Rh

PPa

b

c

d

e

f

a

ARRAA

KVARAVHTI

1

arpcR[ttt

0d

Vaccine 29 (2011) 6592– 6597

Contents lists available at ScienceDirect

Vaccine

j ourna l ho me pag e: www.elsev ier .com/ locate /vacc ine

isk of rheumatoid arthritis following vaccination with tetanus, influenza andepatitis B vaccines among persons 15–59 years of age

aula Raya, Steven Blackb, Henry Shinefieldc, Aileen Dillond, Diane Carpentere, Edwin Lewisa,at Rossa, Robert T. Chenf, Nicola P. Kleina, Roger Baxtera,∗, for the Vaccine Safety Datalink Team1

Kaiser Permanente Vaccine Study Center, Oakland, CA, USACenter for Global Health, University of Cincinnati Children’s Hospital, Cincinnati, OH, USAUniversity of California San Francisco Department of Pediatrics, CA, USAThe Permanente Medical Group, San Francisco, CA, USAKaiser Permanente Division of Research, Oakland, CA, USACenters for Disease Control and Prevention, Atlanta GA, USA

r t i c l e i n f o

rticle history:eceived 2 March 2011eceived in revised form 27 June 2011ccepted 29 June 2011vailable online 16 July 2011

eywords:accinedverseheumatoidrthritisaccination

a b s t r a c t

Background: Associations between vaccinations, particularly hepatitis B, and onset of rheumatoid arthritis(RA) have been reported, but examined in few large-scale studies.Method: Onset of RA cases and dates of vaccination against hepatitis B, tetanus, and influenza were iden-tified in a retrospective chart review of approximately 1 million Kaiser Permanente Northern Californiamembers ages 15–59 years from 1997 through 1999. In a cohort analysis, rates of new-onset RA werecompared between vaccinated and unvaccinated within 90, 180, and 365 days. In a case-control analysis,rates of vaccination during exposure intervals (90, 180, 365, and 730 days) were compared between casesand controls using conditional logistic regression.Results: 378 RA cases were included in the cohort analysis; 37 additional cases were included in thecase-control analysis. In the cohort analysis the relative risks of RA onset within 90, 180, or 365 days ofhepatitis B vaccination were not significant (R.R. = 1.44, p = 0.53; R.R. = 1.67, p = 0.22; R.R. = 1.23, p = 0.59

epatitis Betanusnfluenza

respectively). We found a possible association between RA and influenza vaccine in the previous 180and 365 days in the cohort analysis (R.R = 1.36, p = 0.03; R.R. = 1.34, p = 0.01 respectively), but in the case-control analysis, cases were no more likely than controls to have received any of the three vaccines.Conclusions: In this large retrospective study we found no statistically significant association betweenexposure to hepatitis B vaccine and onset of RA. A possible association between RA and influenza vacci-

y was

nation in the cohort stud

. Introduction

Arthropathies, both acute and chronic, have been reported indults following vaccination against diseases such as tetanus [1],ubella [2,3], and hepatitis B [4–7], but studies of causality haveroduced conflicting results [8,9]. The evidence linking tetanus vac-ine to arthritis is limited primarily to case reports (for example,ef. [10]); few cases have been reported to surveillance agencies1], and one large prospective study in Denmark in the 1960s iden-ified no cases of arthritis among 2.5 million doses of monovalent

etanus vaccine and 3.7 million doses of combined diphtheria andetanus vaccine [11], leading the United States Institute of Medicine

∗ Corresponding author. Tel.: +1 510 267 7529.E-mail address: roger.baxter@kp.org (R. Baxter).

1 Members listed in Appendix A.

264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2011.06.112

not borne out in the larger case-control analysis.© 2011 Elsevier Ltd. All rights reserved.

to declare in 1994 that the evidence was insufficient to determinecausality between tetanus toxoid and arthritis [1].

There have also been reports of onset of persistent arthritisafter rubella vaccination. One group of investigators in Canadafound an incidence of persistent or recurrent arthritis followingrubella vaccination as high as 5–11% in small studies of adultwomen [2,3]. A subsequent randomized, placebo-controlled trialfound a significant association between rubella vaccination andacute joint pain, and a marginally significant association with anincreased rate of chronic or recurrent arthralgia [9]. However,this finding was not reproduced in a large retrospective cohortstudy of 4300 women conducted by the Vaccine Safety Datalink(VSD), which found no evidence of an association between rubellavaccination and chronic arthropathy [8]. A review of influenza

and pneumococcal vaccines found them to be safe and immuno-genic in patients with rheumatoid arthritis [12], and researchersfound no increase in autoantibodies after influenza vaccination[13]. More recently, Swedish researchers did not find evidence of an

ne 29 (

a[b

rorpeB1rcifhdBrlttwc

taucatbyaacr

2

2

emitapantB

ew1ytistecp

P. Ray et al. / Vacci

ssociation between common vaccines and rheumatoid arthritis14] in a large case-control study where vaccination status wasased on questionnaires.

