In this study, we found that about 10.3% of the general German population suffer annually from unintentional non-fatal injuries that require medical treatment. The overall incidence of non-fatal injuries was higher among men than women, and showed different age patterns with regards to gender and place of occurrence. The overall incidence of non-fatal injuries did not substantially vary with the educational levels. However, the places of occurrence (work, home, leisure time and traffic accident) did differ by educational level.
In our survey, we collected a lot of sociodemographic information on the injured individuals, such as age, sex and educational level. Upon analysing these sociodemographic data, we found that work-related injuries had the highest incidence among men. Among both men and women, the incidence of work-related injuries was highest among those with the lowest educational level. Injuries caused by traffic-related accidents most often occurred among young adults (18–29 years) and the elderly (70+ years), and varied in the type of road use (younger: car accidents; elderly: accidents as pedestrians). Altogether, the proportion of pedestrians and cyclists among the people injured in traffic accidents was very high in our study (cyclists: 21.8%, pedestrians 36.6%). Among women, the most frequent place of occurrence was at home. In summing together two important places of accidents, home and leisure time, we observed that they constituted the largest proportion of all reported accident-related injuries (49.5%).
The fact that the injured subjects required medical treatment (as based on self-reports) implies that these accidents were an underestimated burden of disease in Germany. Our survey data may supplement official statistics because we measured injuries (at home, during leisure-time and traffic accidents) that are not necessarily represented in official statistics. We also provided important information on the injury victims, such as age, sex, education level, type of body part injured and hospital admission (the latter two not shown). Such information is necessary to identify high-risk groups and plan specific strategies for injury prevention. Our analyses indicated that young men were especially at high risk of non-fatal injuries. It is also well-documented through cause of death statistics that the risk of fatal injuries in this group is also relatively high [2]. Another high-risk group is vulnerable road users, particularly pedestrians.
Comparing frequency and places of occurrence among published statistics is problematic because categorisations may differ. For example, our estimated one-year incidence of non-fatal injuries (10.3%) was in line with the annual estimates published by the Federal Institute for Occupational Safety and Health [16]. Further agreement was seen in the relative frequency of certain places of accidents. For example, the Institute’s estimation and our results demonstrated that home and leisure injuries (overall) accounted for the most frequent types of injuries (Bundesanstalt für Arbeitsschutz und Arbeitsmedizin 2004: 63.2% [16], Telephone health survey 2004: 49.5%). However, comparing injuries due to leisure-time accidents among the different reported statistics is complicated because the Institute categorises injuries at public places, e.g., when walking or cycling, as leisure-time accidents if they occur during leisure time. On the other hand, in our survey, these injuries were considered to be caused by traffic accidents. In addition, the Federal Institute for Occupational Safety and Health also counted injuries that did not require medical treatment. It is difficult to compare our incidence estimates of traffic accident-related injuries to official road traffic accident statistics by the German Federal Statistical Office because the methodologies in reporting differ. The same is true when trying to compare our incidence estimates of work-related injuries to the workplace accident statistics reported by the statutory accident insurance carriers. For example, our analyses show, like the official traffic accident statistics, a decreasing impact of injuries due to car accidents with increasing age, but a rising number of injured pedestrians in the oldest age groups [17]. However, the percentage of injured pedestrians and cyclists is much higher in our data than in the official road traffic statistics. These accidents may be underreported in the official statistics because they contain only those that are registered by the police.
According to European Union-wide statistics, between 2005 and 2007, about 60 million people - nearly an eighth of EU residents - sought medical treatment for an injury annually (data from the 27 Member States of the European Union) [1]. We obtained a similar incidence, although the EU statistics also included children’s injuries as well as intentional and fatal accidents. Again, the comparison of data on injuries across countries is problematic. Although the collection of data on injuries is considered an important priority in many countries, the actual data collection currently performed in member states varies both in the methodology and in the degree of comprehensiveness of data collection [18]. EU statistics show that home and leisure injuries are the most frequent types of injuries (74.0%) [1]. Our analysis of non-fatal injuries among adults similarly showed that home and leisure together were the most frequent settings for an accident (see above). The difference in the relative frequencies of injuries is due to different inclusion criteria. According to our findings regarding injured pedestrians and cyclists, it is estimated that about half of all hospital-treated injuries are not registered in police statistics [1]. When considering the broader consequences (including economics) of injuries, the European Union is currently seeking to harmonise the methodology of collecting data on this subject across its member countries.
There are several factors that limit our results. First, by nature of the survey used, our results are restricted to non-fatal injuries, as we were only able to interview accident survivors. Moreover, severe injuries that may have resulted in brain damage were underrepresented in our study because the inclusion criteria of the survey required the respondents’ ability to undergo telephone interviews.
Second, we relied completely on self-reports, which are always susceptible to error. As we did not carry out a validation study, we could not report on the sensitivity and specificity of self-reported injuries. We know from other studies that participants remember earlier and/or minor injuries less well, as is also true for injuries in children [19, 20]. We asked adults about their own injuries that were treated medically, and so, this did not include minor injuries. Our accident incidences are comparable to those of the Federal Institute for Occupational Safety and Health. Nevertheless, it is possible that a small proportion of injuries were not specified.
Third, the response rate was low. Comparing the distribution of age in our survey with official statistics revealed an underrepresentation of the elderly, especially those over 80 years. However, if we want to calculate age-standardised rates, it is important to take them into consideration because they contribute to the result. Special studies on the frail elderly and those of advanced age will provide a more accurate picture for this group. Comparing the distribution of educational levels in our survey with census data revealed an oversampling of subjects with a higher educational level. Nonetheless, standardisation of the incidence by age- and gender-specific weights based on census data did not substantially change the incidence estimates. Since the telephone survey was conducted in the German language, the number of participants with an immigrant background was below average. There is only little information about their risk of accidents from official statistics. However, we have indications of deviations, for example, more accidents at work in foreigners than in Germans [21]. However, analysing the telephone surveys of the Robert Koch Institute, there was no clear trend as to whether the health status would have improved or not if more immigrants had been included in the survey [22].