• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of croatmedjFree full text at www.cmj.hrAboutSubscribeSubmitInfo for AuthorsFree full text at www.cmj.hr
Croat Med J. Feb 2006; 47(1): 134–141.
PMCID: PMC2080367

Burden of Cancer in Serbia



To provide a comprehensive assessment of burden of selected cancers in Serbia.


We calculated disability adjusted life years (DALY) – the sum of the years of life lost (YLL) from premature mortality and the years lived with disability (YLD) – for cancers of stomach, colon and rectum, lung, breast, and cervical cancer for central Serbia and Vojvodina, Serbia and Montenegro. The obtained values were compared with the corresponding values for European region as estimated by the World Health Organization. The study was conducted between October 2002 and September 2003. The cancer burden was estimated for the year 2000.


Observed cancers were responsible for 133 689 DALYs (73 197 for men and 60 482 for women). There were significantly more losses because of premature death than disease disability (95.2% vs 4.8% in men P<0.001, and 93.2% vs 6.8% in females, P<0.001). The cancer burden was dominated by lung cancer in men and breast cancer in women. The cancer burden was very small before the age of 35.


DALYs per 1000 population were higher in Serbia than in theEuropean region for all observed cancers except for stomach cancer. The participation of a burden caused by disability in the total burden of selected cancers was lower in Serbia than in other European countries, with the greatest differences in colorectal, breast, and cervical cancers.

There is growing literature on the use of summary measures of population health, which combine the information on mortality and non-fatal health outcomes (1,2). One of them, called disability-adjusted life years (DALY) was introduced by the World Bank (3) and subsequently used by the World Health Organization and Harvard University for detailed assessment of global burden of disease and injuries (4). Afterwards, many investigators have performed and published studies on burden of diseases and injuries using DALY, which reflect both the years of life lost because of premature death and years lived with disability (5-7).

Morbidity and mortality of cancers have a great impact on population of Serbia. During many decades, malignant neoplasms have occupied the second place as a cause of death, with mortality rates for some cancers showing a slow but steady increase (8,9). In 2000, cancer accounted for 19% of all causes of deaths (21% in men, and 17% in women) in the population of Serbia without Kosovo and Metohia. In 2000, stomach cancer (International Classification of Diseases, version 10 [ICD10], code C16), colorectal cancer (ICD10 codes C18 – C21), lung cancer (ICD10 codes C33 – C34), breast cancer (ICD10 code C50), and cervical cancer (ICD10 code C53) were responsible for 50% of all cancer incidence cases (45% in men and 54% in women) and for 65% of deaths caused by all malignant neoplasms (63% in men, and 66% in women) (10). Lung cancer, colorectal cancer, and stomach cancer were among ten leading causes of deaths in men, whereas breast cancer, colorectal cancer, lung cancer, stomach cancer, and cervical cancer were among twelve leading causes of death in women (11). We chose these cancers to provide a comprehensive assessment of premature mortality and disability attributable to cancers in Serbia by calculating disability adjusted life years (DALY).


The Serbian Burden of Disease Study was carried out by the European Agency for Reconstruction Project Team, using methods largely based on those developed for the Global Burden of Disease study (4). The project commenced in October 2002 and was completed in September 2003. The study was conducted in Central Serbia and Vojvodina, Serbia and Montenegro.

The burden of selected cancers was estimated for the year 2000 by the use of DALY, which combines a measurement of premature mortality and disability (12-15). This indicator is the aggregation of years of life lost (YLL) and years lived with disability (YLD) at the population level and thus reflects the “burden of disease” in a population (DALY = YLL + YLD).

Population data

The population estimates for the calculation of values of needed indicators were based on the census adjustments for 2000, not including the migration component (16). The total number of observed population was 7 551 000. Direct method, with the European population as the standard, was used for standardization (17).

Mortality data

The Regional Offices of Statistics compile information on death certificates. The Serbian Office of Statistics receives these data from each region and provides a unit record file of deaths with diagnosis, death date, age, sex, and place of residence. For the burden of disease estimates, we considered all deaths of people with place of residence in Serbia that occurred anywhere in or out of Serbia and were registered in 2000.

The completeness of Serbian Office of Statistics 2000 mortality database was 98%. In addition to missing data, two other problems with the validity and reliability of cause of death data were present: not assigning a specific code but using “senility” or some other ill-defined code, and assigning the wrong code due to diagnostic fashion or carelessness.

