Abstract Background Worldwide, about 5.8 million people die each year due to injuries. In Sweden, the corresponding number amounts to 3000. There are large differences among injury types regarding the age-profile of the fatalities and as most of them occur in older age groups, counting the absolute number of injury fatalities does not fully reflect the size of the burden of injury. Methods Using age- and sex-specific life expectancy tables in combination with data on external causes of injury, the number of injury fatalities in Sweden for the time period 1972–2014 is converted to a sum of potential years of life lost (PYLL). We then fit cause and group-specific spline regression models to the data to estimate temporal trends in both fatality counts and PYLL. Results There has been a steady reduction in the number of injury fatalities and in the sum of PYLL from the early 1970s to around the year 2000. Since then, there has been an increase in the number of injury fatalities and in the sum of PYLL. The upward trend is mainly explained by an increasing number of deaths due to poisonings and suicide, specifically among younger men. Conclusions The increases in suicide and poisoning mortality offset the reductions in downward trending causes of injury mortality during the last decades. The share of PYLL is larger than the share of fatalities for both suicides and poisonings implying that an aging population does not cause the increase. Introduction Worldwide, about 5.8 million people die each year because of injuries, accounting for 10% of all deaths.1 All age groups are affected, but injuries have a particular impact on young people and people in their working years (ibid). In the USA, unintentional injuries are the leading cause of death up to the age of 442 and in Sweden, injuries are the most common cause of death up to the age of 39, causing 79% of all deaths among men and 68% among women.3 According to data from the Global Burden of Disease study,4 Sweden is comparable with the average Western European country in terms of injury-related death rates (45.8 per 100 000 person-years in Sweden vs. 43.8 in Western Europe), potential years of life lost (PYLL) (1096 vs. 1061 per 100 000 person-years) and disability-adjusted life years (2139 vs. 2099 per 100 000 person-years) per population. The rates in Western Europe are roughly two-thirds and one-third of the rates in Central and Eastern Europe, respectively, and about half that of the global rate. Counts of the absolute number of injury fatalities in a population in a given time period are the primary means used to quantify the burden of injury mortality.5 However, as most fatalities occur in older age groups, mortality data do not fully acknowledge the burden of premature mortality, which is an important indicator of the health status in a population.6,7 There are large differences between injury types when it comes to the age profile of the fatalities. While the average Swedish fall-related fatality in 2014 caused a loss of approximately 9 life years compared with the life expectancy at the age of death, the average road traffic fatality lost 32 years of life and the poisoning fatality on average 40 years. The sum of PYLL adds a temporal dimension to mortality counts by combining information about the age at death with potential years of life left to live at that time.8 PYLL is the sum of years of life lost (YLL) due to premature fatalities from a particular cause in a given population.8 By giving weight to each year lost, PYLL values fatalities at young age more than those occurring at older age.7 From a public health perspective, reducing the number of life years lost is important since it reflects a reduction in premature fatalities and adds a dimension of social and economic consequences in quantifying the burden.8 Calculations of life years lost are part of the most well-known burden of injury presentation, which is included in the Global Burden of Disease Study (GBD) by the World Health Organization (WHO). In GBD, the burden is represented by the sum of disability-adjusted life-years (DALY’s), which is an index calculated by adding YLL and years of life lived with disability (YLD) (due to injuries).9 YLD’s are calculated by the multiplication of the number of cases with a given health outcome by disability weights reflecting the health loss sustained. In the most recent GBD study, 85% of the burden of injuries consisted of YLL. Whereas the GBD studies are not performed yearly, the present paper estimates the burden of injury mortality in Sweden using PYLL according to the cause of death register and life tables in Sweden for each year in the period of 1972–2014, allowing a more detailed trend analysis. Methods The PYLL concept involves estimating the average time a person would have lived in the absence of premature death.8 Using age- and sex specific life expectancy tables in combination with statistics on type of