The Burden of Tick-Borne Encephalitis in Disability-Adjusted Life Years (DALYs) for Slovenia

Background Tick-borne encephalitis (TBE) presents an increasing burden in many parts of Europe, Asian Russia, Siberia, Asian former USSR and Far East. Incidence can be considered as one way to express the burden. A more comprehensive measure concerns disability-adjusted life years (DALYs), better characterizing the full burden of TBE. TBE burden in DALYs has not yet been estimated, nor has it been specified by the Global Burden of Disease (GBD) studies. Objective The purpose of the present study is to estimate the burden of TBE in Slovenia, expressed in DALYs, both from the population and individual perspectives. We discuss the impact of TBE burden on public health and potential strategies to reduce this burden in Slovenia. Methods The burden of TBE is estimated by using the updated DALYs' methodology first introduced in the GBD project. The DALYs᾽ calculations are based on the health outcomes of the natural course of the disease being modelled. Corrections for under-reporting and under-ascertainment are applied. The impact of uncertainty in parameters in the model was assessed using sensitivity analyses. Results From the population perspective, total DALYs amount to 3,450 (167.8 per 100,000 population), while from the individual perspective they amount to 3.1 per case in 2011. Notably, the consequences of TBE present a larger burden than TBE itself. Conclusions TBE presents a relatively high burden expressed in DALYs compared with estimates for other infectious diseases from the GBD 2010 study for Slovenia. Raising awareness and increasing vaccination coverage are needed to reduce TBE and its consequences.


Introduction
The burden of TBE is mostly expressed by incidence only [9,30,33]. The burden of the disease can be better evaluated if information on the incidence/prevalence, mortality and sequelae are all combined into a composite measure which can be found in the disability-adjusted life years (DALYs) [34]. DALYs reflect lost years of healthy life, as defined by the World Health Organization (WHO) [35]. DALYs, as one criteria, can be used for effective planning and prioritizing of scarce and therefore limited public health resources [34]. The burden of TBE, expressed in DALYs, has not yet been assessed nor has it been specified in the Global Burden of Disease (GBD) studies [36].
The purpose of the present study is to estimate the burden of TBE in Slovenia, expressed in DALYs from both the population and individual perspectives. Furthermore, we discuss strategies to reduce this public health burden in Slovenia. Data used reflect 2011 as the most recent year with data availability. Additionally, we consider the trend in DALYs to analyse how reflective 2011 is for other recent years. Our study can help to guide health policy and action locally as well as provide suggestions globally to mitigate the burden of TBE.

Brief overview of the DALYsʼ methodology
The DALYsʼ methodology was first introduced by Murray and co-workers in the Global Burden of Disease (GBD) project [37] using the following equation: where YLLs are the number of life years lost due to premature death and YLDs are the number of life years lost due to disability, weighted with a factor between 0 (perfect health) and 1 (death) reflecting the severity of the disability. According to the GBD-terminology and the International Classification of Functioning, Disability and Health, [38] the term disability refers to any short-term or long-term health loss in terms of functional capacity such as mobility, self-care, participation in normal activities, pain and discomfort, anxiety and depression and cognition.
In the original GBD study in 1990 and further WHO updates [39][40][41][42][43][44][45][46][47], 3% and 0% discount rates and both presence and absence of age weights in the DALYs' calculations were used. Discounting means that the value of health is weighted less in the future than in the present due to time preference related to growth in life expectancy [48]. Age weights give less weight to years of healthy life lost at young ages and older ages, with productivity as the main rationale [48]. For the GBD 1990 study, the reference standard life table has a life expectancy at birth of 82.5 years for females and 80.0 years for males [48].
Since the GBD 1990 study was published, there have been intense debates on the key methodological choices used for the DALYs' calculation, especially regarding the discounting, age weights, disability weights, the years lost per death and the incidence estimates for the YLDs calculation [48][49][50][51][52][53][54]. As one outcome of these debates, the new GBD 2010 study, done by Institute of Health Metrics and Evaluation, used a simplified calculation of DALYs without discounting and age weights, and the YLDs are calculated from prevalence estimates rather than incidence estimates [48,[55][56][57][58]. The GBD 2010 study also involves some modifications of disability weights for the DALYsʼ calculus and uses a new reference standard life table for the YLLs calculation. The new GBD 2010 reference standard life table has a life expectancy at birth of 86 years for both males and females and expresses an aspiration for high and healthy life expectancy [55].
