Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Clin Infect Dis. Author manuscript; available in PMC 2010 May 15.
Published in final edited form as:
PMCID: PMC2743911

Herpes zoster risk factors in a national cohort of veterans with rheumatoid arthritis



Herpes zoster occurs more commonly in patients taking immunosuppressive medications, though the risk associated with different medications is poorly understood.


Retrospective cohort study including 20,357 patients who were followed in the Veterans Affairs healthcare system and treated for rheumatoid arthritis from October 1998 through June 2005. Cox proportional hazards regression was used to determine risk factors for herpes zoster, and herpes zoster-free survival. Chart review was performed to validate the diagnosis of herpes zoster.


The incidence of herpes zoster was 9.96 per 1000 patient-years. In time-to-event analysis, patients receiving medications used to treat mild rheumatoid arthritis were less likely to have an episode of herpes zoster than patients receiving medications used to treat moderate and severe rheumatoid arthritis (p<0.001). Independent risk factors for herpes zoster included older age, prednisone use, medications used to treat moderate and severe rheumatoid arthritis, malignancy, chronic lung disease, renal failure, and liver disease. Among patients receiving tumor necrosis factor-alpha antagonists, etanercept (HR 0.62) and adalimumab (HR 0.53) were associated with lower risk of herpes zoster. There was excellent agreement between ICD-9-CM diagnosis of herpes zoster and diagnosis by chart review (kappa = 0.92).


Risk factors for herpes zoster included older age, prednisone use, medications used to treat moderate and severe rheumatoid arthritis, and several comorbid medical conditions. These results demonstrate that the Department of Veterans Affairs’ national administrative databases can be used to study rare adverse drug events.

Keywords: Herpes zoster, Adverse drug event, Patient safety, Rheumatoid arthritis


Herpes zoster (HZ), a reactivation of latent varicella-zoster virus infection, causes substantial morbidity, especially among elderly and immunocompromised patients [1]. Established risk factors for HZ include older age and immunosuppressive medications [24]. Other potential risk factors have been identified by some studies but not others, including female gender [3,5,6], malignancy [7,8], kidney disease [6,8], and acquired immune deficiency syndrome (AIDS) [9].

While it has been established that immunosuppressive medications contribute to excess risk of HZ, the comparative risk of HZ among different immunosuppressive medications is unclear. Enhanced understanding of the risk of HZ associated with different classes of immunosuppressive medications may provide clinicians with useful information as they prescribe immunomodulatory drugs to individual patients. This newfound knowledge may also influence the evaluation and implementation of HZ immunization strategies in patients prior to receiving immunosuppressive medications.

Adverse drug events, such as infection after immunosuppression, are between the fourth and sixth most common cause of death in the United States [10]. Under the Food and Drug Administration’s current drug approval process, pre-marketing clinical trials for a new drug typically include 500 to 3000 exposed patients. This allows identification of common adverse drug events, but not rare ones [11]. Current systems for detection and investigation of rare adverse drug events in the United States are inadequate [12]. A 2007 report by the Institute of Medicine emphasized the need to improve post-marketing surveillance of adverse drug events by increasing and improving programs that utilize data from large automated healthcare databases [13]. The Department of Veterans’ Affairs (VA) national administrative databases present an ideal opportunity to characterize and investigate rare adverse drug events. In this study, we used the VA’s national administrative databases to examine HZ risk, risk factors, treatments and outcomes in a large national cohort of veterans with rheumatoid arthritis (RA), with a particular focus on the contribution of different classes of immunosuppressive medications to the risk of HZ.


This study was approved by the institutional review boards of the participating institutions.


The Austin Automation Center (AAC) is the VA’s centralized repository of administrative data. We obtained all inpatient and outpatient International Classification of Diseases, Version 9, Clinical Modification (ICD-9-CM) diagnosis codes, encounter data, and demographic data from the inpatient and outpatient SAS datasets of the AAC. We obtained inpatient and outpatient pharmacy data on our patients from the VA’s Pharmacy Benefits Management (PBM) database. Data from AAC and PBM were merged into a single database using common unique identifiers.

