Table 1 Characteristics of included studies

Year

2014

2009

2016

2014

2010

2015

2013

2015

2016

Design

Retrospective Matched Cohort

Retrospective Matched Cohort

Retrospective Matched Cohort

Self-Controlled Case Series

Retrospective Matched Cohort

Self-Controlled Case Series

Retrospective Cohort

Retrospective Cohort

Retrospective Matched Cohort

Data Source, Country

The Health Improvement Network (THIN) general practice database, United Kingdom

Taiwan National Health Insurance Research Database (NHIRD), Taiwan

Korea Health Insurance Database, Korea

UK Clinical Practice Research Datalink (CPRD) general practice database, United Kingdom

Taiwan National Health Insurance Research Database (NHIRD), Taiwan

Medicare administrative data claims, USA

Danish Civil Registration System (CRS), Denmark

Västra Götaland County Primary Health Care Register and the Swedish Patient Register, Sweden

Rochester Epidemiology Project, USA

HZ diagnosis period

2002–2010

1997–2001

2003–2013

1987–2012

2003–2004

2006–2011

1995–2008

2008–2010

1986–2011

Number of Cases (HZ and HZO)

106,601

7,760

70,424

6,584

658

42,954

117,926

13,296

4,478

Number of Controls

213,202

23,280

695,755

0

1,974

0

4,503,054

~1,500,000

16,800

Follow-Up Period

24 years

1 year

11 years

1 year

1 year

1 year

14 years

1 year

3 years

Gender, % female

59

52

49

57

51

71

51

60

62

Mean age at zoster diagnosis, years

59

47

41

Median age at stroke onset 77 years

57

80

Not reported

59

68

% Receiving AV Therapy

Not reported

Not reported

Not reported

55%

24% of cases

100%

50%

Not reported

Not reported

Case Inclusion Criteria

Adults 18 years of age or older diagnosed with incident HZ or HZO (index date as recorded in database)

Adults 18 years of age or older presenting to ambulatory clinic with incident HZ or HZO (index date as recorded in database)

Adults 18 years of age or older with incident HZ

Adults 18 years of age or older with incident HZ or HZO and incident stroke

Adults 18 years of age or older presenting to ambulatory clinic with incident HZO (index date as recorded in database)

Adults 65 years of age or older with evidence of incident HZ or HZO and an incident ischemic/nonspecific stroke

Adults 18 years of age or older who received acyclovir 800 mg in packs of 35

All individuals recorded in either of the two databases with incident HZ

Adults 50 years and older with incident HZ

Case Exclusion Criteria

Patients who experienced cardiovascular/stroke event (MI, TIA, stroke) before index date and those with recurrent HZ

Patients who had been diagnosed with stroke before the index date

Patients who had been diagnosed with stroke before HZ

Patients with evidence of HZ, postherpetic neuralgia, or stroke before the study period. Patients with incident episodes of TIA and subarachnoid hemorrhage or risk factors for subarachnoid hemorrhage. Patients with encephalitis 12 months after stroke. Patients with nonspecific cerebral aneurysms.

Patients diagnosed with HZO during the previous 1-year period. Patients diagnosed with any type of stroke prior to index ambulatory care visit. Patients with systemic lupus erythematosus, rheumatoid arthritis, multiple sclerosis, HIV, malignancy, and use of steroids or immunosuppressants for more than 1 month within 1 year prior to index date.

Individuals with evidence of vascular events or HZ before observation period. Secondary inpatient diagnosis of zoster or cardiovascular events. Individuals with subarachnoid hemorrhage (or established risk factors of) or encephalitis diagnosed up to 12 months after stroke

Patients who received a second acyclovir prescription of same strength and pack size. Patients with a stroke or TIA diagnoses before the start of the follow up period as per the ICD10 codes below as well as ICD8 codes 430-438

Patients diagnosed with HZ during the previous 1-year

Recurrent herpes simplex infection, history of stroke great than 1 month before index date, MI before index date

HZ Definition

Read codes corresponding to diagnoses of incident HZ (index date as recorded in database)

Incident HZ diagnosis using ICD9 codes 053x (index date as recorded in database)

ICD10 codes corresponding to diagnoses of incident HZ (index date as recorded in database)

