentage point change inprobability of uptake (95% CI)

p-Valueb

Requirements and recommendationsNo requirement or recommendation –c –c

Requirement with penalty 39.7 (25.8,53.6) <0.001 49.0 (28.0,70.0) <0.001Requirement without penalty 31.5 (14.4,48.7) <0.001 30.5 (4.1,56.9) 0.024Recommendation 9.2 (−3.2,21.6) 0.146 19.5 (9.4,29.6) <0.001Other workplace policies and programsOffered vaccination onsite 28.8 (17.2,40.5) <0.001 13.4 (3.0,23.8) 0.012Received vaccination reminders 6.7 (−1.6,15.0) 0.112 5.5 (−3.8,14.9) 0.246Offered rewards for vaccination −3.7 (−23.7,16.3) 0.719 −3.4 (−22.8,16.0) 0.733Patient contactProvides hands-on patient care −2.4 (−11.8,7.0) 0.616 −3.4 (−12.6,5.9) 0.476Work around seriously ill patients 8.6 (0.3,16.9) 0.042 9.2 (−0.4,18.7) 0.06Contact with influenza patients 7.2 (0.1,14.3) 0.048 5.2 (−2.5,12.8) 0.186SettingHospital –c –c

Ambulatory, outpatient, clinic 13.1 (3.6,22.6) 0.007 −3.7 (−15.2,7.7) 0.523Long term care 6.9 (−3.2,16.9) 0.181 1.9 (−9.6,13.4) 0.747Other 5.9 (−6.7,18.6) 0.359 6.0 (−7.5,19.6) 0.384OccupationMD, PA, NP, or dentist –c –c

Nurse −1.4 (−14.9,12.1) 0.841 −5.8 (−22.1,10.6) 0.489Allied health professional −4.1 (−16.7,8.4) 0.516 −12.9 (−27.7,1.9) 0.088Administration, management −3.5 (−19.0,11.9) 0.654 −15.7 (−33.0,1.5) 0.074Support staff and other −8.6 (−23.5,6.3) 0.256 −20.4 (−37.4,−3.5) 0.018

beingc l othe

tHc

uvwpeaHmtipcc

A

ffitsdfp

R

[[

[

[

[

[

[

a Estimates based on linear regressions that include controls for gender, race, andontacts and characteristics listed in the table. The models are also controlled for al

b p-Values indicate statistical significance of individual coefficient estimates.c Reference category.

he prevalence or nature of penalties that employers imposed onCP who choose to remain unvaccinated and on differences in vac-ination rates associated with different types of requirements.

In summary, our findings suggest a strong relationship betweenptake and exposure to work-based efforts to promote influenzaaccination. This relationship appeared to be particularly stronghen HCP are subject to vaccination requirements paired withenalties for non-compliance. As such, our findings providempirical support for vaccination requirements as a strategy forchieving substantial increases in influenza vaccination amongCP. Although hospital employees, regardless of occupation, wereost likely to be vaccinated, work setting was less important

han the specific programs employed in the workplace. This find-ng highlights the potential benefits of implementing strategies toromote influenza vaccination among HCP working in any health-are setting, particularly those that are directly involved in patientare.

cknowledgement

This work was performed under contract with the U.S. Centersor Disease Control and Prevention. The authors have no specificnancial interests, relationships, or affiliations that are relevant tohe topic of influenza vaccination of healthcare personnel that con-titute conflicts of interest. Dr. Jurgen Maurer analyzed the surveyata presented in this manuscript. The authors are grateful for help-ul comments from Arthur Kellermann and James Singleton androgramming support from Rick Li.

eferences

[1] Centers for Disease Control and Prevention. Estimates of deaths associated withseasonal influenza—United States, 1976–2007. MMWR Morb Mortal Wkly Rep2010;59(33):1057–62.

[2] Thompson WW, Shay DK, Weintraub E, Brammer L, Bridges CB, CoxNJ, et al. Influenza-associated hospitalizations in the United States. JAMA2004;292(11):1333–40.

[

[

recommended for influenza vaccination based on age, health status, and personalr variables in the table.

[3] Nichol KL, Wuorenma J, von Sternberg T. Benefits of influenza vaccina-tion for low-, intermediate-, and high-risk senior citizens. Arch Intern Med1998;158(16):1769–76.

[4] Mullooly JP, Bennett MD, Hornbrook MC, Barker WH, Williams WW, PatriarcaPA, et al. Influenza vaccination programs for elderly persons: cost-effectivenessin a health maintenance organization. Ann Intern Med 1994;121(12):947–52.

[5] Mullooly JP, Bridges CB, Thompson WW, Chen J, Weintraub E, Jackson LA,et al. Influenza- and RSV-associated hospitalizations among adults. Vaccine2007;25(5):846–55.

[6] Immunization Action Coalition. Honor roll for patient safety: manda-tory influenza vaccination for healthcare workers; 2010. Available from:http://www.immunize.org/laws/influenzahcw.asp [cited 14.04.10].

[7] Salgado CD, Farr BM, Hall KK, Hayden FG. Influenza in the acute hospital setting.Lancet Infect Dis 2002;2(3):145–55.

[8] U.S. Centers for Disease Control and Prevention. CDC estimates of 2009H1N1 influenza cases, hospitalizations and deaths in the United States,April 2009–March 13, 2010; 2010. Available from: http://www.cdc.gov/h1n1flu/estimates/April March 13.htm [4.09.10].