Study sample

The study period was October 1, 1998 to June 30th 2005. The study sample included all veterans who had an ICD-9-CM code diagnosis of RA during the study period and who, after at least a 4-month history of receiving medications from the VA during the study period, subsequently received a first prescription for a disease-modifying anti-rheumatic drug (DMARD).

Patients were excluded who: 1) had a diagnosis of HZ at any time prior to receiving a DMARD, or 2) did not have at least two separate outpatient or inpatient clinical encounters during the study period.

Patient records were censored at the end of the last prescribed medication course, the last clinical encounter, or at the time of first occurrence of HZ, whichever came last. Thus recurrent HZ was excluded from our analyses. Uncensored patients were followed through December 31st, 2006.



Based on an algorithm validated by Singh et al [14], the diagnosis of RA required both: 1) occurrence of an ICD-9-CM code for RA on at least one occasion in either the inpatient or outpatient record, and 2) the receipt of a prescription for a DMARD on at least one occasion. See appendix for ICD-9-CM codes accepted for definition of RA.

Medication group

RA treatments were subdivided into medication groups, based on their place in the armamentarium of RA therapeutics. Group 1 (treatment of mild disease) included hydroxycholoroquine, sulfasalazine, auranofin, injectable gold, and penicillamine. Group 2 (treatment of moderate disease) included methotrexate, leflunomide, azathioprine, cyclophosphamide, cyclosporine, and anakinra. Group 3 (treatment of severe disease) included the tumor necrosis factor-alpha (TNF) antagonists (etanercept, infliximab, and adalimumab), which are typically used after failure to respond adequately to medications in groups 1 and 2. Though treatment of RA is highly individualized, existing RA treatment guidelines support this classification system [15,16].


DMARDs were defined as all medications in medication groups 1, 2, and 3.

Patient-time in medication group

Time zero was defined as the date of the first prescription for a medication in a given medication group. The patient continued in that medication group until: 1) a medication from a higher-numbered medication group was prescribed or 2) censorship. Data from different discrete time periods within a single patient record could be included in different medication groups, if the patient’s medication group number increased over time. A patient could only be in one medication group at any point in time.


The first occurrence in the study period of an ICD-9-CM code accepted for the definition of HZ (see appendix), after at least one clinical encounter in the study period without such an ICD-9-CM code.

Hospitalization for HZ

Hospitalization for which an ICD-9-CM code for HZ is in the field indicating the diagnosis primarily responsible for hospitalization.

HZ complications

See appendix for accepted ICD-9-CM codes.

HZ treatments

Oral acyclovir, valacyclovir, famciclovir, and intravenous acyclovir were defined as HZ treatment if they were prescribed within 90 days before or after HZ diagnosis.

Comorbid medical conditions

We used ICD-9-CM code definitions developed by Elixhauser et al for use with administrative data.17 Our variable for malignancy was a composite of ICD-9-CM codes used to define metastatic cancer, solid tumor without metastases, and lymphoma [17].


The accuracy of ICD-9-CM diagnosis of HZ was measured against the gold standard, medical record review by a physician. Of patients in our cohort from the St. Louis VA, medical records were randomly selected for review using a random number generator. Of 101 patients with an HZ diagnosis code, 50 were reviewed, and of 3,860 without an HZ diagnosis code, 150 were reviewed. Medical record review was performed by one author (JRM), using a methodology described in the appendix.