injury, the plain number of injury fatalities due to injuries in Sweden for the time period 1972–2014 is converted to a sum of PYLL. The definition of ‘premature’ death might be discussed and is reflected by the selected cut-off age, estimating the number of years a person would have lived had the injury not occurred.7 In some cases, a cut-off age of 65 years is chosen to stress the loss of productive life years,10 in other cases 75 or 85 years is used.11 Another measure often used is based on the life expectancy,9 at birth or at age of death, the latter being used in this case. For example, if a man died in 1987 at the age of 47, he would lose 29.64 years of life, which was the expected remaining life length for a man at that specific age collected from the 1987 life expectancy tables. A woman of the same age lost 34.77 years, due to women living longer. In 2014, the PYLL for men and women dying at the age of 47 have risen to 34.79 and 37.92 years, respectively. This also implies that deaths at any age contribute to the sum of PYLL, since there is some remaining life expectancy at any age. The sum of PYLL is calculated by the formula below, where i = age at death, d = number of deaths at age i and Li= life expectancy at age i ∑i=0∞di(Li) We then fit cause and group-specific spline regression models to the data to estimate temporal trends in both fatality counts and PYLL, which allows us to model the trends flexibly without any explicit assumptions regarding their functional form. To avoid overfitting the models to the observed data, we use leave-one-out cross-validation to determine the optimal value for the smoothing parameter so that the out-of-sample prediction error is minimized.12,13 The fitted values from these models are used in all our calculations below to minimize the impact of outlier events and random noise. Data regarding fatalities are collected from the Cause of Death register at the Swedish National Board of Health and Welfare14 and from life tables produced by Statistics Sweden.15 The quality of data in Swedish registers is high, allowing for the study of external death causes for a long period of time.16 However, there have been two revisions of the International Classification of Diseases (ICD) during 1972–2014, which is important to keep in mind when studying trends.17 ICD 8 was used until 1986, ICD 9 1987–1996 and ICD 10 from 1997 to the end of the studied period. There is a sudden decrease in the number of fall fatalities due to the change from ICD 9 to ICD 10, also acknowledged by Janssen et al.17 At the same time, there is a corresponding increase in the miscellaneous category. About two-third of the miscellaneous category consist of the category ‘other accidents’, where the circumstances around the injury fatality are unclear. The remaining one-third of fatalities includes causes of death with limited numbers of fatalities such as exposure to forces of nature, electric current and contact with venomous animals or plants. Results Overall, the number of fatalities and the sum of PYLL due to injuries show a similar trend in the period of 1972–2014, as presented in figure 1. There has been a steady reduction in the number of injury fatalities and thus also in the number of life years lost from the early 1970’s to around the year 2000. Since then, there has been an increase in the number of injury fatalities (A) and in the sum of PYLL (B). When looking at trends separated by sex, the number of injury fatalities and sum of PYLL among men mirrors the overall trend. For women, there is a slight upward turn in the number of injury fatalities since around year 2000, but the sum of PYLL does not increase. Figure 1 View largeDownload slide Number of injury fatalities and sum of PYLL in Sweden, 1972–2014 Figure 1 View largeDownload slide Number of injury fatalities and sum of PYLL in Sweden, 1972–2014 Not all types of injuries, as presented in figure 2, share the upward turn. The burden of road traffic accidents, drownings, fires and suffocation and interpersonal violence are decreasing both in terms of number of fatalities and sum of PYLL. Falls, unintentional poisonings and injuries with unknown intent are increasing since the mid-90s and at the end of the study period, suicides show an increase as well. Figure 2 View largeDownload slide Number of injury fatalities and sum of PYLL by external cause category in Sweden, 1972–2014 Figure 2 View largeDownload slide Number of injury fatalities and sum of PYLL by external cause category in Sweden, 1972–2014 For suicides, the increase in sum of PYLL is steeper than the increase in number of fatalities, indicating relatively young victims. For falls, it is the other way around, reflecting that those dying from falls are mainly elderly. In Appendix A, the trends from figure 2 are separated by sex. This shows an upturn in the number of PYLL among men due to interpersonal violence, not visible in the number of fatalities. In