Most of the evidence that hepatitis B vaccination may play aole in the development of arthritis has come from case reportsr case series [4–7,15]. In some of these patients, revaccinationeproduced symptoms or worsened symptoms that were alreadyresent [15]. In Ontario, Canada [7], a cluster of 5 male firefight-rs developed arthritis within 2–3 weeks of receiving hepatitis

vaccination, as did 6 female health care workers; 10 of these1 cases met American College of Rhematology (ACR) criteria forheumatoid arthritis (RA) [16]. Notably, several different lots of vac-ine were used, and despite the overlap in occupation the affectedndividuals worked in different locations; a possible contributingactor is that 9 of the 11 had genetic risk factors for RA. Othersave used the Vaccine Adverse Event Reporting System (VAERS)atabase to investigate a possible association between hepatitis

vaccination and RA as well as other autoimmune events andeported a significantly higher risk of these adverse events fol-owing receipt of hepatitis B vaccine than following receipt ofetanus and tetanus–diphtheria vaccines [17,18]. However, reportso VAERS may not reflect true rates of event occurrence, may varyith vaccine and time, and should not be construed as denoting

ausality [19].Here we used data from a large health maintenance organiza-

ion to investigate the occurrence of RA within defined intervalsfter vaccination for hepatitis B in individuals 15–59 years old,sing cohort and case-control designs. The ages of 15–59 werehosen to maximize both new onset RA and use of the vaccine,t a time of universal vaccination of infants, but a recommenda-ion for vaccination of adults only with risk factors, determinedy occupation, sexual orientation, drug use and travel. Personsounger than 15 were less likely to receive the diagnosis of RA,nd those over 59 were less likely to receive the vaccine. Welso analyzed the rates of RA after influenza and tetanus vac-ines, the two most common vaccines given to those in this ageange.

. Methods

.1. Study population

The study population was drawn from members of the North-rn California Kaiser Permanente Health Plan, a large healthaintenance organization with a total membership of approx-

mately 3.2 million. All medical care for members is providedhrough the health plan, and clinic visits and treatments arell documented in comprehensive databases. Vaccinations arerovided to members without charge, and almost all vaccinesre received in the system, and recorded by the Kaiser Immu-ization Tracking System (KITS). The study was approved byhe Kaiser Foundation Research Institute’s Institutional Reviewoard.

We identified a cohort of health plan members continuouslynrolled from January 1, 1995, through December 31, 1999, whoere aged 15 through 59 years during the period from January 1,

997, through December 31, 1999. Continuous enrollment for 2ears prior to the study start date of January 1, 1997, was required sohat complete medical records for all members would be availablen order to determine pre-existing disease and vaccine exposuretatus. Persons with visits for RA and other inflammatory condi-

ions prior to their follow-up start date were excluded. Vaccinexposure status was determined from KITS and supplemented byhart reviews. Chart reviews went back as far as 1986 to determinerior vaccination status.

2011) 6592– 6597 6593

2.2. Medical record reviews

Inpatient and outpatient records of potential cases werereviewed by trained medical record analysts (MRAs) to determinecase status, intervals between the first visit with an RA diagnosis inthe chart and actual onset date, and vaccine exposure status. MRAswere instructed to review all records for each subject that presentedfor medical care any time between January 1, 1995, and the dateof review and to record any diagnoses of arthritis or joint com-plaints during the period from January 1, 1997, through the date ofreview. Medical records were available for all subjects in the study.Dates of symptom onset and initial diagnosis were recorded forany diagnoses of RA, Juvenile Rheumatoid Arthritis, Still’s disease,arthritis, arthralgia, or other joint complaints. Related diagnostictest results, radiologic images, and medications prescribed for thejoint complaints were abstracted, and all immunization recordswere searched dating back to January 1, 1986.

Following the medical record abstractions, potential cases werecategorized into three groups based on the ACR criteria for rheuma-toid arthritis (included in Appendix A) [16]. Cases with a definitediagnosis of RA that clearly met the ACR criteria were accepted as“definite” cases, whereas cases without any evidence of arthritis orjoint complaint during the study period and who were also with-out a history of any inflammatory joint condition were rejected. Allother potential cases were referred for adjudication by the studyrheumatologist (A.D.). For all cases, the rheumatologist also pro-vided a specific or estimated onset date of related symptoms.