The GBD studies estimate the global burden of diseases, including communicable, maternal, neonatal, nutritional, non-communicable diseases and injuries [48]. In 2008, the European Centre for Disease Prevention and Control urged to develop a new methodology to estimate the burden for communicable diseases in European Member States and EEA/EFTA countries by using a pathogen-based approach, taking into account subsequent sequelae and complications associated with infection [59,60]. No age weights or discounting are considered [59]. Adjustment to correct for under-ascertainment and under-reporting is applied [59,61]. Under-ascertainment refers to cases who do not seek medical care due to mild or absence of symptoms, or who have knowledge that the disease is self-limiting [61,62]. Due to under-ascertainment, cases will not enter the notification or surveillance system [61,62]. Under-reporting refers to cases who seek medical care but are not captured by the surveillance or notification systems because the infection/pathogen is not diagnosed or is misdiagnosed, misclassified or miscounted [61,62]. Further development of this methodology is in progress [60]. Using this methodology to estimate the full burdens of different communicable diseases enables comparisons among these diseases within and between countries to help guide public health policy and action in Europe [59]. In the present study, the updated methodology [60,61,62] will be applied for one individual country.
The DALYs methodology to calculate the burden of TBE On the basis of the natural course of the disease, the following health outcomes of a model were included for the calculation of DALYs [32]: death due to TBE, the acute disease of the second stage with signs of meningitis, meningoencephalitis and/or meningoencephalomyelitis and mild, moderate and severe neurological sequelae. Mild neurological sequelae (emotional liability, tiredness and intermittent headache) can be assumed not to have a significant impact on patientsʼ daily activities, social and working capacity. Moderate neurological sequelae (ataxia of gait, paresis of extremities, cognitive disorders, pronounced dementia or severe deafness) will however affect patients' daily activities, social and working capacity. In patients with severe neurological sequelae, social life and working capacity can be seriously affected and in a few cases patients need institutional care. In the present study, these neurological sequelae are considered as permanent neurological sequelae.
Data in Table 1 were used to calculate the burden for TBE in a straightforward disease burden model programmed in Microsoft Excel. The reported age-dependent number of cases with acute disease of the second stage (n ra ) was obtained from the National Institute of Public Health (NIJZ) data [63] and was used as the initial model input. These numbers of cases were corrected for under-estimation by a factor γ. This factor γ is the product of a factor β a to correct for under-ascertainment and a factor β r to correct for under-reporting. Both correction factors β a and β r are calculated based on the schematic presentation of disease progression presented in Fig 1. [32]. In the present study the factor β a was calculated on the basis of targeted specific information. Firstly, the natural course of TBE takes into account that around 90% of symptomatic infections develops acute disease of the first stage with symptoms similar to flu and about one third develops acute disease of the second stage with CNS involvement and the remaining 10% of infections involves more serious disease of the CNS [33,64]. Secondly, all reported cases were assumed to concern patients with acute disease of the second stage with CNS involvement in Slovenia [63].
Specifically, factor β r was calculated taking into account that 55% of TBE cases are reported; i.e., 45% of TBE cases are assumed misdiagnosed, misclassified or miscounted. The age-dependent estimated numbers of cases with acute disease of the second stage (n ea ) were calculated as follows: e Disability weights for acute disease of the second stage are not available, therefore the disability weights of Japanese encephalitis from GBD 2004 were updated and subsequently taken into account [68]. f The maximum and minimum values of probabilities for mild, moderate and severe permanent neurological sequelae for those aged 15 years and over were taken from the highest and lowest values from both studies in Slovenia [65] and Lithuania [4].
g The maximum and the minimum values of probabilities for mild, moderate and severe permanent neurological sequelae for under the age of 15 years were taken from both studies in Slovenia [72,73].
h Disability weights for mild, moderate and severe permanent neurological sequelae due to TBE are not available, therefore the disability weights for each of these sequelae as well as their maximum and minimum values were calculated as 1 minus available quality weights. The quality weights for mild, moderate and severe permanent neurological sequelae and their maximum and the minimum values were derived from published data [69].
j The maximum and minimum values for factor βd is calculated as ± 2 times the base case value. k The probability of death due to TBE was calculated from the average reported deaths due to TBE during the period 2002 to 2011 (1) [63] divided by all reported TBE cases from 2011 (247) [63].