According to the Global Burden of Disease methodology, ill-defined causes of deaths were reassigned. The 8.7% of deaths assigned to ill-defined and senility codes (U161) were redistributed proportionally, by sex and age, across the group of communicable, maternal, perinatal, and nutritional conditions among those 0-5 years old, and across the group of non-communicable diseases among those >5 years old. Cancers of unspecified sites (U077), which made 3.4% of all deaths, were redistributed proportionally by age and sex across all specified sites. Finally, 0.1% of deaths with missing age data were redistributed proportionally by sex and cause.

YLL was determined by the average life expectancy at the age of death while discounting future years by 3%. The 3% discount rate, recommended by the International Panel on Cost Effectiveness in Health and Medicine (18), was chosen to allow for international comparisons. YLL were calculated by multiplying age-specific mortality rates by age-specific standard expected for YLL and population numbers. Standard expected YLL were derived from the standard life table West 26 (19) with a life expectancy at birth of 82.5 years for women. Life table West 25 (20) for women was used for men with life expectancy at birth of 80 years.

The Global Burden of Disease Study weighted a year of healthy life lived at young age and older age lower than for other ages, because there is a broad social preference to value a year lived by a young adult more highly than a year lived by a young child or at older ages (4). Although not all investigators agree that young age and older age should be given less weight, in Serbian Burden of Disease Study 2000, we decided to use this approach to be able to compare our results with the results of other studies.

Incidence data

YLD was calculated on the basis of incidence data. The incidence data were obtained from the Cancer Registry of central Serbia for 1999 (10) and from Cancer Registry of Vojvodina for 1998 (21). From data for Central Serbia, only those for Belgrade population were used because they were more reliable than the data for the rest of the region. Global Burden of Disease (22), Dutch study (23) models, and provisional disability weights for cancer stages were used for all cancers. Since Dutch weights do not exist for cervical cancer, we applied weights used in Victoria study (24).

It was accepted that, according to the Global Burden of Disease estimate, 20% of cancers were untreated in the region of the former socialist economies of Europe (4).

For all cancers observed, including breast cancer, we assumed patients who survived five years to be in remission and we took the five-year survival rate as the cure rate. Patients who were cured were assumed to be without disability (22). For patients who died, average time to death was assumed to follow an exponential distribution, so that the mean survival time was estimated by fitting this distribution to available survival data (after 1, 3, and 5 years) by Kaplan-Meier product limit method (24,25). The duration of primary diagnosis, therapy, remission, and preterminal phase of each cancer were discussed with experts. The experts’ opinion was also asked for the frequency of small cell and non-small cell lung cancer, both operable and inoperable ones, and for the incidence of breast cancer by size. Since there are no data on survival of cancer patients in Serbia, we used 5-year survival data for Slovenia in the period 1985-1989 (26). Both Serbia and Slovenia had been part of former Yugoslavia before 1991.


The incidence and mortality rates for selected malignant neoplasms in central Serbia and Vojvodina in 2000 (Table 1) were compared for reliability with the same data for Slovenia (27) in 1999 (Table 2). Cancer incidence rates in these two populations were similar, although the frequency of lung cancer and cervical cancer was higher in Serbia, and the frequency of stomach and colorectal cancers was higher in male population of Slovenia. The incidence was almost the same for breast cancer. Mortality rates of stomach and colorectal cancers were similar, but mortality rates for lung, breast, and cervical cancers were higher in Serbia.

Table 1
Incidence and mortality rates from selected cancers by sex in Serbia, 2000*
Table 2
Incidence and mortality rates from selected cancers by sex in Slovenia, 1999*

Observed cancers were responsible for 73 197 DALYs in men and for 60 482 DALYs in women. There were significantly more YLL than YLD losses for all cancers (95.2% vs 4.8% in men, and 93.2% vs 6.8% in women), showing that the burden of cancer was dominated by mortality rather than long-term disability (Table 3).

Table 3
Burden of selected cancers by site and sex in Serbia, 2000*

Lung cancer accounted for the largest part of the cancer burden in men, followed by colorectal and stomach cancers, whereas the burden of breast cancer was negligible. For women, the burden of breast cancer was the greatest, followed by lung cancer, colorectal cancer, cervical cancer, and stomach cancer (Table 3).