2.3. Analysis

All RA cases classified as definite or probable were included inthe analyses described below. Exposure status (i.e., whether thecase was considered as vaccinated within a specific period beforedeveloping RA) was based on the date that the RA symptoms beganrather than date of first RA diagnosis.

2.3.1. Cohort analysisTo assess whether exposure to vaccine resulted in a higher

incidence of RA, rates of new-onset RA were compared duringvaccine-exposed and vaccine-unexposed intervals within follow-up periods of 90, 180, and 365 days. The entire study populationcontributed follow-up time. Vaccinated individuals contributedunexposed follow-up time until they were vaccinated, after whichthey were considered vaccinated for the remainder of the timeinterval analyzed. The follow-up time began on January 1, 1997,or on the 15th birthday, whichever came later. The follow-up timeended on December 31, 1999, the 60th birthday, or the RA onsetdate, whichever came earlier. Relative risks were adjusted for age,sex, and race using Poisson regression because they are factors thatmay be associated with both RA and vaccine exposure [20–23]. Inaddition, because we noticed a marked increase in health care vis-its by the RA cases in our population during the year prior to thedate identified as the first onset of RA symptoms, the analyses alsoadjusted for the number of health care visits within 1 year. We cal-culated the maximum attributable risk for hepatitis B vaccine usingthe incidence of disease in the exposed and the upper limit of theadjusted relative risk 95% confidence interval.

2.3.2. Case-control analysisIn the cohort analysis described above, the number of cases was

limited. In order to increase the number of cases analyzed and morefinely control for health care utilization, a matched case-control

analysis was undertaken that included as cases the incident casesfrom the cohort analysis and additional new onset cases identifiedfrom the study population whose symptoms began during 1996.These cases were identified in the first phase of the study, but the

6594 P. Ray et al. / Vaccine 29

Table 1Demographics of subjects in cohort analysis.

Age in years Person–Yearsa

Male Female Total

15–29 279,658 298,242 577,89930–44 426,762 494,161 920,92345–59 513,102 575,276 1,088,377

otwo

cbR1eba

TA

TR

TO

O

Total 1,219,522 1,367,678 2,587,199

a Includes about 1 × 106 individuals.

nset was found to be prior to 1997, so they had been rejected fromhe original study. Accepting them as onset in 1996 meant that weere only certain of one medical record for at least 1 year prior to

nset, rather than 2 years.Each incident case was matched with three randomly selected

ontrols on age within 1 year and a categorical utilization variableased on the number of clinic visits during the year prior to theA symptom onset date of the matched case (0, 1–2, 3–5, 6–9, or

0+ visits). The odds of vaccination with hepatitis B during variousxposure intervals (90, 180, 365, and 730 days) were comparedetween cases and controls using conditional logistic regressionnd adjusting for sex and race and exact number of utilization

able 2ge and sex of subjects in case-control analysis.

Age (years) at symptom onset Cases

Females N (%) Males N (%) Tot

13–19 6 (85.7) 1 (14.3) 720–29 6 (85.7) 1 (14.3) 730–39 43 (76.8) 13 (23.2) 5640–49 101 (82.1) 22 (17.9) 12350–59 114 (69.1) 51 (30.9) 16560–64 40 (70.2) 17 (29.8) 57Total 310 (74.7) 105 (25.3) 415

able 3elative risk of rheumatoid arthritis in the cohort analysis.

Vaccine Exposure interval (Days) Vaccinated individuals with RA (n)

Hepatitis B 90 3

180 6

365 7

Tetanus 90 10

180 19

365 27

Influenza 90 19

180 62

365 113

a Adjusted for race, sex, and exact number of utilization visits.

able 4dds ratio of exposure to hepatitis B, influenza, and tetanus vaccines in patients with new

Vaccine Exposure interval (Days) Exposed cases, n (%) Exposed co

Hepatitis B 90 4 (1.0) 7 (0.6)

180 8 (1.9) 11 (0.9)

365 10 (2.4) 20 (1.6)

730 12 (2.9) 34 (2.7)

Tetanus 90 10 (2.4) 37 (3.0)

180 21 (5.1) 61 (4.9)

365 29 (7.0) 111 (8.9)

730 60 (14.5) 206 (16.6)

Influenza 90 21 (5.1) 85 (6.8)

180 64 (15.4) 178 (14.3)

365 113 (27.2) 309 (24.8)

730 132 (31.8) 373 (30.0)

.R., odds ratio; CI, confidence interval. Adjusted for race.