doi:10.1371/journal.pone.0144988.t001 The estimated number of cases for mild (n emild ), moderate (n emod ), and severe (n esev ) permanent neurological sequelae and deaths due to TBE (n ed ) were calculated, using the annual transition probabilities p mild , p mod , p sev and p d of moving from acute disease of the second stage to health outcomes (mild, moderate, severe permanent neurological sequelae and deaths due to TBE, respectively). The age-dependent annual transition probabilities for mild, moderate, and severe permanent neurological sequelae were obtained from national studies [29,65]. The age-dependent annual probabilities of deaths due to TBE (p d ) were calculated from the NIJZ data [63] and were corrected for under-reporting due to under-diagnosis by a factor β d . The factor β d for TBE was calculated based on a published study [66] So, the age-dependent estimated numbers of cases for mild, moderate, severe permanent neurological sequelae and deaths due to TBE (n ed ) were calculated as follows: These estimated numbers of cases for the various health outcomes were subsequently incorporated in the DALYs calculation for TBE. YLLs were calculated by multiplying the estimated age specific numbers of deaths due to TBE (n ed ) with the remaining life expectancy (e) at that age. For life expectancy, the latest GBD 2010 reference standard life table [58] was used.
The YLDs acute for the acute disease of the second stage were calculated by multiplying the estimated number of cases with acute disease of the second stage (n ea ) with the appropriate disability weights (Dw a ) and duration of hospitalization (l a ).
The YLDs seq for mild, moderate and severe permanent neurological sequelae were calculated by multiplying the estimated number of cases for mild (n emild ), moderate (n emod ) and, severe (n esev ) permanent neurological sequelae with its disability weights (Dw mild , Dw mod and Dw sev ) and remaining life expectancies (e mild , e mod and e sev ).
The YLDs for all health outcomes (acute disease of the second stage, mild, moderate and severe permanent neurological sequelae) were aggregated to get the total YLDs.
The full burden for TBE is the sum of YLLs and YLDs as follows: DALYs ¼ n ed Â e þ n ea Â Dw a Â l a þ n emild Â Dw mild Â e mild þ n emod Â Dw mod Â e mod þ n esev Â Dw sev Â e sev ð11Þ The burden of the disease for one year is presented from the population (total DALYs and DALYs per 100,000 populations) and the individual perspectives (DALYs per case) [62].
Neither discounting nor age weighting was applied in the DALYs calculus.  [58]. An incidence-based approach was used to estimate YLDs. In the absence of discounting and age weights, this approach converges to incidence multiplied by estimated duration times with disability weighting for sequelae [67]. In a prevalence-based approach without discounting and age weights, YLDs would follow from prevalence of sequelae times disability weights [67]. As prevalence is approximately incidence times duration, our incidence-based YLDs across all ages would approximately equal the prevalence-based YLDs, in particular in the absence of discount and age weighting [67]. Notably, when discounting and age weights are applied, the prevalencebased YLDs for all ages may be quite different from the incidence-based YLDs [67].
As no disability weights for the acute disease of the second stage of TBE exist, age-dependent disability weights were derived from the GBD 2004 updated study for Japanese encephalitis [68]. Also, no disability weights for the various levels of TBE-related neurological sequelae severities exist, and such disability weights for the mild and moderate and severe permanent neurological sequelae were derived as 1 minus the respective quality weights for each of these sequelae. The quality weights for mild, moderate and severe permanent neurological sequelae were derived from a published study [69]. As mentioned, all base-case parameter values are listed in Table 1.
