BackgroundThe objectives of this study were to characterize the burden of herpes zoster, as well as the longitudinal and incremental changes of healthcare service utilization among individuals with herpes zoster and postherpetic neuralgia (PHN) compared to those without.MethodsUsing the National Health Insurance Research Database (NHIRD), we established a herpes zoster cohort of people diagnosed with herpes zoster between 2004 and 2008 as study cases. Another subset of the NHIRD, which was randomly selected from all elderly beneficiaries between 2004 and 2008 served as a non-herpes-zoster elderly control pool. Each case was then assigned one matched control according to age, gender, index date and propensity score. PHN cases were defined as those with persisting pain for more than 90 days after the onset of herpes zoster.ResultsBetween 2004 and 2008, about 0.6 million patients were newly diagnosed with herpes zoster. The incidence increased with age, and most cases were identified during the summer period. Herpes zoster cases were found to have higher consumption of all types of healthcare services in the first year after the index date. Such increases were particularly obvious for patients with PHN, who showed incremental increases on average of 16.3 outpatient visits, 0.4 emergency room visits and 0.24 inpatient admissions per year.ConclusionsThe incidence of herpes zoster increased with age and changed according to the seasons. Patients with herpes zoster were associated with higher healthcare utilization and this increase in healthcare utilization was most obvious for herpes zoster patients with PHN.Electronic supplementary materialThe online version of this article (10.1186/s12889-018-5247-6) contains supplementary material, which is available to authorized users.

Empirical data of medication-related hospitalization are very limited. We aimed to investigate the associations between 12 high risk medication categories (diabetic agents, diuretics, nonsteroidal anti-inflammatory drugs (NSAIDs), anticoagulants, antiplatelets, antihypertensives, antiarrhythmics, anticonvulsants, antipsychotics, antidepressants, benzodiazepine (BZD)/Z-hypnotics, and narcotics) and unplanned hospitalizations. A population-based case-time-control study was performed using Taiwan Adverse drug events (ADEs) and their associated morbidity and mortality 1 represent a significant burden on the healthcare system 2 . Systemic reviews have estimated that approximately 5-10% of total hospital admissions are related to an ADE 3,4 . For the years 2004-2005 in the US, there were more than 700,000 ADE-related emergency department visits annually, and 1 of every 6 led to subsequent hospitalizations 5 . The annual cost of ADE in the US has been estimated at more than $ 136 billion 6 . In addition, the elderly are the most vulnerable to ADE-associated unplanned hospitalization 7 . For the elderly, the percentage of hospitalization attributed to ADE is estimated at 3.4-16.6% 8 . Since a great proportion of ADE-related hospitalizations are preventable 4 , efficient identification of ADE-related hospitalizations is therefore crucial to help highlight area which clinicians and policy-makers can put efforts in. Previous studies have adopted different approaches to identify ADE-related hospitalizations, such as analysis of spontaneous reporting data 9 , medical chart review 5,7,8,[10][11][12][13][14] , and screening diagnostic codes from electronic medical databases [15][16][17] . However, these methods are plagued with under-reporting issues and reporting bias. In addition, some of the methods are personnel-costly and time-consuming, resulting in studies with small sample sizes that cannot be generalized 10,[12][13][14] . Furthermore, it is difficult to interpret the results between different medication categories in existing studies, since most studies only provide descriptive data on