Jaw-Shyang Jih1,2, Yi-Ju Chen1,3, Ming-Wei Lin4,5, Yu-Chun Chen6,7, Tzeng-Ji Chen6,7, Yu-Lin Huang1,2, Chih-Chiang Chen1,2, Ding-Dar Lee1,2, Yun-Ting Chang1,2,8, Wen-Jen Wang1,2 and Han-Nan Liu1,2,9

Departments of 1Dermatology, 4Institute of Public Health, 6Faculty of Medicine, National Yang-Ming University, Departments of 2Dermatology, 5Medical Research and Education, 7Family Medicine, Taipei Veterans General Hospital, 3Department of Dermatology, Taichung Veterans General Hospital, Taichung, 8Department of Dermatology, National Yang Ming University Hospital, I-Lan and 9Department of Dermatology, National Defence Medical Center, Taipei, Taiwan

To analyse the epidemiological characteristics and related costs of herpes zoster in Taiwan, a nationally representative cohort of 1,000,000 individuals from the National Health Insurance register was followed up from 2000 to 2006 and their claims data analysed. Overall, 34,280 patients were diagnosed with zoster (incidence 4.89/1000 person-years) and 2944 patients (8.6%) developed post-herpetic neuralgia 3 months after the start of the zoster rash (incidence 0.42/1000 person-years). People with older age, diabetes, and immunocompromising conditions were at higher risk of developing zoster and post-herpetic neuralgia. The overall hospitalization rate for zoster was 16.1 cases per 100,000 person-years. The cost for each home care case and per hospitalized case were approximately €53.30 and €1224.70, respectively. Further research into the cost-effectiveness of zoster vaccine is needed. Key words: cost; epidemiology; herpes zoster; post-herpetic neuralgia.

(Accepted June 8, 2009.)

Acta Derm Venereol 2009; 89: 612–616.

Yun-Ting Chang, Department of Dermatology, Taipei Veterans General Hospital, Shih-Pai, 112 Taipei, Taiwan. E-mail: ytchang@vghtpe.gov.tw

Herpes zoster (HZ) or shingles is a clinical manifestation caused by activation of the varicella-zoster virus (VZV) that has remained latent in the sensory ganglia and dorsal nerve roots following varicella infection. This disease is characterized by a unilaterally grouped vesicular rash with radicular pain, which is generally limited to a single dermatome (1, 2). The estimated lifetime risk of developing zoster in those exposed to varicella is 10–30%, while the incidence and severity of HZ increases with age; more than 50% of all persons in whom HZ develops are older than 60 years (3–5). The most common complication of HZ is post-herpetic neuralgia (PHN), which has been variably defined as any pain one month, 3 months, or 4 months after rash onset (6, 7). Possible risk indicators for the occurrence of PHN include older age, female sex, presence of a prodrome, greater rash severity, and greater acute pain (6).

In Europe and the USA, the incidence of HZ is 1.2–4.8 cases per 1000 person-years (8). However, to the best of our knowledge, the epidemiology of HZ and PHN has not been well investigated in Asia. In Taiwan, the National Health Insurance (NHI) programme covers most of the population (the coverage rate in 2000 was 96.16%). Most medical institutions (93%) have been contracted to the Bureau of NHI. Inhabitants of Taiwan are free to choose Western medicine or traditional Chinese herbal medicine, and can choose to visit either public or private medical facilities (9). Therefore, information from the NHI database is considered appropriate for assessment of epidemiological features of HZ in Taiwan. The objectives of this study are to estimate the epidemiological characteristics of HZ in Taiwan, to describe the Western and Chinese herbal management of cases, and to calculate the costs of home care and hospital care based on the NHI reference costs.

PATIENTS AND METHODS

We conducted a retrospective cohort study based on the NHI programme. The NHI programme was initiated in Taiwan in 1995 and covers almost all of the population (21,653,555 beneficiaries at the end of 2001, equivalent to a coverage rate of 96.6%). In 1999, the Bureau of NHI began to release all claims data in electronic format to the public under the National Health Insurance Research Database (NHIRD) project, and NHIRD has been used extensively in many epidemiological studies (10, 11).