Data analyses

Incidence of HZ was calculated as number of events per 1000 patient-years. For descriptive and bivariate analysis, dichotomous variables were analyzed using chi-square test and Fisher’s exact test where appropriate. Continuous variables were analyzed using student’s t-test. A two-sided p-value of <0.05 was considered statistically significant. Risk of outcomes was described using hazard ratios and 95% confidence intervals. Time-to-event analysis was performed using Cox proportional hazards regression. In regression modeling, membership in each medication group was modeled as a time-dependent dummy variable to account for the change in medication groups over time. Each drug was modeled separately and adjusted for age, gender, race, and time-dependent comorbid diagnoses. The race variable as it was recorded in the AAC has been shown to correlate poorly with patient self-report [18]. Despite this limitation, we obtained this variable and included it in our analyses in order to control for it. Comorbidities were treated as time-dependent variables. For validation of the HZ diagnosis, agreement between ICD-9-CM code and the gold standard was calculated using the kappa statistic. All analyses were performed using SAS software version 6.12 (SAS Institute, Cary, NC). The time-to-event graph was creating using R software version 2.5.1 (R Foundation, Vienna, Austria).


There were 20,816 patients who met our inclusion criteria. After applying exclusion criteria, our study cohort consisted of 20,357 patients with more than 26 million patient-days (Figure 1).

Figure 1
Study Flow Diagram.

Patient demographics, comorbid medical conditions, and RA treatments are shown in Table 1. Among the 28.4% of our sample with malignancy, the most common type of malignancy was prostate cancer (29%). The frequency of malignancy in our sample is consistent with the frequency seen in other administrative datasets [17], after considering the additional cases found when outpatient claims data are added [19]. Only 12.0% of patients had none of the examined comorbid conditions.

Table 1
Demographic and Clinical Characteristics

There were 713 episodes of HZ in the study cohort, with an overall incidence of 9.96 episodes per 1000 patient-years. Incidence of HZ was significantly higher in medication group 2 compared to medication group 1 (11.18 per 1000 patient-years vs. 8.00 per 1000 patient-years, p<0.0001), and in medication group 3 compared to medication group 1 (10.60 per 1000 patient-years vs. 8.00 per 1000 patient-years; p<0.0001). Incidence was similar between medication groups 2 and 3. In time-to-event analysis (Figure 2), patients in each of medication groups 2 and 3 had shorter HZ-free survival than group 1. HZ-free survival was significantly different between medication groups (p<0.001).

Figure 2
After adjusting for demographic data and comorbid medical conditions, time-to-herpes zoster was different among groups 1, 2, and 3 (p<0.001).

In the study cohort, 66.0% of patients with HZ were treated with oral antiviral medications, and 4.6% received intravenous acyclovir (Table 2). Hospitalization for HZ occurred in 4.9% of patients, and incidence was similar between medication groups.

Table 2
Treatments and Outcomes for Patients with Herpes Zoster (HZ)

Independent risk factors for HZ included older age, prednisone, medications in groups 2 and 3, malignancy, chronic lung disease, renal failure, and liver disease (Table 3). The hazard ratio of group 2 medications was similar to the hazard ratio of group 3 medications (1.34 vs. 1.38, p=0.67).

Table 3
Multivariate Risk Factors for Herpes Zoster (HZ)

There were 96 patients with incident HZ among the 3,661 patients prescribed TNF antagonists. Of these, 59 occurred on etanercept, 33 occurred on infliximab, and 4 on adalimumab. Among the TNF antagonists, etanercept (HR 0.62, 95% CI 0.40–0.95) and adalimumab (HR 0.53, 95% CI 0.31–0.91) were associated with lower risk of HZ compared to infliximab (Table 4).

Table 4
Risk Factors for Herpes Zoster (HZ) Among Patients Receiving Tumor Necrosis Factor-Alpha Antagonists (Medication Group 3)

Medical record review to validate the ICD-9-CM code for HZ demonstrated that 45 of the 50 patients with an HZ ICD-9-CM code had an acute episode of HZ in the 90 days prior to receiving the code, for a positive predictive value of 90%. All five of those who did not have an acute episode of HZ in the prior 90 days had had HZ in the past. One of the 150 patients without an HZ ICD-9-CM code (0.67%) had an episode of acute zoster during the study period, for a negative predictive value of 99.3%. The kappa statistic for agreement between the ICD-9-CM code diagnosis and the medical record diagnosis was 0.92, indicating excellent agreement.