In sensitivity analyses, we tested how the results change with changes in the input parameters' values. One-way and probabilistic sensitivity analyses (PSA) were conducted. With oneway sensitivity analyses one parameter was changed while other parameters were kept constant at the base case values. The maximum and minimum values of each parameter were defined by the range from Table 1. The results of one-way sensitivity analyses were represented as a tornado diagram for the population perspective. PSA was conducted in the form of simulations using @Risk 1 (Palisade Corporation) and 1000 repetitions. Uniform distributions were used between maximum and minimum values as defined by the ranges from Table 1. Results of the PSA were presented using 95%-uncertainty cut-offs. Table 2 shows that from the population perspective total DALYs amount to 3,450 or 167.8 per 100,000 population, while from the individual perspective they amount to 3.1 per case. The disease burden is dominated by permanent neurological sequelae (93.9%), followed by the burden of premature death due to TBE (5.6%), while the burden due to the acute disease of the second stage reflects the smallest proportion (0.5%). Within the total permanent neurological sequelae, the burden of moderate sequelae presents the biggest proportion (77.6%) followed by the burden of severe sequelae (14.0%) and the burden of mild sequelae (2.4%). Fig 2 shows that, from the population perspective, the burden of TBE expressed in DALYs in the age group from 15 to 54 years amounts to 68% of the total burden. The burden in the working age population from 15 to 64 years amounts to 85%. Within working age populations, the burden in the prime-aged workers group from 25 to 54 years of age amounts to 53%, followed by the burden in the youth workers group from 15-24 years of age and older workers group from 55 to 64 years (16-17%).  The burden expressed as the number of reported cases in the age group from 15 to 54 years amounts to 51% of the total burden and to 72% in the working age population aged 15 to 64 years. The burden in the prime-aged workers group amounts to 43%, followed by the burden in the older workers group (21%), while the burden in youth workers group amounts to 8%.

Sensitivity analyses
With one-way sensitivity analyses, it was tested which parameters have the greatest influence on the modelʼs results. From Fig 4, the ordering of parameters according to their influence on DALYs can be seen. The disability weights of moderate permanent neurological sequelae have the greatest influence on estimated DALYs.
Results of the probabilistic sensitivity analyses presented in Fig 5 show a 95% uncertainty interval from 2,394 to 5,774 DALYs. It shows that, for example, with 0.95 likelihood total DALYs are higher than 2,394, whereas DALYs are higher than 2,631 with 0.9 likelihood.

Discussion
DALYs as a composite measure enable the combination of information on incidence, mortality and sequelae associated with TBE infection. Therefore, DALYs provide unique insight into the The Burden of Tick-Borne Encephalitis in DALYs disease burden [34]. From the population perspective, total DALYs amount to 3,450 (167.8 per 100,000 population) in 2011 in Slovenia, while from the individual perspective they amount to 3.1 per case for that same year. The majority of the estimated burden is centered in adults with potential major impact on productivity losses. 1.5% of TBE burden in the age group from 5-14 years seems low compared to the burden in adults. However, as these are children and disease in children is often considered as of even higher subjective burden than in adults, a greater relative burden of disease may well be conceived than reflected just by the bare percentages. Also, the present study emphasizes the impact of neurological sequelae. These sequelae have significant impact on patients᾽ quality of life, social and working activities. Cognitive impairments are perhaps one of the most relevant sequelae of TBE and potentially are crucial in the work environment.
In the present study, TBE country-specific data were used for the DALYs calculation. The study estimates are based on the best evidence the authors could find, however, some data and information were not available for Slovenia, such as disability weights for TBE, both correction factors β d (under-reporting of deaths due to TBE) and β r (under-reporting of cases of the acute disease of the second stage). In Slovenia, only cases with the acute disease of the second stage with CNS [75] and death due to TBE [63] are routinely reported, DALYs have previously not been used to measure the burden of TBE and application of correction factors for TBE as done in the present study reflects an innovative approach in DALYs' calculations. The disease burden model developed in the present study could be used as a representative model for TBE DALYs' calculations for other countries. Cases with asymptomatic infection and symptomatic cases including acute disease of the first stage are not captured by notification or surveillance systems in Slovenia [75] nor by our correction factors, causing potential under-estimation of our DALYs' estimates.
Disability weights of mild and severe permanent neurological sequelae, as addressed in the one-way sensitivity analyses, appeared to have no crucial impact on DALYs. Additionally, The Burden of Tick-Borne Encephalitis in DALYs correction for under-reporting of deaths due to TBE shows only a slight impact on YLLs in the full disease burden. It is estimated that at the European level (Albania, Austria, Croatia, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Italy, Latvia, Lithuania, Poland, Russian Federation, Slovak Republic, Slovenia, Sweden, Switzerland) only 30-40% of TBE cases are reported [76], meaning that 60-70% of TBE cases are misdiagnosed, misclassified and miscounted. These figures are derived from varying systems of TBE surveillance and notification and varying diagnostic procedures among these countries in the investigated time period up to 2000 [7]. As TBE has been mandatory notifiable since 1977 in Slovenia [7] and awareness about TBE might increased in the last years, we decided to use a conservative estimate of 55% of TBE cases being reported in Slovenia. Taking into account the proportion of 30-40% of TBE reported cases [76], our total DALYs estimation would obviously still reflect an under-estimate.