addition, the difference in the sum of PYLL lost by men compared with women due to fall injuries is larger than the corresponding difference in the number of fatalities. However, the most striking difference between men and women in this respect is shown in the unintentional poisoning trend, where the increase in both the number of fatalities and sum of PYLL is much steeper for men than for women. In figure 3, the relative ranks of the projected sum of PYLL by cause of death are presented stratified by decade. The largest burden in terms of PYLL is caused by suicide, which has been the case during the whole study period. There has been a sharp increase in sum of PYLL due to poisonings, moving from ranks 7 and 8 in the 70s, 80 s and 90 s to rank 2 in the 2010s, more than doubling the sum of PYLL. Falls, being the second most important cause of death in number of fatalities (miscellaneous disregarded), are less important in terms of PYLL on ranks 4 and 5 throughout the time period. The estimated average number of fatalities and PYLL per period is available in Supplementary Appendix B, tables B1 and B2. Figure 3 View largeDownload slide Changes in rank (1 = highest burden) in projected life years lost due to deaths by external causes in Sweden from 1972 to 2014, stratified by decade and cause category Figure 3 View largeDownload slide Changes in rank (1 = highest burden) in projected life years lost due to deaths by external causes in Sweden from 1972 to 2014, stratified by decade and cause category The distribution of the burden of injury fatalities for the whole time period and for the latest 5 years can be seen in Supplementary Appendix C. Suicides account for 27.59% of the fatalities in 1972–2014, but for 34.47% of life years lost. When looking specifically at the later years, in 2010–2014, the suicide share of fatalities has decreased to 24.65% but the share of PYLL has increased to 36.49%. In 1972–2014, falls account for 17.66% of the fatalities, but the share of PYLL is 7.34%. In 2010–2014, the share of fatalities has increased to 19.63%, but the share of PYLL has been steady at 7.85%. Unintentional poisonings are getting more common, increasing the share of fatalities from 4.66% in 1972–2014 to 9.12% in 2010–2014. The share of PYLL increased from 6.17% to 15.42% in 2010–2014. The upper panel of figure 4 presents the number of fatalities and the sum of PYLL due to suicides, unknown intent and unintentional poisonings compared with all other external causes. These three causes are related to each other, since only confirmed suicides are coded as suicides, meaning that a significant share of suicides are reported in the category of unknown intent.18 Similarly, it is often not possible to distinguish unintentional and intentional poisonings. Figure 4 View largeDownload slide The impact of the increases in suicide (including unknown intent) and poisoning mortality since the year 2000 on the overall trend in the sum of PYLL due to injuries in Sweden, as shown via stratification Figure 4 View largeDownload slide The impact of the increases in suicide (including unknown intent) and poisoning mortality since the year 2000 on the overall trend in the sum of PYLL due to injuries in Sweden, as shown via stratification The trend visible in the diagram to the left, where the number of fatalities due to suicides, unknown intent and poisonings is increasing, whereas the number of fatalities due to other external causes is decreasing, is strongly enhanced when looking at the sum of PYLL diagram to the right. In the lower panel, poisonings are decomposed from suicides and unknown intent. Trends show that the sum of PYLL increases more than the number of deaths for both suicides/unknown intent and poisoning. The implications are that fatalities due to suicides/unknown intent and poisonings are relatively young people. In addition, figure 4 shows that the upturn in all-injury fatalities and sum of PYLL shown in figure 1 is caused by an increase in suicides, unknown intent and poisonings. Discussion As there exists no ideal measure of the burden of injury, several measures can be used together to provide different perspectives.5 Every premature death can be seen as a social and economic loss to society, and one way to account for this is by estimating the sum of PYLL.19 Also, there are studies indicating that many people prefer saving younger lives20 and presenting the burden of injuries in terms of years of potential life lost is one way of acknowledging that preference. The sum of PYLL might also be a more relevant indicator of the indirect costs in terms of production losses of injury deaths than the pure number of fatalities due to placing more weight on death at younger age. Although similar information can be provided by presenting the number of fatalities by age groups, there is a point in summarizing it into a single, composite measure instead of dealing with tables where the