In the present study, a total of 1,000,000 persons (approximately 5% of Taiwan’s population), were randomly selected from Taiwan’s NHIRD. Data from 495,816 men (49.5%) and 504,184 women (50.4%) were evaluated. Attrition of the study cohort was observed due to various reasons such as mortality and emigration. We linked, through individuals’ personal identification numbers, to both the ambulatory care and inpatient claims in order to identify all cases of HZ. The NHI electronic data files provided patient identification numbers, gender, date of birth, diagnostic codes, prescription drugs dispensed, medical cost, medical care facilities and specialties. Cases were defined with the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD9-CM) code as zoster (from 053.0 to 053.9) present in either an inpatient or outpatient service claim. We used the claims data for the years 2000 to 2006 to investigate the incidence rate, demographic characteristics, treatment modalities, risk factors, and economic burden of patients with HZ. PHN was defined as visiting a physician again with a coding of zoster more than 90 days after the first onset in addition to receiving treatment for neuralgia.

For each cohort year, the incidence was calculated as the number of patients with HZ divided by the total population of that cohort year. After analysing our database and reviewing documents and journals, the following diseases were considered as possible co-morbid diseases associated with HZ: diabetes mellitus (DM) (ICD-9 code 250), systemic lupus erythematosus (SLE) (ICD-9 code 710), human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) (ICD-9 codes 042), breast cancer (ICD-9 code 174-175), liver cancer (ICD-9 code 155), and lymphoma//leukaemia (ICD-9 code 200–202 and 204–208).

The NHI provides comprehensive coverage, including out-patient/in-patient care, laboratory tests, prescription drugs, etc. The additional expense of NHI enrollees includes payment for services not covered by the NHI as well as “user fees” and “co-payments” for NHI-covered services. The medical care expenditure data used in the present study include medical expenditures of outpatient services and hospitalization of the study subjects. All medical costs presented in the study were then converted from Taiwan dollars ($NT) to Euros using an exchange rate of 31.07:1, based on the average exchange rate during 2000 to 2006. The study was approved by the institution review board of the hospital.

Data were analysed using SAS statistical software (Version 8.2; SAS Institute, Cary, NC, USA), SPSS (Version 15; SPSS Inc.) and Excel (Microsoft Office 2007). We used a Poisson regression model with age, sex, and each possible associated disease as covariates to analyse their association with HZ and PHN. Rate ratios (RRs) and 95% confidence interval (CI) from the Poisson regression model after controlling for each of the other covariates in the model were used to estimate the magnitude of the association between risk factors and HZ/PHN. A p-value less than 0.05 was considered as statistically significant.

RESULTS

Demographic data of patients with zoster

In this study cohort, there were 34,280 patients (16,332 males and 17,948 females) with a claim of primary diagnosis of HZ during the 7-year study period. Among these patients with HZ, 18,569 cases (56%) were diagnosed by dermatologists, followed by family doctors (9%) and internists (7%). A total of 11,421 patients (33.3%) were 60 years or older. The incidence rate of HZ in all age groups was 4.89 cases/1000 person-years (Fig. 1). There was no significant gender difference in the incidence of HZ in Taiwan (Females: 5.09 cases/1000 person-years; Males: 4.71 cases/1000 person-years). When the results were stratified according to age, it was found that the incidence increased with age. The highest incidence rate was in patients over 80 years of age (13.69 cases/1000 person-years), while the lowest (2.07 cases/1000 person-years) in patients younger than 20 years. Overall, 1125 patients were admitted with HZ discharge diagnosis. The hospitalization rate for HZ was 16.1 per 100,000 person-years. The largest proportion (59.5%) of hospitalizations was in adults older than 60 years of age. The average number of inpatient days per zoster admission was 8.3 days, and this increased with age from 6.06 days in 0–20 year olds to 9.19 days in elderly people (over 60 years of age).

Fig. 1. Age- and sex-specific incidence rate (and 95% confidence intervals) of herpes zoster during 2000 to 2006 in Taiwan.

Co-morbid diseases

The co-morbid diseases in patients with HZ included DM (7062 cases, 20.60%), lymphoma/leukaemia (233 cases, 0.68%), breast cancer (314 cases, 0.92%), liver cancer (388 cases, 1.13%), SLE (284 cases, 0.83%), and HIV/AIDS (61 cases, 0.18%). A multivariate Poisson regression model demonstrated that patients with HZ were more likely to have DM, lymphoma/leukaemia, breast cancer, liver cancer, SLE, and HIV/AIDS even after controlling for confounders, including age, sex, and other potential risk factors (Table I).

