Total trauma follow-up costs
The objective of the present study was to estimate trauma follow-up costs for Germany. A margin of total annual trauma follow-up costs was calculated in the amount of EUR 11.1 billion for the lower bound and EUR 29.8 billion for the upper bound, respectively. The correspondence of the Australian result with Germany’s lower bound should be considered with utmost caution, since both cost studies are based on completely different methods and also include different cost areas. In contrast to the Australian study , the German cost calculation  does not take crime and deadweight losses into consideration, whereas the areas health, education, productivity and social services have been considered equally.
In both studies, prevalence is based on a population survey with similar results for the lifetime prevalence of physical and sexual abuse (17.8% in Australia  and approximately 15.9% in Germany [10, own calculation]). However, the Australian study uses only the one-year prevalence of 0 to 17- year-olds and does not include emotional abuse or neglect. Thus, the number of people concerned is much lower in the Australian study, whereas the total costs per person must lie close to those of Germany’s upper bound: the upper bound costs (EUR 18,087) multiplied by Australia’s prevalence (3.7% of Germany’s population aged 0–15 years: 412.147, age range as presented by the German Federal Statistical Office ) would yield – with EUR 7.5 billion – a result quite close to the Australian one.
The Canadian result  is very close to the German upper bound, but relies on higher prevalence rates (cf. Table 3), which have been used for cost calculation in a sophisticated way. Canadian costs comprise expenses related to the legal system, social services, education, health, employment and personal costs, with the employment sector being the most expensive one, accounting for 72% of the total costs (CAD 11.3 billion of a total of CAD 15.7 billion). Other than in the Australian and US-American studies, Canada has based their cost calculation in the employment sector on a large population survey, which combined information on income with physical and sexual abuse in the respondents' history (Ontario Health Survey Mental Health Supplement (OHSUP)).
With a loss of productivity of over 70% in the moderate case scenarios and over 50% in the pessimistic ones, the German cost-benefit-analysis  ranks close to the Canadian result. Since costs are oriented towards individual life courses in both countries – in Canada on the basis of a population survey and in Germany on the basis of individual case scenarios – this result could in fact point in the right direction, namely to regarding productivity losses as the main cost driver of societal trauma follow-up costs. The other two studies [52, 53] rely on less specific, aggregated data. In Australia , losses of productivity rank far behind the other areas, while an approximate proportion of almost 30% can be derivated for the United States .
The total costs in the United States  are considerably lower than those in other countries, even though the cost sectors taken into consideration largely correspond. However, on the one hand, child neglect is not included for methodological reasons, and on the other hand, the number of child abuse victims is not determined on the basis of a population survey, but official information sources are used . Despite the attempt to calculate institutionally unknown cases, the dark figure remains largely unconsidered. There is naturally no precise information regarding the magnitude of this dark figure. Wetzels  indicates an optimistic estimate at the ratio of one to ten. This estimate projected on the US-American study would yield a result of EUR 18 billion instead of EUR 1.8 billion and thus above the Australian costs and within the German cost frame.
The two results from Australia  and the United States , which contain intangible costs, cannot be compared with the German result (without intangible costs). In the Australian study, intangible costs make up for 1.7 times, in the US-American study even 6.7 times of the tangible costs. In spite of this large difference it can be stated on the transnational level that intangible costs as a measure for personally experienced burden considerably exceed the actual expenses in the form of tangible costs in any case.
Generally, the comparison of the four aforementioned results of trauma follow-up costs is limited due to the time lag of altogether fifteen years between the individual studies, which have certainly influenced prices, services and their utilization. Additionally, national service organization and funding structures, e.g. of the health care systems, are fundamentally different . These variations presumably influence the availability and the assessment of costs and their assignment to various sectors, so that differences in costs are to be expected a priori due to structural conditions.
The calculated amount of trauma follow-up costs clearly has economic relevance, constituting 0.44% (lower bound) and 1.20% (upper bound) of Germany’s 2008 Gross Domestic Product (EUR 2,489.4 billion) . The figures have an additional relevance due to the fact that with early and effective intervention or prevention, they reveal a considerable saving potential [9, 51, 56–58].
Basically, trauma follow-up costs were determined by following a conservative approach. This is reflected in several details, for example in the restriction to a 21% share of only severely affected cases [4, 10]. Results of risk- and resilience research lie around this value for the share of traumatized individuals with long-term consequences caused by trauma-related disorders [50, 68].
Moreover, total costs have only been taken into account for the age group from 15 up to including 64 years. Consequently, direct costs are only considered to a small extent and indirect costs of older ages remain completely excluded. With existing trauma-related disorders, it can be assumed that the age-related, increasing instability of life situation leads to further health problems, which again incur additional costs in higher age. In general, trauma-related disorders do not tend to decrease in higher age [1, 69], but elderly people are often severely impaired due to e.g., insufficient specialized care .
Last but not least it should be noted that types of traumatization other than sexual, physical and emotional abuse and neglect are left unconsidered in the present study, so that no statements can be made on their prevalence or on follow-up costs. However, it appears reasonable to limit the derivation of trauma follow-up costs to child abuse and neglect, since other current data are not available and international cost studies [52–58] refer to these types of traumatization exclusively or predominantly, so that results can be better compared with each other.
When trying to estimate whether the true costs may tend towards the lower or the upper bound one has to keep in mind that the cost scenarios are based on early childhood traumatization, whereas the prevalence data include the entire childhood and adolescence as defined by the CTQ. Since trauma-related disorders tend to be more severe the earlier traumatization was experienced [2, 70], this discrepancy leads in the direction of the lower bound.
The international comparison supports both the lower and the upper bound of the cost margin – depending on the respective study. Due to methodological differences the results have to be interpreted rather as crude reference points, though. Despite all limitations, the comparison shows that the cost margin calculated for Germany is well associated with other countries’ results.
Limitations associated with the use of already existing data are particularly given by the fact that these data have been collected with different objectives and are not well-matched. The question arises, in particular, to what extent the cost scenarios – determined under the assumption of child endangerment  – can be projected on the number of traumatized individuals identified in epidemiologic studies [4, 10]. While various age limits of the investigated populations can be approximated, it cannot be stated with certainty whether the cost scenarios described by Meier-Gräwe and Wagenknecht  are based on the same kind of traumatization as the determination of prevalence by Häuser et al. .
The prevalence of traumatization has been determined by Häuser et al.  retrospectively, which may represent yet another source of error – due to blurred or imprecise memories. However, several studies of this kind illustrate the fact very well that the number of errors is to be estimated rather low and of conservative type, in other words, that the results tend to underestimate reality [3, 5, 11, 20, 45, 71, 72].
Another significant uncertainty lies in the cost data themselves. The authors of the cost study explain in detail that due to the lack of data, several parts had to be completed by expert knowledge and international literature . The complete case scenarios are thus but a construct, which has been developed as close to reality as possible with the help of various information sources.
The problem of low availability and unsatisfactory quality of the data with regard to the estimation of trauma follow-up costs does not only exist in Germany but it is criticized in all cost studies [52–58]. Consequently, results are consistently presented as fragmentary and underestimating. Since it can therefore be assumed that all cost studies deviate from reality in the same direction – with the extent of deviation being unclear – a comparison is possible and reasonable. Nevertheless, it can only be valued as a comparison of cost dimensions, not of amounts calculated precisely to the cent, solely due to the different methodologies. In view of the generally weak data, it should be noted that by using more precise procedures, only the illusion of higher precision could be created. This is not the intention of the authors.