[9] Nichol KL, Treanor JJ. Vaccines for seasonal and pandemic influenza. J Infect Dis2006;194(Suppl. 2):S111–8.

10] Cox NJ, Subbarao K. Influenza. Lancet 1999;354(9186):1277–82.11] Centers for Disease Control and Prevention. Influenza vaccination of health-

care personnel: recommendations of the healthcare infection control practicesadvisory committee (HICPAC) and the advisory committee on immunizationpractices (ACIP). MMWR Morb Mortal Wkly Rep 2006;55(RR-2).

12] Horcajada JP, Pumarola T, Martínez JA, Tapias G, Bayas JM, de la Prada M, et al. Anosocomial outbreak of influenza during a period without influenza epidemicactivity. Eur Respir J 2003;21(2):303–7.

13] Sartor C, Zandotti C, Romain F, Jacomo V, Simon S, Atlan-Gepner C, et al. Dis-ruption of services in an internal medicine unit due to a nosocomial influenzaoutbreak. Infect Control Hosp Epidemiol 2002;23(10):615–9.

14] Fiore AE, Uyeki TM, Broder K, Finelli L, Euler GL, Singleton JA, et al. Preventionand control of seasonal influenza with vaccines: recommendations of the Advi-sory Committee on Immunization Practices (ACIP), 2009. MMWR Recomm Rep2009;58(RR-8):1–52.

15] AARC. JCAHO releases 2007 patient safety goals, issues influenza con-trol standard. In: In the news; 2006. Available from: http://www.aarc.org/headlines/jcaho 2007 goals.cfm.

16] Use of influenza A (H1N1) 2009 monovalent vaccine: recommendations of theAdvisory Committee on Immunization Practices (ACIP), 2009. MMWR Recomm

Rep 2009;58(RR-10):1–8.

17] Caban-Martinez AJ, Lee DJ, Davila EP, LeBlanc WG, Arheart KL, McCollister KE,et al. Sustained low influenza vaccination rates in US healthcare workers. PrevMed 2010;50(4):210–2.

18] Center for Disease Control and Prevention. Interim results: influenza A (H1N1)2009 monovalent and seasonal influenza vaccination coverage among health-

ine 29

[[

[

[

[

[

[

[

[

[

[

[

[

[32] SteelFisher GK, Blendon RJ, Bekheit MM, Lubell K. The public’s response to the2009 H1N1 influenza pandemic. N Engl J Med 2010;362(22):pe65.

[33] Maurer J, Uscher-Pines L, Harris KM. Perceived seriousness of seasonal

K. Harris et al. / Vacc

care personnel—United States, August 2009–January 2010. MMWR MorbMortal Wkly Rep 2010;12:357–84.

19] They should know better. In: The New York times; 2010 [New York, NY].20] Poland GA, Tosh P, Jacobson RM. Requiring influenza vaccination for health care

workers: seven truths we must accept. Vaccine 2005;23(17–18):2251–5.21] Infectious Disease Society of America and Society for Healthcare Epidemi-

ology of America. Nation’s leading infectious diseases experts call formandatory flu vaccine for all healthcare personnel; 2010. Available from:http://www.idsociety.org/Content.aspx?id=16980.

22] American College of Physicians. American college of physicians recom-mends flu vaccination for health care workers; 2007. Available from:http://www.acponline.org/pressroom/hcw.htm.

23] National Influenza Vaccine Summit. Prevent Influenza Now!. Available from:http://www.influenzasummit.org/profs workers.asp [cited 09.04.09].

24] U.S. Bureau of Labor Statistics. Standard occupational classification; 2010.http://www.bls.gov/soc/ [accessed 14.10.10].

25] U.S. Census Bureau. North American industry classification system (NAICS);

2010. http://www.census.gov/eos/www/naics/ [accessed 14.10.10].

26] U.S. Census Bureau. U.S. interim projections by age, sex, race, and hispanicorigin: 2000–2050; 2004. Available from: http://www.census.gov/population/www/projections/usinterimproj/.

27] Update on influenza A (H1N1) 2009 monovalent vaccines. MMWR Morb MortalWkly Rep 2009;58(39):1100–1.

(2011) 2978–2985 2985

28] Rakita RM, Hagar BA, Crome P, Lammert JK. Mandatory influenza vaccina-tion of healthcare workers: a 5-year study. Infect Control Hosp Epidemiol2010;31(9):881–8.

29] Talbot TR. Improving rates of influenza vaccination among healthcareworkers: educate; motivate; mandate? Infect Control Hosp Epidemiol2008;29(2):107–10.

30] Polgreen PM, Septimus EJ, Parry MF, Beekmann SE, Cavanaugh JE, SrinivasanA, et al. Relationship of influenza vaccination declination statements andinfluenza vaccination rates for healthcare workers in 22 US hospitals. InfectControl Hosp Epidemiol 2008;29(7):675–7.

31] Palmore TN, Vandersluis JP, Morris J, Michelin A, Ruprecht LM, SchmittJM, et al. A successful mandatory influenza vaccination campaign usingan innovative electronic tracking system. Infect Control Hosp Epidemiol2009;30(12):1137–42.

and A(H1N1) influenzas, attitudes toward vaccination, and vaccine uptakeamong U.S. adults: does the source of information matter? Prev Med2010;51(2):185–7.