In this large sample of patients in the VA healthcare system with RA, we analyzed HZ-free survival as well as outcomes, treatments, and risk factors for HZ. While the occurrence of HZ in a patient with RA is relatively uncommon, we were able to identify 713 such occurrences among 20,357 patients, in more than 71 thousand patient-years, even after applying stringent inclusion criteria designed to ensure the quality of our data and the validity of our conclusions. We demonstrated in this population that the risk of HZ while taking TNF antagonists is similar to the risk while taking group 2 medications, and that the risk associated with infliximab exceeds the risk associated with other TNF antagonists.

HZ incidence was highest among patients taking group 2 and 3 medications, and the risk was similar between the two groups. Prior studies have conflicting results in this area. Listing et al studied 1,529 patients with RA receiving DMARDs, and demonstrated that the risk of herpes virus infections was similar among patients receiving TNF antagonists and those receiving other DMARDs [20]. The small study size (only 17 episodes of HZ were seen) made it unlikely that significant differences would be detected. Wolfe et al studied 10,614 patients with RA, and found that cyclophosphamide, azathioprine, prednisone, leflunomide, and some non-steroidal anti-inflammatory drugs were risk factors for HZ, but that TNF antagonists and methotrexate were not [4]. However, data on both medication exposure and HZ incidence were collected by self-administered patient questionnaire, which are likely to be less temporally accurate than electronic prescription and diagnosis data. In addition, our study includes more than twice as many patient-years as this study. Smitten et al studied a database of more than 160,000 commercially-insured patients, and found that biologic DMARDs were associated with higher risk of HZ than other DMARDs; however, this study did not differentiate between groups of non-biologic DMARDs and did not include several covariates known to impact HZ risk [21].

The incidence of HZ in our sample (9.96 per 1000 patient-years) was similar to the incidence seen in other studies. Donahue et al [22] and Insinga et al [23] found incidences of 2.2 and 3.2 per 1000 patient-years, respectively, in administrative databases sampling the general US population, not specifically selected for RA. Studies examining cohorts of patients with RA found incidences similar to ours: A study using an administrative database of commercially-insured RA patients in the United States found incidence of 9.83 per 1000 patient-years [21], and study using a registry of patients with RA identified an incidence of 13.2 per 1000 patient-years [4]. The higher incidence of HZ in the latter study may be explained by a higher rate of infliximab use among patients in that study, or by overestimates of HZ based on patient self-report. While it is possible that inaccuracies in diagnosis coding explain the difference, our validation results suggest that our incidence estimate is robust for this population.

We found that older age and certain comorbid medical conditions (malignancy, chronic lung disease, renal failure, and liver disease) were independent risk factors for HZ. Older age and cell-mediated immune defects are well-described risk factors for HZ [24]. Prior studies have described malignancy [7,8], renal disease [6,8], and chronic lung disease [24] as HZ risk factors. Female gender was not found to be an HZ risk factor in our study. Some prior studies have identified female gender as a risk factor [3,5,6], possibly due to gender-specific patterns of healthcare utilization [24]. Our result may differ because of the inclusion of previously-ignored covariates, differences in gender-specific healthcare utilization between female veterans and non-veterans, or because the low proportion of women in our sample (9.5%) made that part of our analysis relatively underpowered.

Among patients on TNF antagonists, adalimumab and etanercept use were significantly associated with lower risk of HZ. This finding is consistent with those of previous studies, which show higher rates of a composite endpoint of infection [20] as well as tuberculosis and other granulomatous infections [25] among patients who received infliximab compared to etanercept. There are several differences in the properties of TNF antagonists that may explain differences in infection risk. In contrast to etanercept, infliximab binds to both soluble and transmembrane forms of TNF, induces apoptosis of monocytes and T-cells, and induces expression of different leukocyte genes [2527].