choice of age groups might affect the interpretation. In addition, differences in trends between the number of fatalities and sum of PYLL, i.e. that different age groups are subject to different trends, might indicate that development is driven by multiple mechanisms9 and that there is a need to look deeper into underlying causes and preventive measures. Studying the sum of PYLL in Sweden in 1972–2014 stresses the importance of the burden of suicide and poisonings further. Not only are suicides the most common external cause of death, in terms of the sum of PYLL the share of the total burden of injury mortality is even larger and growing. Unintentional poisonings show a worrying trend, both increasing in terms of number of fatalities and, even more so, in terms of PYLL thus indicating that those affected are relatively young. The rising sum of PYLL due to poisoning is in line with a growing drug use disorder in terms of opioid dependence.21 This has also been discussed in the Swedish context, where opioid deaths have been proposed to cause 70–75% of the drug deaths.22 Furthermore, which is also the case for suicide, the fact that the share of PYLL is larger than the share of fatalities implies that an aging population does not cause the increase. To investigate this further, we study age- and sex-specific differences in the trends in Supplementary Appendix B, figures B1 and B2, where it becomes clear that men contribute more greatly to the increase in the sum of PYLL from suicides and poisonings than women. It appears that young men (aged 20–29 years) are the driving force behind the large increase in poisoning fatalities during the last two decades, even though increases are apparent in most age- and sex-specific groups. Regarding suicides, we find that the upturn in the last decade is primarily caused by suicides among men aged 20–29 years and seems to be mostly isolated to this group. The upward trends for suicide and poisonings are similar to the findings by Case and Deaton,23 where an increase in midlife mortality in the USA among specifically (white) men due to poisonings and suicide is identified. Case and Deaton, as well as Stein et al.,24 point out that the development identified in the US, where increasing mortality rates in what is called ‘despair deaths’ (suicides, poisonings and liver disease) have turned the all-cause mortality trend upwards for some subgroups, has not been present in corresponding groups in other high-income countries. However, in a later study, Case and Deaton present mortality trends in 1999–2015 for those 45–54 years for a number of countries.25 Based on that, although not offsetting the all-cause downward trend, mortality from alcohol, drugs and suicide increased in. e.g. UK, Canada, Australia and Ireland. In France, Germany and the Netherlands, on the other hand, both trends were decreasing. Although the upward trend is not strong enough to affect the all-cause mortality trend (results available upon request), this study shows that the same pattern can be traced to the Swedish setting. In comparison with the US, younger men seem to be affected. However, the increases in suicide and poisoning mortality are clearly large enough to offset the large reductions in road traffic mortality and other downward trending causes of injury-related fatalities during the last decades. This indicates an on-going transition from technical to social injury risks, where future countermeasures may need to focus more on the latter, which poses many new challenges for injury prevention since behavioral risk factors have been shown more difficult to handle.26 Further research would be needed to study whether the development of non-fatal injuries, also an important part of the burden of injury, shares the same trends as the fatalities. Supplementary data Supplementary data are available at EURPUB online. Acknowledgements We are grateful for comments and suggestions by Henrik Jaldell and Niklas Jakobsson. Conflicts of interest: None declared. Key points There are large differences among injury types regarding the age-profile of fatalities and one way of acknowledging this is by converting the number of fatalities to a sum of PYLL. The steady reduction in the number of injury fatalities and sum of PYLL from the 70 s to around the year 2000, but since the trend has turned upwards again. An increasing number of fatalities, mainly among younger men, due to poisonings and suicides offset the reductions in downward trending causes of injury mortality in Sweden. The share of PYLL is larger than the share of fatalities for both suicides and poisonings, implying that an aging population does not cause the increase in injury mortality. References 1 Injuries and Violence 2010: The Facts. World Health Organization; Geneva, 2011. Available at: http://apps.who.int/iris/bitstream/handle/10665/44288/9789241599375_eng.pdf?sequence=1 (May 2018, date last accessed). 