The cancer burden was very small before the age of 35 (Figure 1). The DALY rates per 1000 male population were the highest in the 55-74 age group for lung cancer and in the 65-74 age group for stomach and colorectal cancer. For women, DALY rates were the highest in the 55-74 age group for lung cancer, and in those aged ≥65 years for stomach and colorectal cancers. The burden of breast and cervical cancers was greater in younger age groups, ie, those aged 45-64 years and 45-54 years, respectively.

Figure 1
The burden of cancer by site, sex and age in Serbia, 2000. Triangles – stomach cancer; open circles – colorectal cancer; rhombs – cervical cancer; squares – breast cancer; closed circles – lung cancer.

We compared the incidence, YLD, and DALY per 1000 population, as well as YLD/DALY ratio (%) of selected cancers in Serbia and European region divided in European subregions A, B, and C (Table 4). The incidence of breast cancer in Serbia was almost the same as that in the European region. The incidence of stomach and colorectal cancers was lower and the incidence of lung and cervical cancers was higher in Serbia. The most pronounced difference was found in the incidence of cervical cancer, which was 2.5 times higher in Serbia than in the European region. DALY per 1000 population in Serbia was higher than in the European region for all observed cancers, except for stomach cancer. The differences between Serbia and the European region in participation of YLD and YLL in the total burden (DALY) were presented through YLD/DALY ratio. The proportion of YLD in DALY for stomach and lung cancers in Serbia was similar to that estimated for the European region. For colorectal, breast, and cervical cancer, the proportion of YLD in DALY was smaller even in comparison to European B and European C regions.

Table 4
Incidence, years lived with disability (YLD)*, disability-adjusted life years (DALY) per 1000 population (all persons), and YLD/DALY ratio from selected cancers in Serbia and European region

The burden of the same cancers was also estimated by Dutch weights used for calculation of YLD (Table 5). Since Dutch weights were much higher, the values of YLD and DALY were also higher than those obtained by Global Burden of Disease study weights. The same was true for YLD/DALY ratios. Participation of YLD in DALY for lung cancer in Serbia was similar to that found in Victoria study (24), but YLD/DALY ratio for stomach, colorectal, breast, and cervical cancers in Serbia was lower.

Table 5
The burden of cancers by site in the population of Serbia and selected countries, 2000


The total burden of lung, stomach, colorectal, breast, and cervical cancers in Serbia in 2000 amounted to 133 689 DALYs. There were significantly more losses on account of premature death than disease-related disability. The participation of a burden caused by disease-related disability in the total burden of selected cancers was lower in Serbia than in other European countries, with the greatest differences being found for colorectal, breast, and cervical cancers.

Although we considered our mortality data reliable, percentage of deaths assigned to ill-defined codes was not small. However, we could expect that among ill-defined causes of deaths, cardiovascular diseases were more frequent than malignant diseases.

On the basis of comparison with cancer incidence in Slovenia (27), we assumed that data from the Cancer Register for Belgrade and for Vojvodina were reliable enough to estimate cancer incidence in Serbia. Slovenia was chosen for comparison for several reasons. Serbia and Slovenia had been part of former Yugoslavia before 1991, the health service in Slovenia was closely related to that in Serbia, and Cancer Registry in Slovenia was one of the oldest services of this type in Europe (26). With some exceptions, incidence rates of observed cancers were similar in these two populations. Higher incidence rates of lung and cervical cancers in Serbia were expected, because smoking rate is one of the highest in Europe and screening for cervical cancer has not been established yet.

Method used for estimation of DALY was largely based on the one developed for the Global Burden of Disease study (4). After discussion with experts for selected cancers, the Global Burden of Disease model and provisional disability weights for cancer stages were used (22). The estimation of YLD by Dutch disease model and disability weights (23) was also performed to allow for comparison with other population data analyzed by the same method. As expected, however, we obtained similar information from these two models, at least when the burden of cancer is concerned. Only the participation of YLD in DALY was higher when Dutch weights were applied. In Australian Burden of Disease study (5), the two sets of weights were used concurrently because there was a high correlation between them for 54 conditions, indicating that the two sets of weights valued the same conditions similarly. Although it is possible to create own disease model and to estimate own disability weights, in our opinion, the same disease model and disability weights should be used in all studies to make comparison between populations possible. Since we had no reliable data about treated and non-treated cases, we accepted that, according to the Global Burden of Disease estimate, 20% of the cancer cases were untreated (4).