This study has significant strengths. The large size and national scope of this database allows not only the capture of a large number of rare events, but also the application of strict inclusion criteria to maximize the quality of the data. We limited our sample to patients with multiple VA clinical encounters, at least 4 months of drug prescriptions excluding DMARDs prior to DMARD initiation, and no HZ diagnosis prior to first DMARD prescription. Our data spans an eight-year period, allowing for the identification of events that occur long after drug initiation. Because our patients comprise a real-world cohort that was not subject to the eligibility criteria often present in randomized controlled trials of therapeutic agents, our data are generalizeable to similar real-world populations.

Our study is limited by the fact that we did not include medication dose in our analysis. Duration, however, was taken into account in our survival models. With 14 different DMARDs plus prednisone, the inclusion of dose variability in the analysis would have added significant complexity to the interpretation of the study, and detracted from our primary aims. Our population was more than 90% male, and thus our results may not be generalizable to a female population. Although administrative data are imperfect sources of clinical information, they are very useful for large scale epidemiologic research. The validity of different ICD-9 codes varies widely [28,29]. Our validation demonstrates that the diagnosis of HZ in our data is very accurate, and our approach to RA diagnosis has also been shown to be accurate in prior studies [14].

RA severity can impact risk of infection [30]. We did not evaluate the impact of RA severity on HZ risk, because there is no accepted tool for measuring RA severity from administrative data. One study evaluating the relative contributions of RA treatment and RA severity to infection risk found that approximately one-third of the excess risk of infection was related to RA severity, while two-thirds was related to treatment [20]. Thus, the excess infection seen in higher medication groups was likely caused in part by medication, and in part by RA severity.

A varicella-zoster virus vaccine that prevents HZ is currently available in the US and recommended for use in individuals who are 60 years of age or older. Because of the risk of disseminated disease caused by the live attenuated vaccine strain of the virus, it is not recommended for patients receiving higher-dose immunosuppressive therapy, though the Advisory Committee on Immunization Practices has recently advocated the safety of the vaccine in selected patients on lower-dose corticosteroids and some DMARDs, and at least 14 days prior to planned immunosuppression [2,31,32]. However, clinical data supporting these recommendations are sparse, and our results on comparative risk of infection among treatment types may inform future efforts to target this vaccine to populations most likely to benefit.

This study describes the risk of HZ associated with specific medications used to treat RA. These data may inform clinical decision-making in prescribing treatment for RA, as well as future HZ vaccine testing and targeting in immunosuppressed populations. Our data show that the administrative databases of the Department of Veterans’ Affairs can be a useful source of data for the identification and analysis of rare adverse drug events.


All authors have no conflicts of interest. The research reported here was supported by the Department of Veterans Affairs, Veterans Health Administration, Health Services Research and Development Service project number IAF 06–026, and NIH K12RR023249 and KL2RR024994. The views expressed in this article do not necessarily represent the views of the Department of Veterans Affairs.


Austin Automation Center, a centralized repository of administrative data within the Department of Veterans’ Affairs
Acquired Immune Deficiency Syndrome
Computerized Patient Records System, the VA’s electronic health record
Disease-modifying Anti-rheumatic Drug
Herpes Zoster
International Classification of Diseases, Version 9, Clinical Modification, a system of assigning coded medical diagnoses to clinical encounters
Pharmacy Benefits Management database, a database containing prescription pharmacy data for the Department of Veterans’ Affairs
Rheumatoid arthritis
Department of Veterans’ Affairs


ICD-9 codes accepted for diagnosis of RA

RA (714.0)

Felty’s syndrome (714.1)

RA with visceral or systemic involvement (714.2)

Rheumatoid lung (714.81)

ICD-9 codes accepted for diagnosis of herpes zoster

Herpes zoster (053)

Herpes zoster with meningitis (053.0)

Herpes zoster with other nervous system complications - nonspecific (053.1)

Herpes zoster with unspecified nervous system complications (053.10)

Geniculate herpes zoster (053.11)