(2011) 6592– 6597

visits. We noted an possible increase in utilization among casesduring the year prior to the electronic onset date of RA, which weinterpreted as possibly indicating earlier onset of symptoms beforediagnosis. Therefore, we added a 730-day exposure interval to the90, 180, and 365 day intervals used in the cohort analysis. Simi-lar comparisons were done for exposures to tetanus and influenzavaccines.

3. Results

Approximately 1 million individuals met the criteria for inclu-sion in the study population (Table 1). They contributed a total of2,587,111 person–years. Members of the study population received212,268 doses of hepatitis B vaccine; 656,488 doses of influenzavaccine; and 249,182 doses of tetanus vaccine during the studyperiod.

Overall, inpatient and outpatient medical records of 2354 poten-tial RA cases were reviewed (Fig. 1). A total of 628 cases (26.7%)

were accepted, including those classified as probable RA by therheumatologist. Of these, 550 cases (535 RA and 15 JRA) met ourcase definition for definite RA based on the ACR criteria. All RAcases were in individuals 16 years of age or older. In total, 378

Controls

al N (%) Females N (%) Males N (%) Total N (%)

(1.7) 9 (42.9) 12 (57.1) 21 (1.7) (1.7) 15 (62.5) 9 (37.5) 24 (1.9)

(13.5) 101(60.8) 65 (39.2) 166 (13.3) (29.6) 218 (58.3) 156 (41.7) 374 (30.0) (39.8) 280 (57.8) 204 (42.2) 484 (38.9)

(13.7) 88 (50.0) 88 (50.0) 176 (14.2) (100.0) 711 (57.1) 534 (42.9) 1245 (100.0)

Crude R.R. Adjusted relative riska 95% CI p value

0.71 1.44 0.46, 4.51 0.530.78 1.67 0.74, 3.77 0.220.52 1.23 0.58, 2.63 0.59

1.61 1.36 0.72, 2.54 0.341.55 1.31 0.82, 2.09 0.261.10 0.93 0.62, 1.37 0.70

1.19 0.72 0.45, 114 0.162.24 1.36 1.03, 1.80 0.032.31 1.34 1.06, 1.69 0.01

onset rheumatoid arthritis in the case-control analysis.

ntrols, n (%) Crude O.R. Adjusted O.R. 95% CI p value

1.7 1.5 0.4, 5.2 0.552.2 2.0 0.8, 5.1 0.141.5 1.4 0.6, 3.1 0.391.1 1.0 0.5, 2.12 0.91

0.8 0.8 0.4, 1.6 0.481.0 1.1 0.6, 1.8 0.840.8 0.8 0.5, 1.2 0.310.9 0.8 0.6, 1.12 0.20

0.7 0.7 0.4, 1.2 0.141.1 1.1 0.8, 1.6 0.571.1 1.1 0.9, 1.5 0.431.1 1.1 0.8, 1.4 0.59

P. Ray et al. / Vaccine 29 (2011) 6592– 6597 6595

* Met ARA criteria.** All required ARA criteria could not be verified from chart.RA, Rheumatoid arthritis; JRA, juvenile RA; MRA, medical record analyst.

2354 potential cases reviewed

979 referred to the consulting rheumatologist

970 rejected by the MRA405 accepted as RA or JRA cases by the MRA

287 rejected with no inflammatory condition

78 identified as Probable RA**

469 rejected with other inflammatory conditions

145 accepted as RA or JRA cases*

Total accepted cases* = 550 (535 RA and 15 JRA)

Total including probable RA** = 628 (607 RA and 21 JRA)

Total cases with onset 1996 – 1999* = 365

Total including probable RA with onset 1996–1999 = 415

F s of KD

psac

paw3n

ova1n9ta93vltC

c(Tv

ig. 1. Selection of RA cases from the charts of approximately 1 million memberecember 31, 1999.

robable and definite cases had their symptom onset during thetudy follow-up period (1997–1999); an additional 37 probablend definite cases had onset during 1996 and were included in thease-control analysis.

The cohort analysis encompassed approximately 2.6 millionerson–years, with subjects about evenly divided between malend female (Table 1). The case-control analysis included 415 casesith definite or probable RA, with a female-to-male ratio of about

:1, and 1245 controls with a less pronounced female predomi-ance (Table 2).