We used 5-year survival data for Slovenia from 1985 to 1989 (26) because it was more appropriate than to use data on some other neighboring country or Europe as a whole. The reason was that in Serbia we could not expect positive changes in survival during the preceding 10 years, ie, in the period 1991-2000. The civil strife in 1991, which led to the war and the break-up of the former Yugoslavia, United Nation’s economic sanctions imposed on Serbia and Montenegro, and the ensuing economic crisis had a highly adverse effect on the population health and the quality of health care (31).

We may conclude that DALYs, used as the summary measure of population health in the present study, can be considered a good estimate of cancer burden in Serbia. The differences in cancer burden and especially the differences in YLD/DALY ratio between Serbia and European region, and between Serbia and Victoria, New Zealand, and the Netherlands, could be explained by higher incidence and/or higher fatality of almost all observed cancers in Serbia. Early diagnosis and early and adequate therapy of lung cancer and to a less extent of stomach cancer do not influence significantly their prognosis. That is why YLD/DALY ratio for these cancers is similar both in the developed and in the developing countries. On the other hand, early diagnosis and therapy of cervical, breast, and colorectal cancers have considerable influence on their outcome. Consequently, the participation of YLD in DALY is lower in countries like Serbia, in which regular screening for these cancers has not been organized yet, than in countries in which early detection was established years ago. It should be also mentioned that YLD/DALY ratio can be affected by age distribution of the diseased – the participation of YLL in DALY is greater when fatal disease occurs in younger age groups.

The most important advantage from a policy perspective of the approach taken in this study is that estimates of burden of cancers are expressed in terms of a summary health-outcome measure, which combines mortality and morbidity. These estimates should guide future health strategies and interventions and would allow for monitoring of the improvements in health and the performance of the health care system. However, there are views that burden of disease assessment should be supplemented with positive measures of health expectancy or health adjusted life expectancy (HALE). According to Tobias (28), these two families of summary measures of population health could be measured in such a way to make them not only conceptually, but also quantitatively complementary. In any case, the estimate of disability adjusted life years for selected cancers should be repeated after several years in order to assess the effects of preventive strategies, which are in the process of implementation.


This work was supported by European Agency for Reconstruction (EAR), 2002-2003, and by the Ministry of Science and Ecology of Serbia, contract No. 101927, 2002-2005.