Postherpetic trigeminal neuralgia (053.12)

Postherpetic polyneuropathy (053.13)

Herpes zoster myelitis (053.14)

Herpes zoster with other nervous system complications - specific (053.19)

Herpes zoster with ophthalmic complications (053.2)

Herpes zoster dermatitis of eyelid (053.20)

Herpes zoster keratoconjunctivitis (053.21)

Herpes zoster iridocyclitis (053.22)

Herpes zoster with other ophthalmic complications (053.29)

Herpes zoster with other specified complications (053.7)

Otitis externa due to herpes zoster (053.71)

Herpes zoster with other specified complications (053.79)

Herpes zoster with unspecified complication (053.8)

Herpes zoster without complication (053.9)

ICD-9 codes accepted for diagnosis of complications of herpes zoster

All of the codes accepted for the diagnosis of herpes zoster except Herpes zoster (053) and Herpes zoster without complication (053.9)

Medical record review methodology for validation

Each patient’s computerized medical record in the VA’s CPRS (Computerized Patient Record System) was electronically queried using the CPRS search function for the following text strings: “zost”, “herp”, “shingl”, “HZ”, “VZ”, and “varicel”. Any note containing these text strings was manually reviewed for evidence of HZ diagnosis. HZ was defined as health care provider report of “shingles,” “zoster,” “herpes zoster,”, “HZ,” or “varicella reactivation” in 90 days prior to the code. “Probable” and “likely” diagnoses were accepted, while “possible,” “questionable,” and diagnoses included in a differential diagnosis list were not accepted.