Table I. Multivariate analysis for the co-morbid diseases in patients with herpes zoster vs. controls

Co-morbid diseases	RR (95% CI)a	p-value
Diabetes mellitus	1.522 (1.478–1.565)	< 0.001
Lymphoma/leukaemia	1.908 (1.670–2.179)	< 0.001
Breast cancer	1.568 (1.399–1.758)	< 0.001
Liver cancer	1.191 (1.076–1.318)	< 0.001
Systemic lupus erythematosus	2.115 (1.876–2.385)	< 0.001
HIV/AIDS	1.527 (1.172–1.990)	< 0.001

aModel adjusted for age and sex.

RR: rate ratio; CI: confidence interval; HIV: human immunodeficiency virus; AIDS: acquired immunodeficiency syndrome.

Post-herpetic neuralgia

A total of 4543 patients (13.3%) had persistent neuralgia one month after the start of the zoster rash (incidence 0.64/1000 person-years) and 2944 patients (8.6%) developed PHN 3 months after the start of the zoster rash (incidence 0.42/1000 person-years) (Fig. 2). Zoster patients 60 years or older were more likely to develop PHN than those younger than 60 years (RR: 2.34) (Table II). Moreover, more than 20% of zoster patients over 80 years old would develop PHN. In contrast, the frequency of developing PHN was similar in males and females (8.82% vs. 8.37%). Other independent risk factors for PHN included diabetes (RR: 1.35), lymphoma/leukaemia, and SLE.

Fig. 2. Percentage of post-herpetic neuralgia (PHN) in patients with herpes zoster.

Table II. Multivariate analysis for potential risk indicators for post-herpetic neuralgia at 3 months in patients with herpes zoster

Potential risk indicator	RR (95% CI)a	p-value
Age ≥ 60 years	2.344 (2.171–2.532)	< 0.001
Female gender	0.953 (0.886–1.025)	0.195
Diabetes mellitus	1.351 (1.246–1.467)	< 0.001
Lymphoma/leukaemia	1.735 (1.319–2.282)	< 0.001
Breast cancer	0.748 (0.526–1.063)	0.105
Liver cancer	0.864 (0.651–1.148)	0.315
Systemic lupus erythematosus	2.268 (1.749–2.942)	< 0.001
HIV/AIDS	0.475 (0.264–0.856)	0.013

aModel adjusted for age and sex.

RR: rate ratio; CI: confidence interval; HIV: human immunodeficiency virus; AIDS: acquired immunodeficiency syndrome.

Management of the disease

Table III shows the distribution of prescribed drugs and other treatment modalities, including acupuncture and herbal medicine. Non-steroidal anti-inflammatory drugs (NSAIDs) were the most frequently prescribed medications, and were used by 61.1% of patients. The other common treatments included acetaminophen (49.3%), topical and systemic antiviral agents (23.6%), systemic corticosteroids (14.7%), tricyclic antidepressants (8.9%), anticonvulsants (6.6%), and opiates (3.9%). A minority (3.6%) of patients also took Chinese herbal medicine and 0.3% received acupuncture.

Table III. Treatment modalities in Taiwanese patients with herpes zoster

Treatment modalities	% of zoster patients
NSAIDs	61.1
Acetaminophen	49.3
Antiviral agents	23.6
Topical antiviral agent	13.4
Systemic antiviral agent	12
Systemic corticosteroid	14.7
TCA	8.9
Anticonvulsants	6.6
Carbamazepine	5.2
Gabapentin	1.9
Opiates	3.9
Chinese herbal medicine	3.6
Topical NSAIDs	2.1
Acupuncture	0.3
Lidocaine cream	0.1

NSAIDs: non-steroidal anti-inflammatory drugs; TCA: tricyclic antidepressants.

Medical care expenditure

The total cost of the 34,003 home care HZ cases was €1,811,603.80 and the cost per case was €53.30 ($NT1655). On the other hand, the total cost of the 1125 hospitalized cases was €1,377,791 and the cost per treated hospitalized case was €1224.70 ($NT38,051). With a mean inpatient stay of 8.3 days, the average expenditure per treated case was €147.60 per day. Further analysis by age groups showed that patients 60 years or older accounted for 59.5% of the total medical care costs.