Trend analysis of the TBE burden may be interpreted as showing a slight decrease in DALYs for the time period between 2004 and 2011, likely mainly due to increasing awareness of the disease among the general population. Despite increasing awareness during the last years, vaccination coverage of 13% in terms of persons receiving one or two doses of TBE vaccine was low in 2009 [77]. Notably, vaccination seems far too low in Slovenia for efficient prevention and control of TBE [31]. Still, the trend analysis shows that the estimated burden of 3,450 DALYs for 2011 is likely representative for the national disease burden.
In a report of the Dutch National Institute of Public Health & the Environment, burdens are presented as measured in DALYs, incidence and mortality for seven selected infectious diseases (influenza, measles, HIV, campylobacteriosis, infection with enterohaemorrhagic Escherichia The Burden of Tick-Borne Encephalitis in DALYs coli, salmonellosis and tuberculosis) and compared between some European countries [34]. It was concluded that the relative burden of these diseases expressed in DALYs is different compared to the relative burden expressed just by incidence or mortality data [34]. In a related study [78] explicitly recommend to calculate DALYs also for other infectious diseases in Europe to prioritize interventions.
DALYs have previously not been used to measure the burden of TBE. TBE burden, expressed in DALYs is not included in GBD studies, and results of the present study give novel information in that respect. Comparison of the TBE burden expressed in DALYs in the present study to DALYs for other infectious diseases from the GBD 2010 study for Slovenia [79] is possible. However, this comparison should be considered with caution as no GBD disability weights for TBE exist. In the GBD 2010 study for Slovenia [79], the cluster for communicable, maternal, neonatal and nutritional disorders presents 25,501 DALYs per year. Within this group [79], lower respiratory infections present the highest burden with 7,545 DALYs amounting to 29.6%, followed by diarrheal diseases with 1,225 DALYs (4.8%), otitis media with 954 DALYs (3.7%), upper respiratory infections with 951 DALYs (3.7%), tuberculosis with 776 DALYs (3%) and HIV/AIDS with 547 DALYs (2.1%). The potential proportion of our estimated TBE burden in this cluster amounts to 13.5%, demonstrating a relatively high burden in Slovenia. Thus, the present study can serve as an informative estimation of the TBE᾿s national burden and the importance of the TBE on population's health in Slovenia. TBE can be considered to have a high impact on public health and to present a challenge for more efficient health policies and actions to reduce TBE in Slovenia. Such action may lead to huge population health benefits on national scales.
Austria is the only European country where an extended vaccination campaign was launched in 1981. In Austria, vaccination is free of charge only for people with an occupational risk of TBE, while for the rest of the Austrian population a part of the costs are still covered by health insurance [9,80]. Vaccination coverage of persons receiving at least one dose of vaccine increased from 6% at the beginning of the programme to 88% in 2006, with 58% being regularly vaccinated within the recommended schedule [27]. This has led to drastic reductions in TBE in all age groups [27]. In the years between 2000 and 2006, about 2800 cases were prevented in Austria [27].
TBE is endemic in Slovenia, one of the countries with the highest incidence worldwide [30]. In Slovenia, vaccination is recommended for the general population from the age of one. TBE vaccine is free of charge only for those who are potentially exposed during fieldwork while the rest of the population has to pay for the vaccination [81,82]. Furthermore, because of limited public health resources in Slovenia, it was recommended [30] to provide only selective access to free of charge vaccination for specific age groups in specific regions. The findings of the present study may suggest the need to prioritize prevention of TBE disease further and reallocation of limited public health resources in Slovenia, for example, for an extended TBE vaccination for all groups, ages and regions, potentially free of charge or at a reduced price. It has been shown before that extended TBE vaccination can be beneficial from an economic perspective [80]. Extended vaccination may result in major health benefits to the population as a whole as well as cost savings to the health-care system.

Conclusions
TBE presents a relatively high burden in Slovenia, expressed in DALYs both from the population and individual perspectives. Public-health impact may justify reallocation of scarce budgets to better control TBE. In particular, continued awareness raising and corresponding increased vaccination coverage is needed to reduce TBE and its consequences in Slovenia.