In the cohort analysis, the numbers of individuals who devel-ped RA within 1 year of vaccination ranged from 7 for hepatitis Baccine to 113 for influenza vaccine. After adjustment for sex, race,nd number of utilization visits, the relative risks of RA onset 90,80, or 365 days after exposure to hepatitis B vaccine were not sig-ificant (R.R. = 1.44 (p = 0.53, 95% CI 0.46–4.51); R.R. = 1.67 (p = 0.22,5% CI 0.74–3.77); R.R. = 1.23 (p = 0.59, 95% CI 0.58–2.63) respec-ively) (Table 3). The relative risk of exposure to tetanus vaccine waslso not significant for any interval (90 days: R.R. = 1.36 (p = 0.34,5% CI 0.72–2.54); 180 days: R.R. = 1.31 (p = 0.26, 95% CI 0.82–2.09);65 days: R.R. = 1.34 (p = 0.01, 95% CI 0.62–1.37)). In contrast, indi-iduals who received influenza vaccine were significantly moreikely to develop RA within 180 or 365 days following immuniza-ion (R.R = 1.36 (p = 0.03, 95% CI 1.03–1.80); R.R. = 1.34 (p = 0.01, 95%I 1.06–1.69) respectively).

The maximum attributable risk calculations for hepatitis B vac-

ine are as follows. We divided the number of exposed casesn = 6) by the upper limit of the 180-day confidence interval (3.77;able 3) to determine the number of cases expected withoutaccination (n = 1.59) and the number attributable to vaccination

aiser Permanente Northern California ages 15–59 from January, 1, 1997, through

(6 − 1.59 = 4.41). The 4.41 excess cases among 212,000 doses yieldsa rate of 2.1 per 100,000 doses.

Of the additional cases with RA onset during 1996 that wereincluded in the case-control analysis, three were exposed to hep-atitis B vaccine and two were exposed to tetanus vaccine, for atotal of 10 exposed to hepatitis B vaccine, 29 exposed to tetanusvaccine, and 113 exposed to influenza vaccine within 1 year beforethe onset of their RA (Table 4). Cases were not significantly morelikely than controls to have received any of the three vaccines inany of the intervals examined (Table 4). For hepatitis B vaccine for90 days R.R. = 1.5 (p = 0.55, 95% CI 0.4–5.2); for 180 days R.R. = 2.0(p = 0.14, 95% CI 0.8–5.1); for 365 days R.R. = 1.4 (p = 0.39, 95% CI0.6–3.1); and for 730 days R.R. = 1.0 (p = 0.91, 95% CI 0.5–2.12). Fortetanus vaccine for 90 days R.R. = 0.8 (p = 0.48, 95% CI 0.4–1.6); for180 days R.R. = 1.1 (p = 0.84, 95% CI 0.6–1.8); for 365 days R.R. = 0.8(p = 0.31, 95% CI 0.5–1.2); and for 730 days R.R. = 0.8 (p = 0.20, 95% CI0.6–1.12). For influenza vaccine for 90 days R.R. = 0.7 (p = 0.14, 95%CI 0.4–1.2); for 180 days R.R. = 1.1 (p = 0.57, 95% CI 0.8–1.6); for 365days R.R. = 1.1 (p = 0.43, 95% CI 0.9–1.5); and for 730 days R.R. = 1.1(p = 0.59, 95% CI 0.8–1.4). Among the subjects included in the case-control analysis, blacks and Hispanics had significantly higher riskof having RA than whites (Table 5).

4. Discussion

Individual cases of RA and other autoimmune disease occurring

after vaccination have been reported for years [3–7,9,24]. In a fewinstances, such as the 1976–1977 influenza vaccine and Guillain-Barré syndrome [25,26] and the RA 27/3 rubella vaccine and acutearthritis [1], a close association was found. However, for many

6596 P. Ray et al. / Vaccine 29

Table 5Ethnicity of subjects in case-control analysis.