Read or ICD10 codes corresponding to diagnoses of incident HZ

N/A

HZ diagnosis using ICD-9 codes 053x

AND antiviral therapy 7 days before or after diagnosis

Patients who filled a single prescription of acyclovir 800 mg for 35 tablets

Incident HZ diagnosis using ICD-10 code B02.X

Incident HZ diagnosis using ICD-9 codes, and confirmed by medical record review

HZO Definition

Read codes corresponding to diagnoses of incident HZO

Incident HZO diagnosis using ICD9 code 053

N/A

Read codes corresponding to diagnoses of incident HZO

Incident HZO diagnosis based on the ICD9 code 053.2

HZO diagnosis using ICD-9 codes

AND antiviral therapy 7 days before or after diagnosis

N/A

N/A

N/A

Stroke Definition

Read codes corresponding to stroke, diagnosis

Stroke/TIA diagnosis as per ICD9 codes 430-438

Stroke/TIA diagnosis as per ICD10 codes

Read or ICD10 codes corresponding to stroke diagnosis

Stroke/TIA diagnosis as per ICD9 codes 430-438

Ischemic/nonspecific stroke diagnosis as per ICD9 codes 436, 433x1, or 434x1

Diagnosis in hospital of stroke/TIA as per ICD10 I60-64 and G45

Stroke diagnosis as per ICD-10 codes I61-I64 (excluding I62)

Stroke diagnosis using ICD-9 codes

Control selection

Patients who had no record of HZ, matched (2:1) by age (+/−2 years), sex, and general practice

Patients with no HZ or stroke before 2001, matched (3:1) on age and sex, and defined their index date as their first ambulatory care visit in 2001

Patients without HZ, matched on age group

Self controlled

Selected from remaining patients, matched (3:1) on age group and gender, and defined their index date as their first ambulatory care visit in 2004

Self controlled

Patients who had no prior history of acyclovir, valacyclovir, or famciclovir use (as a proxy for unexposed)

Total remaining population without HZ

Matching each patient with HZ with up to 4 patients whose birthday was +/− 1 year, who were the same sex, and no HZ in the past 5 years

Confounders (Adjusted for)

Age, sex, obesity, smoking, high cholesterol recording, hypertension, diabetes, ischemic heart disease, atrial fibrillation, intermittent arterial claudication, carotid stenosis, and valvular heart disease

Age, sex, hypertension, diabetes, coronary heart disease, hyperlipidemia, renal disease, atrial fibrillation, heart failure, heart valve/myocardium disease, carotid/peripheral vascular disease, monthly income, urbanization level, and geographical region

Age, male gender, hypertension, hyperlipidaemia, ischaemic heart disease, diabetes,

heart failure, peripheral vascular disease, arterial fibrillation or atrial flutter, renal disease and valvular

heart disease

Confounders are implicitly controlled for due to study design

Age, sex, hypertension, diabetes, hyperlipidemia, coronary heart disease, chronic rheumatic heart disease, other forms of heart disease, and medication habits

Confounders are implicitly controlled for due to study design

Age, sex, calendar period, acute MI, atrial fibrillation, education, cancer, medications (antihypertensives, drugs used to treat dyslipidemia and atrial fibrillation, immunosuppressive drugs)

Age and sex

Age, sex, hypertension, dyslipidemia, coronary artery disease (including MI), arrhythmias, congestive heart failure, diabetes, depression, chronic obstructive pulmonary disorder, vasculopathies, stroke, and anxiety

Relevant study outcomes

Stroke HR after >1 year follow up, with analyses stratified by age <40 and ≥40 years, and by HZO.

Stroke HR after 1 year follow up since HZ, with analyses stratified by age (<45 and ≥ 45 years), gender and by HZO

Stroke HR after 11 years of follow up, stratified by age 18–30, 30–40, 40–50, 50–60, 60–70, and >70 years

Stroke IR 1–4, 5–12, 13–26, and 27–52 weeks after HZ or HZO +/−head and

neck involvement, with analyses stratified by those

who received antiviral therapy and by HZO

Stroke HR after 1 year follow up since HZO, with analyses stratified by those

who received antiviral therapy

Stroke IR at 1, 2–4, 5–12, 13–26, and 27–52 weeks since HZ diagnosis, with analyses stratified by gender, HZO

Stroke IRR after <2 weeks, 2–52 weeks, and >1 year follow up since HZ, with analyses stratified by age <40, 40–59, and ≥60 years and by gender

Stroke IRR after 1 year follow up, with analyses stratified by age <40, 40–49, 50–59, 60–69, 70–79, and ≥80 years and gender

Stroke OR 3 and 6 months and 1 and 3 years after zoster.