1. Field M, Gold MR, editors. Summarizing population health: directions for the development and application of population metrics. Washington (DC): National Academy Press; 1998.
2. Murray CJ, Salomon JA, Mathers CD, Lopez AD, editors. Summary measures of population health: concepts, ethics, measurement and applications. Geneva: World Health Organization; 2002.
3. World Bank. World development report 1993: investing in health. New York: Oxford University Press; 1993.
4. Murray CJ, Lopez AD, editors. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Global Burden of Diseases and Injury Series, Vol. 1. Cambridge: Harvard University Press; 1996. p. 125-33.
5. Mathers CD, Vos TE, Stevenson CE, Begg SJ. The burden of disease and injury in Australia. Bull World Health Organ. 2001;79:1076–84. [PMC free article] [PubMed]
6. Melse JM, Essink-Bot ML, Kramers PG, Hoeymans N. A national burden of disease calculation: Dutch disability-adjusted life-years. Dutch Burden of Disease Group. Am J Public Health. 2000;90:1241–7. [PMC free article] [PubMed]
7. Bowie C, Beck S, Bevan G, Raftery J, Silverton F, Stevens A. Estimating the burden of disease in an English region. J Public Health Med. 1997;19:87–92. [PubMed]
8. Vlajinac H, Marinkovic J, Kocev N, Adanja B, Sipetic S, Pekmezovic T, et al. Mortality trend in Serbia, excluding the provinces, 1973-1994. Srp Arh Celok Lek. 2000;128:309–15. [in Serbian]. [PubMed]
9. Adanja B, Gledovic Z, Pekmezovic T, Vlajinac H, Jarebinski M, Zivaljevic V, et al. Mortality trends of malignant tumours of digestive organs in Belgrade, Yugoslavia, 1975-1997. Dig Liver Dis. 2000;32:386–91. [PubMed]
10. Cancer registry of Central Serbia – Cancer incidence and mortality in Central Serbia, 1999. Beograd: Dr Milan Jovanović – Batut Institute of Public Health of Serbia, report No. 1; 2002.
11. Serbian Office of Statistics. Mortality database 2000. Belgrade: SOS; 2000.
12. Anand S, Hanson K. Disability-adjusted life years: a critical review. J Health Econ. 1997;16:685–702. [PubMed]
13. Murray CJ, Acharya AK. Understanding DALYs (disability-adjusted life years). J Health Econ. 1997;16:703–30. [PubMed]
14. Williams A. Calculating the global burden of disease: time for a strategic reappraisal? Health Econ. 1999;8:1–8. [PubMed]
15. Murray CJ, Lopez AD. Progress and directions in refining the global burden of disease approach: response to Williams. Health Econ. 2000;9:69–82. [PubMed]
16 Atanasković-Marković Z, Bjegović V, Janković S, Kocev N, Laaser U, Marinković J, et al. Burden of disease and injury in Serbia. Belgrade: Republic of Serbia Ministry of Health; 2003. Available from: www.sbds.sr.gov.yu. Accessed: October 25, 2004.
17. Ahmad OB, Boschi-Pinto C, Lopez A, Murray CJ, Lozano R, Inoue M. Age standardization of rates. A new WHO standard. Global Programme on Evidence for Health Policy Discussion Paper Series: No. 31. Geneva: WHO; 2000. p. 3.
18. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. New York (NY): Oxford University Press; 1996.
19. Coale A, Guo G. Revised regional model life tables at very low levels of mortality. Popul Index. 1989;55:613–43. [PubMed]
20. Coale A, Demeny P, Vaughan B. Regional model life tables and stable populations. 2nd ed. New York (NY): Academic Press; 1983.
21. Institute of Oncology Sremska Kamenica. Cancer registry of Vojvodina – cancer incidence and mortality in Vojvodina, 1998. Novi Sad: Institute of Oncology Sremska Kamenica; 2001.
22 Mathers C, Boschi-Pinto C. Global burden of cancer in the year 2000: Version 1 estimates. In: Global burden of disease 2000. Geneva: WHO; 2003. Available from: http://www3.who.int/whosis/. Accessed: November 10, 2004.
23. Stouthard ME, Essink-Bot ML, Bonsel GJ, Barendregt JJ, Kramers PG, van de Water HP, et al. Disability weights for diseases in the Netherlands. Rotterdam: Department of Public Health, Erasmus University; 1997.
24. Department of Human Services (DHS). The Victorian burden of disease study: morbidity. Melbourne: Public Health and Development Division; 1999.
25. Lee ET. Statistical methods for survival data analysis. Belmont (CA): Lifetime Learning Publications; 1980.
26. Berrino F, Capocaccia R, Esteve J, Gatta G, Hakulinen T, Michele A, et al, editors. Survival of cancer patients in Europe. The EUROCARE-2 Study. Lyon, France: IARC Scientific Publication No. 151; 1999. p. 320.
27. Cancer Register of Slovenia. Cancer incidence in Slovenia 1999. Ljubljana: Institute of Oncology Ljubljana. Report No. 41; 2002.
28 Tobias M, Borman B, Forbes G, Hemona P. The burden of disease and injury in New Zealand. Public Health Intelligence Occasional Bulletin No.1. Wellington: Ministry of Health; 2001. Available from: http://www.moh.govt.nz Accessed: April 12, 2003.
29. Stouthard ME, Essink-Bot ML, Bonsel GJ, Dutch Disability Weights Group. Disability weights for diseases: a modified protocol and results for a Western European region. Eur J Public Health. 2000;10:24–30.
30. Kruijshaar ME, Barendregt JJ, European Disability Weights Group. The breast cancer related burden of morbidity and mortality in six European countries: the European Disability Weights project. Eur J Public Health. 2004;14:141–6. [PubMed]
31. Vlajinac HD, Marinkovic JM, Kocev NI, Adanja BJ, Pekmezovic TD, Sipetic SB, et al. Infectious disease mortality in Central Serbia. J Epidemiol Community Health. 1997;51:172–4. [PMC free article] [PubMed]

Articles from Croatian Medical Journal are provided here courtesy of Medicinska Naklada
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • PubMed
    PubMed citations for these articles