Cases N (%) Controls N (%) OR (95% CI)* p value*

Asian 33 (8.0) 137 (11.0) 0.75 (0.5–1.1) 0.16Black 54 (13.0) 102 (8.2) 1.63 (1.1–2.4) 0.01Hispanic 55 (13.2) 118 (9.5) 1.44 (1.0–2.1) 0.05White 239 (57.6) 736 (59.1) Referent ReferentMulti-ethnic 2 (0.5) 2 (0.2) 3.38 (0.5–24.4) 0.23Other** 1 (0.2) 7 (0.6) 0.46 (0.1–3.7) 0.46Unknown 31 (7.5) 143 (11.4) 0.66 (0.4–1.0) 0.06

otsahiaataao

tfrb[

bbncieatBiv

mp[ttcpmhocfeu

iotlw

l

* Conditional logistic regression.** Includes 5 native Americans, 1 case and 4 controls.

ther vaccines and diseases, including hepatitis B vaccine and RA,he link remains uncertain. RA is thought to develop in geneticallyusceptible individuals in response to environmental stimuli, andctivation of the immune system through infection or vaccinationas been postulated as a possible trigger for the onset or worsen-

ng of autoimmune diseases [27,28]. Individuals with HLA variantsssociated with RA have been among those reported to developrthritis after hepatitis B vaccination [7,15]. In some cases, addi-ional doses of the vaccine worsened the symptoms, supporting

causal link. In contrast, vaccination of patients with active RAgainst hepatitis B did not worsen clinical or laboratory measuresf the disease in a small, controlled study [29].

Very few controlled studies have been performed to resolvehese conflicting findings. Although large case-control studies haveound no association between hepatitis B vaccine and multiple scle-osis [30,31], to our knowledge, only one other large scale study haseen reported to look at the association between RA and vaccines32].

In our initial cohort analysis, we found a possible associationetween RA and influenza vaccine in the previous 180 or 365 days,ut the extended case-control analysis, bringing in more data, didot demonstrate this increase. We found no statistically signifi-ant association between exposure to hepatitis B vaccine and RAn either the cohort or case-control study, although non-significantlevations in risk were seen for intervals of 90, 180, and 365 daysfter vaccination. The power of this study was probably low, and weherefore cannot entirely rule out an association between hepatitis

vaccination and RA, but using the upper limit of the confidencenterval we were able to rule out more than 2.1 cases per 100,000accine recipients.

In our case-control study of a large population of health planembers, the incidence of RA increased with age and occurred

redominantly in females, consistent with previous knowledge20,21]. However, the incidence of RA in our population was lowerhan that found in other populations, possibly because rates con-inue to rise into the 7th decade [33], but our age cutoff was 59. Thisontributed to the low power in the study. We also found that RAatients had higher numbers of health care visits than other planembers in the year before their diagnosis. This increased use of

ealth care in RA patients may be related to the insidious onsetf many cases of RA [34,35], and the increased exposure to thelinic may also increase the likelihood of vaccination [23]. However,urther matching on utilization showed that this is a minor influ-nce (data not shown) and may be accounted for by differences intilization by sex.

There is a potential for residual confounding in our study, if anyndication or contraindication for vaccination were also a causef RA i.e., if hepatitis B disease, occupation, sexual orientation, orravel were a also a risk for RA. We are not aware of any established

ink between RA and Hepatitis B, medical profession, travel, or men

ho have sex with men.The biggest limitation of the study, however, and a lesson

earned, is that observational studies of chronic diseases with

(2011) 6592– 6597

insidious onset, such as RA, are limited by difficulty in deter-mining the date of onset of illness. Electronic records reveal avisit for a condition, but many people present for medical caresome time after symptoms begin, after exhausting home and over-the-counter remedies. Chart review of cases, though costly andtime-consuming, greatly benefited the veracity of the study, buteven with chart review by a rheumatologist it was sometimes dif-ficult to determine the exact onset of symptoms. In addition, wewere limited by sample size; if a very small risk of RA in associationwith vaccines does exist, a larger study would be needed to detectit.

Acknowledgments

This study was funded by the Centers for Disease Control andPrevention Vaccine Safety Datalink Project. The authors would alsolike to thank Pat Ross, Patricia Dameron, Marilyn Foley, KarenForsen, and Frances Sheykhzadeh for their excellent work in col-lecting data from more than 10,000 medical records. Naomi L.Ruff, PhD, contributed to the writing of this article, for which shereceived compensation from the Kaiser Permanente Vaccine StudyCenter.