1. Wareham DW, Breuer J. Herpes zoster. BMJ. 2007;334:1211–1215. [PMC free article] [PubMed]
2. Arvin A. Aging, immunity, and the varicella-zoster virus. N Engl J Med. 2005;352:2266–2267. [PubMed]
3. Thomas SL, Hall AJ. What does epidemiology tell us about risk factors for herpes zoster? Lancet Infect Dis. 2004;4:26–33. [PubMed]
4. Wolfe F, Michaud K, Chakravarty EF. Rates and predictors of herpes zoster in patients with rheumatoid arthritis and non-inflammatory musculoskeletal disorders. Rheumatology. 2006;45:1370–1375. [PubMed]
5. Chidiac C, Bruxelle J, Daures J-P, et al. Characteristics of patients with herpes zoster on presentation to practitioners in France. Clin Infect Dis. 2001;33:62–69. [PubMed]
6. Wung PK, Holbrook JT, Hoffman GS, et al. Herpes zoster in immunocompromised patients: Incidence, timing, and risk factors. Am J Med. 2005;118:e9–e18. 1416. [PubMed]
7. Schimpff S, Serpick A, Stoler B, et al. Varicella-zoster infection in patients with cancer. Ann Intern Med. 1972;76:241–254. [PubMed]
8. Manzi S, Kuller LH, Kutzer J, et al. Herpes zoster in systemic lupus erythmatosus. J Rheumatol. 1995;22:1254–1258. [PubMed]
9. Buchbinder SP, Katz MH, Hessol NA, et al. Herpes zoster and human immunodeficiency virus infection. J Infect Dis. 1992;166:1153–1156. [PubMed]
10. Lazarou J, Pomeranz BH, Corey PN. Incidence of adverse drug reactions in hospitalized patients: A meta-analysis of prospective studies. JAMA. 1998;279:1200–1205. [PubMed]
11. Strom BL. Potential for conflict of interest in the evaluation of suspected adverse drug reactions. JAMA. 2004;292:2643–2646. [PubMed]
12. Keystone EC. Safety of biologic therapies — an update. J Rheumatol Suppl. 2005;32:8–12. [PubMed]
13. Baciu A, Stratton K, Burke SP, editors. Washington DC: National Academies Press; 2007. The future of drug safety: promoting and protecting the health of the public.
14. Singh JA, Holmgren AR, Noorbaloochi S. Accuracy of Veterans Administration databases for a diagnosis of rheumatoid arthritis. Arthritis Rheum. 2004;51:952–957. [PubMed]
15. American College of Rheumatology Subcommittee on Rheumatoid Arthritis Guidelines. Guidelines for the management of rheumatoid arthritis: 2002 update. Arthritis Rheum. 2002;46:328–346. [PubMed]
16. Ledingham J, Deighton C. Update on the British Society for Rheumatology guidelines for prescribing TNF-alpha blockers in adults with rheumatoid arthritis (update of previous guidelines of April 2001) Rheumatology. 2005;44:157–163. [PubMed]
17. Elixhauser A, Steiner C, Harris DR, Coffey R. Comorbidity measures for use with administrative data. Med Care. 1998;36:8–27. [PubMed]
18. Kressin NR, Chang BH, Hendricks A, Kazis LE. Agreement between administrative data and patients’ self-reports of race/ethnicity. Am J Public Health. 2003;93:1734–1739. [PubMed]
19. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol. 2000;53:1258–1267. [PubMed]
20. Donahue JG, Choo PW, Manson JE, Platt R. The incidence of herpes zoster. Arch Intern Med. 1995;155:1605–1609. [PubMed]
21. Insinga RP, Itzler RF, Pellissier JM, Saddier P, Nikas AA. The incidence of herpes zoster in a United States administrative database. J Gen Intern Med. 2005;20:748–753. [PMC free article] [PubMed]
22. Smitten AL, Choi HK, Hochberg MC, Suissa S, Simon TA, Testa MA, Chan KA. The risk of herpes zoster in patients with rheumatoid arthritis in the United States and the United Kingdom. Arthritis Rheum. 2007;57:1431–1438. [PubMed]
23. Ragozzino MW, Melton LJ, III, Kurland LT, Chu CP, Perry HO. Population-based study of herpes zoster and its sequelae. Medicine. 1982;61:310–316. [PubMed]
24. Listing J, Strangfeld A, Kary S, et al. Infections in patients with rheumatoid arthritis treated with biologic agents. Arthritis Rheum. 2005;52:3403–3412. [PubMed]
25. Wallis RS, Broder MS, Wong JY, Hanson ME, Beenhouwer DO. Granulomatous infectious diseases associated with tumor necrosis factor antagonists. Clin Infect Dis. 2004;38:1261–1265. [PubMed]
26. Gardam MA, Keystone EC, Menzies R, et al. Anti-tumour necrosis factor agents and tuberculosis risk: mechanisms of action and clinical management. Lancet Infect Dis. 2003;3:148–155. [PubMed]
27. Haider AS, Cardinale IR, Whynot JA, Krueger JG. Effects of etanercept are distinct from infliximab in modulating proinflammatory genes in activated human leukocytes. J Invest Dermatol. 2007;12:9–15. [PubMed]
28. O’Malley KJ, Cook KF, Price MD, Wildes KR, Hurdle JF, Ashton CM. Measuring diagnoses: ICD code accuracy. Health Serv Res. 2005;40:1620–1639. [PMC free article] [PubMed]
29. Kashner TM. Agreement between administrative files and written medical records: A case of the Department of Veterans Affairs. Med Care. 1998;36:1324–1336. [PubMed]
30. Doran MF, Crowson CS, Pond GR, O’Fallon WM, Gabriel SE. Frequency of infection in patients with rheumatoid arthritis compared with controls: a population-based study. Arthritis Rheum. 2002;46:2287–2293. [PubMed]
31. Oxman MN, Levin MJ, Johnson GR, et al. A vaccine to prevent herpes zoster and postherpetic neuralgia in older adults. N Engl J Med. 2005;352:2271–2284. [PubMed]
32. Centers for Disease Control and Prevention. Prevention of herpes zoster: Recommendations of the Advisory Committee on Immunization Practices. MMWR Early Release. 2008;57:1–40. [PubMed]