DISCUSSION

Unlike other population-based studies (12), the incidence of HZ in Taiwan was estimated not only on the basis of general practitioner, but patients visiting medical centres and dermatologists, have been included in the study. According to our results, the incidence rate of HZ in Taiwan was 4.89 cases/1000 person-years in all age groups and 13.69 cases/1000 person-years in persons over 80 years of age. This incidence is comparable to the results published in Caucasian-based studies (1.2–4.8/1000 person-years) (13–15). To the best of our knowledge, there is no large population-based study of epidemiology of HZ in Asia except for one study in South Korea (16). That study was performed in military personnel and revealed that the annual incidence rate of HZ was 1.41 per 1000 population. However, this study was probably of limited value because only young adult males were enrolled.

Several studies have demonstrated that people with suppressed cell-mediated immunity from immunosuppressive diseases are at higher risk of zoster. In the present study, although most cases showed no risk factors for HZ, malignancy, SLE, and HIV/AIDS were noted in 2.73%, 0.83%, and 0.18% of our cases, respectively. Moreover, DM was found in 20.60% of our cases and multivariate analysis confirmed DM as an independent risk factor for HZ. In fact, a recent study from Israel also suggested that DM was often accompanied by impaired cell-mediated immunity and carried increased risk of HZ (17). Although a previous study reported that female gender was an independent risk factor for HZ (18), there was no significant gender difference in the incidence of HZ in Taiwan.

The incidence of PHN in our study was 0.42 cases/1000 person-years. Studies worldwide showed that the incidence, severity, and complications of HZ all increased with age (6, 7, 19–24). Possible risk indicators for the occurrence of PHN include older age, female sex, presence of a prodrome, greater rash severity, and greater acute pain (6). Our results also showed that zoster patients 60 years or older were more likely to develop PHN than persons younger than 60 years and DM was an independent risk factor for PHN. In contrast, the frequency of developing PHN was similar in males and females. As the data of rash severity and pain intensity were not available in the NHI database, we were unable to clarify their relationship with PHN.

Our study showed that the majority of patients received NSAIDs or acetaminophen. Other common therapy included antiviral agents (23.6%), systemic corticosteroids, and tricyclic antidepressants. Because the expense of using antiviral agents was covered by the NHI programme only in severe HZ infections, many patients might pay for the treatment privately. In Taiwan, Chinese herbs and acupuncture were prescribed in 3.6% and 0.3% of cases, respectively. However, further studies are needed to prove the efficacy of Chinese medicines in the treatment of HZ.

The management of HZ and its complications causes a large economic burden. Compared with the cost of HZ in Europe, the expenditure of HZ is lower in Taiwan. The cost of each home care case was €136.10 in Italy, but €53.30 in Taiwan. The cost of each hospitalized case was €4082.60 in Italy and €1224.70 in Taiwan (25). In Taiwan, a varicella vaccination programme was implemented in 2004. It was suggested that as varicella vaccine coverage in children increased, the incidence of varicella would decrease and the occurrence of HZ might increase (26). However, studies monitoring HZ incidence in the USA have shown inconsistent findings (27). Further study is required to investigate whether the zoster incidence will change after long-term varicella vaccination in Taiwan. A recent study also showed that zoster vaccine was efficacious in reducing the morbidity related to HZ in the immunocompetent elderly population (28). Another study revealed that the cost-effectiveness of zoster vaccine varied substantially with patient age and often exceeded $100,000 per quality-adjusted life year saved (29). As the zoster vaccine is currently not available in Taiwan, our study has provided the background data of HZ and further studies focused on the cost-effectiveness of HZ vaccine in Taiwan are warranted.

There are some limitations of this study. As patients with mild zoster may not have visited a doctor, it is probable that the incidence of mild zoster may be underestimated. It is also likely that people with co-morbid conditions are more likely to present to medical care, which may therefore cause bias.

In conclusion, this is the first large-scale epidemiological study of HZ in Asia and it provides information on the basic epidemiological features and impact of HZ in Taiwan.

ACKNOWLEDGEMENTS

This study is based in part on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health and managed by National Health Research Institutes in Taiwan. The interpretation and conclusions contained herein do not represent those of Bureau of National Health Insurance, Department of Health or National Health Research Institutes.

This study was supported by a grant from Taipei Veterans General Hospital (VGH V98C1-066), Executive Yuan, Taiwan, ROC.

REFERENCES