Appendix A. Appendix A

Other Vaccine Safety Datalink Team Members by Site.Centers for Disease Control and Prevention: Frank DeStefano,

MD, MPH, C. Dexter Kimsey Jr., PhD, MSEH, John Glasser, PhD,MPH, Gina Terraccino, DO, MPH, Jennifer Lloyd, DVM, MSPH, PhillipRhodes, PhD, Emmett Swint, MA; Group Health Cooperative ofPuget Sound: Robert S. Thompson, MD, Robert L. Davis, MD, MPH,Lisa Jackson, MD, MPH, Patti Benson, MPH, Virginia Immanuel,MPH, William Barlow, PhD; Northwest Kaiser Permanente: John P.Mullooly, PhD, Lois Drew, BA, Barbara Mendius, MS; Kaiser Perma-nente of Northern California: Edwin Lewis, MPH, Bruce H. FiremanMA, Tracy A. Lieu, MD, MPH, Robert Hiatt, MD, PhD; Center forVaccine Research, Harbor UCLA Medical Center: Joel I. Ward, MD,Connie M. Vadheim, PhD, S. Michael Marcy, MD, Jennie Jing, MA,Michael Wulfson MD, PhD; Kaiser Permanente of Southern Califor-nia: Marlene Lugg, DrPH, Patricia Osborne, MS; Center for BiologicsEvaluation and Research, Food and Drug Administration: Robert P.Wise, MD, MPH, Suresh Rastogi PhD, Peter Patriarca, MD, Divisionof Vaccine Injury Compensation, Health Resources and ServicesAdministration: Vito Caserta, MD, MPH.

References

[1] Stratton KR, Howe CJ, Johnston Jr RB. Adverse events associated with child-hood vaccines: evidence bearing on causality. Washington, D.C: Institute ofMedicine, Vaccine Safety Committee; 1994.

[2] Tingle AJ, Allen M, Petty RE, Kettyls GD, Chantler JK. Rubella-associated arthritis.I. Comparative study of joint manifestations associated with natural rubellainfection and RA 27/3 rubella immunisation. Annals of the Rheumatic Diseases1986;45(Februry (2)):110–4.

[3] Tingle AJ, Yang T, Allen M, Kettyls GD, Larke RP, Schulzer M. Prospectiveimmunological assessment of arthritis induced by rubella vaccine. Infectionand Immunity 1983;40(April (1)):22–8.

[4] Gross K, Combe C, Kruger K, Schattenkirchner M. Arthritis after hepatitisB vaccination. Report of three cases. Scandinavian Journal of Rheumatology1995;24(1):50–2.

[5] Symmons DP, Chakravarty K. Can immunisation trigger rheumatoid arthritis?Annals of the Rheumatic Diseases 1993;52(December (12)):843–4.

[6] Vautier G, Carty JE. Acute sero-positive rheumatoid arthritis occurringafter hepatitis vaccination. British Journal of Rheumatology 1994;33(October(10)):991.

[7] Pope JE, Stevens A, Howson W, Bell DA. The development of rheumatoid arthri-

tis after recombinant hepatitis B vaccination. The Journal of Rheumatology1998;25(September (9)):1687–93.

[8] Ray P, Black S, Shinefield H, Dillon A, Schwalbe J, Holmes S, et al. Risk of chronicarthropathy among women after rubella vaccination. Vaccine Safety DatalinkTeam. Jama 1997;278(August (7)):551–6.

ne 29 (

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

P. Ray et al. / Vacci

[9] Tingle AJ, Mitchell LA, Grace M, Middleton P, Mathias R, MacWilliam L,et al. Randomised double-blind placebo-controlled study on adverse effectsof rubella immunisation in seronegative women. Lancet 1997;349(May(9061)):1277–81.

10] Jawad AS, Scott DG. Immunisation triggering rheumatoid arthritis? Annals ofthe Rheumatic Diseases 1989;48(February (2)):174.

11] Christensen P. Side reactions to tetanus toxoid. Scientific Publication, vol. 253.Pan American Health Organization; 1972. pp. 36–43; Cited by: Stratton KR,Howe CJ, Johnston RB, Jr. Adverse events associated with childhood vaccines:evidence bearing on causality, Institute of Medicine, Vaccine Safety Committee,Washington, D.C, 1994.

12] Conti F, Rezai S, Valesini G. Vaccination and autoimmune rheumatic diseases.Autoimmunity Reviews 2008;8(2):124–8.

13] Toplak N, Kveder T, Trampus-Bakija A, Subelj V, Cucnik S, Avcin T. Autoimmuneresponse following annual influenza vaccination in 92 apparently healthyadults. Autoimmunity Reviews 2008;8(December (2)):134–8.

14] Bengtsson C, Kapetanovic MC, Kallberg H, Sverdrup B, Nordmark B, KlareskogL, et al. Common vaccinations among adults do not increase the risk of devel-oping rheumatoid arthritis: results from the Swedish EIRA study. Annals of theRheumatic Diseases 2010;69(October (10)):1831–3.

15] Maillefert JF, Sibilia J, Toussirot E, Vignon E, Eschard JP, Lorcerie B, et al.Rheumatic disorders developed after hepatitis B vaccination. Rheumatology(Oxford) 1999;38(October (10)):978–83.

16] Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al.The American Rheumatism Association 1987 revised criteria for the classifi-cation of rheumatoid arthritis. Arthritis and Rheumatism 1988;31(March (3)):315–24.

17] Geier DA, Geier MR. A one year followup of chronic arthritis following rubellaand hepatitis B vaccination based upon analysis of the Vaccine Adverse EventsReporting System (VAERS) database. Clinical and Experimental Rheumatology2002;20(November–December (6)):767–71.

18] Geier DA, Geier MR. A case-control study of serious autoimmune adverseevents following hepatitis B immunization. Autoimmunity 2005;38(June(4)):295–301.

19] Lee K, Hall A. Executive summary of report: Hepatitis B vaccine and utativeassociations with (a) Arthritis (b) Chronic Fatigue Syndrome: WHO; 2005.

20] Alamanos Y, Drosos AA. Epidemiology of adult rheumatoid arthritis. Autoim-munity Reviews 2005;4(March (3)):130–6.

21] Gabriel SE, Crowson CS, O’Fallon WM. The epidemiology of rheumatoidarthritis in Rochester, Minnesota, 1955–1985. Arthritis and Rheumatism1999;42(March (3)):415–20.

[[

2011) 6592– 6597 6597

22] Rangel MC, Shoenbach VJ, Weigle KA, Hogan VK, Strauss RP, Bangdiwala SI.Racial and ethnic disparities in influenza vaccination among elderly adults.Journal of General Internal Medicine 2005;20(May (5)):426–31.

23] Kamal KM, Madhavan SS, Amonkar MM. Determinants of adult influenza andpneumonia immunization rates. Journal of the American Pharmaceutical Asso-ciation 2003;43(May–June (3)):403–11.

24] Toussirot E, Lohse A, Wendling D, Mougin C. Sjogren’s syndrome occurringafter hepatitis B vaccination. Arthritis and Rheumatism 2000;43(September(9)):2139–40.

25] Lasky T, Terracciano GJ, Magder L, Koski CL, Ballesteros M, Nash D,et al. The Guillain-Barre syndrome and the 1992–1993 and 1993–1994influenza vaccines. The New England Journal of Medicine 1998;339(December(25)):1797–802.

26] Schonberger LB, Bregman DJ, Sullivan-Bolyai JZ, Keenlyside RA, Ziegler DW,Retailliau HF, et al. Guillain-Barre syndrome following vaccination in theNational Influenza Immunization Program, United States, 1976–977. AmericanJournal of Epidemiology 1979;110(August (2)):105–23.

27] Albert LJ, Inman RD. Molecular mimicry and autoimmunity. The New EnglandJournal of Medicine 1999;341(December (27)):2068–74.

28] Schattner A. Consequence or coincidence? The occurrence, pathogenesisand significance of autoimmune manifestations after viral vaccines. Vaccine2005;23(June (30)):3876–86.

29] Elkayam O, Yaron M, Caspi D. Safety and efficacy of vaccination against hepati-tis B in patients with rheumatoid arthritis. Annals of the Rheumatic Diseases2002;61(July (7)):623–5.

30] Ascherio A, Zhang SM, Hernan MA, Olek MJ, Coplan PM, Brodovicz K, et al.Hepatitis B vaccination and the risk of multiple sclerosis. The New EnglandJournal of Medicine 2001;344(February (5)):327–32.

31] DeStefano F, Verstraeten T, Jackson LA, Okoro CA, Benson P, Black SB, et al. Vac-cinations and risk of central nervous system demyelinating diseases in adults.Archives of Neurology 2003;60(April (4)):504–9.

32] Bengtsson C, Kapetanovic M, Källberg H, Sverdrup B, Nordmark B, Klareskog L,et al. Common Vaccinations among Adults and the Risk of Developing Rheuma-toid Arthritis; Results from the Swedish EIRA Study. Atlanta, GA: AmericanCollege of Rheumatology Annual Scientific Meeting; 2010.

33] Klein NP, Ray P, Carpenter D, Hansen J, Lewis E, Fireman B, et al. Rates of autoim-

mune diseases in Kaiser Permanente for use in vaccine adverse event safetystudies. Vaccine 2009;(November).

34] Firestein. Kelley’s Textbook of Rheumatology. 8th ed. W.B. Saunders; 2008.35] Fleming A, Crown JM, Corbett M. Early rheumatoid disease. I. Onset. Annals of

the Rheumatic Diseases 1976;35(August (4)):357–60.