2 National Center for Injury Prevention and Control, CDC using WISQARS™, 10 Leading Causes of Death by Age Group 2015. Available at: https://www.cdc.gov/injury/wisqars/pdf/leading_causes_of_death_by_age_group_2015-a.pdf (October 2017, date last accessed). 3 Swedish National Board on Health and Welfare, 2016. Statistik om dödsorsaker, 2015. Available at: http://www.socialstyrelsen.se/publikationer2016/2016-8-3 (May 2018, date last accessed). 4 Institute for Health Metrics and Evaluation (IHME), Seattle, United States. Global burden of disease study, 2016. Available at: http://ghdx.healthdata.org/gbd-results-tool (March 2018, date last accessed). 5 Hendrie D, Miller TR. Assessing the burden of injuries: competing measures. Int Inj Control Saf Promot 2004; 11: 193– 9. Google Scholar CrossRef Search ADS 6 CDC. Premature mortality in the United States: public health issues in the use of years of potential life lost. MMWR Suppl 1986; 35: 1S– 11S. 7 Dranger E, Remington P. U YPLL: a Summary Measure of Premature Mortality Used in Measuring the Health of Communities. Planning 1998; 113: 55– 61. 8 Gardner JW, Sanborn JS. Years of potential life lost (YPLL)-what does it measure? Epidemiology 1990; 1: 322– 9. Google Scholar CrossRef Search ADS PubMed 9 Haagsma JA, Graetz N, Bolliger I, et al. The global burden of injury: incidence, mortality, disability-adjusted life years and time trends from the Global Burden of Disease study 2013. Inj Prevention , 2016; 22: 3– 18. Google Scholar CrossRef Search ADS 10 McDonnell S, Vossberg K, Hopkins R, Mittan B. Using YPLL in health planning. Public Health Rep 1998; 113: 55. Google Scholar PubMed 11 Weerasinghe DP, Yusuf F, Parr NJ. Life lost due to premature deaths in New South Wales, Australia. Int J Enviro Res Public health 2009; 6: 108– 20. Google Scholar CrossRef Search ADS 12 Hastie T, Tibshirani R. Generalized additive models. Statistical Sci 1986; 1: 297– 310. Google Scholar CrossRef Search ADS 13 Chambers JM, Hastie T, editors. Statistical Models in S. Vol. 251. Pacific Grove, CA: Wadsworth & Brooks/Cole Advanced Books & Software, 1992. 14 National Board on Health and Welfare. Dödsorsaksregistret (DOR). Cause of death register. Available at: http://www.socialstyrelsen.se/statistik/statistikdatabas/dodsorsaker (March 2018, date last accessed). 15 Statistics Sweden. Återstående medellivslängd efter kön och ålder 1972-2014. Life tables by sex and age 1972–2014. Available at: http://www.statistikdatabasen.scb.se/pxweb/sv/ssd/START__BE__BE0101__BE0101I/LivslangdEttariga/?rxid=b0ebb78f-ae78-45f6-9f09-1dd5719b1ce2 (March 2018, date last accessed). 16 Brooke HL, Talbäck M, Hörnblad J, et al. The Swedish cause of death register. Eur J Epidemiol 2017; 32: 765– 773. Google Scholar CrossRef Search ADS PubMed 17 Janssen F, Kunst AE. ICD coding changes and discontinuities in trends in cause-specific mortality in six European countries, 1950-99. Bulletin of the World Health Organization 2004; 82: 904– 13. Google Scholar PubMed 18 Wasserman D. Suicidprevention. Hämtad 2013-04-24 från https://www. msb. se/Upload/Forebyggande/kommunala_hp/Skadeprevention/Suicidprevention. pdf Figur 7. Lättare eller svårare att berätta via Internet om det du sökt stöd för jämfört med vid ett personligt möte; 2009. 19 Gurevičius R, Drūtytė G. Years and valued years of life lost due to external causes of death in Lithuania, 2000–2006. Higienos institutas 2009; 73. 20 Carlsson F, Daruvala D, Jaldell H. Preferences for lives, injuries, and age: a stated preference survey. Accid Anal Prev 2010; 42: 1814– 21. Google Scholar CrossRef Search ADS PubMed 21 Sordo L, Barrio G, Bravo MJ, et al. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ 2017; 357: j1550. Google Scholar CrossRef Search ADS PubMed 22 Leifman H. Drug-related deaths in Sweden. Estimations of trends, effects of changes in recording practices and studies of drug patterns. Available at: http://www.can.se/Publikationer/rapporter/drug-related-deaths-in-sweden/: Centralförbundet för alkohol- och narkotikaupplysning, CAN, 2016, (May 2018, date last accessed). 23 Case A, Deaton A. Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. Proc Natl Acad Sci 2015; 112: 15078– 83. Google Scholar CrossRef Search ADS PubMed 24 Stein EM, Gennuso KP, Ugboaja DC, Remington PL. The epidemic of despair among White Americans: trends in the leading causes of premature death, 1999–2015. Am J Pub Health 2017; 107: 1541– 7. Google Scholar CrossRef Search ADS 25 Case A, Deaton A. Mortality and morbidity in the 21st century. Brookings papers on economic activity 2017; 2017: 397. Google Scholar CrossRef Search ADS PubMed 26 Lund J, Aarø LE. Accident prevention. Presentation of a model placing emphasis on human, structural and cultural factors. Safety Sci 2004; 42: 271– 324. Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
The European Journal of Public Health – Oxford University Press
Published: May 26, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera