Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Incidence and mortality of hip fracture among the elderly population in South Korea: a population-based study using the National Health Insurance claims data

Incidence and mortality of hip fracture among the elderly population in South Korea: a... Background: The lack of epidemiologic information on osteoporotic hip fractures hampers the development of preventive or curative measures against osteoporosis in South Korea. We conducted a population-based study to estimate the annual incidence of hip fractures. Also, we examined factors associated with post-fracture mortality among Korean elderly to evaluate the impact of osteoporosis on our society and to identify high-risk populations. Methods: The Korean National Health Insurance (NHI) claims database was used to identify the incidence of hip fractures, defined as patients having a claim record with a diagnosis of hip fracture and a hip fracture-related operation during 2003. The 6-month period prior to 2003 was set as a 'window period,' such that patients were defined as incident cases only if their first record of fracture was observed after the window period. Cox's proportional hazards model was used to investigate the relationship between survival time and baseline patient and provider characteristics available from the NHI data. Results: The age-standardized annual incidence rate of hip fractures requiring operation over 50 years of age was 146.38 per 100,000 women and 61.72 per 100,000 men, yielding a female to male ratio of 2.37. The 1-year mortality was 16.55%, which is 2.85 times higher than the mortality rate for the general population (5.8%) in this age group. The risk of post-fracture mortality at one year is significantly higher for males and for persons having lower socioeconomic status, living in places other than the capital city, not taking anti-osteoporosis pharmacologic therapy following fracture, or receiving fracture-associated operations from more advanced hospitals such as general or tertiary hospitals. Conclusion: This national epidemiological study will help raise awareness of osteoporotic hip fractures among the elderly population and hopefully motivate public health policy makers to develop effective national prevention strategies against osteoporosis to prevent hip fractures. Background society in terms of premature death, long-term disability, Osteoporosis has become a significant public health restricted social functioning, cost of care, lost productiv- problem in recent years, especially with the growth of the ity, and informal caregiver time. Previous studies have elderly population [1]. Hip fracture is the most valid epi- shown that the incidence rates of hip fractures for Asian demiologic marker of osteoporosis, a major cause of mor- people were lower than those for Whites living in North- bidity and mortality in the aging population [1-3], and ern Europe and North America [4-8]. However, recent imposes a considerable burden on patients, payers, and reports have shown an increase in the incidence of hip fractures in Asian populations [4,9,10], and it is projected that 45-70% of global hip fractures will occur in Asia by * Correspondence: hykang@yuhs.ac Graduate School of Public Health, Institute of Health Services Research, 2050 due to the growth of the elderly population [1,11]. Yonsei University, Seoul, South Korea Full list of author information is available at the end of the article © 2010 Kang et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in BioMed Central any medium, provided the original work is properly cited. Kang et al. BMC Public Health 2010, 10:230 Page 2 of 9 http://www.biomedcentral.com/1471-2458/10/230 To reduce the cost of fracture care, it is important to validity of the incidence cases identified from insurance adopt effective prevention strategies to reduce osteopo- claims data. rosis, including dietary supplements such as calcium and All claims records of outpatient visits or hospital vitamin D and pharmacologic therapy. However, the lack admissions of patients aged 50 or older containing a diag- of detailed data on hip fracture epidemiology, particularly nosis of femur fracture (fracture of femur [International in developing countries, hampers the development of Classification of Diseases (ICD)-10 diagnostic code: S72], preventive or curative measures against osteoporosis fracture of the neck of the femur [S72.0, S72.00], pertro- [12]. Thus, knowledge regarding the incidence of hip chanteric fracture [S72.1, S72.10]) and hip fracture- fractures and their outcomes, including mortality, is related operation (open reduction & internal fixation needed to evaluate the impact of osteoporosis on our [ICD-10 procedure code: N0601], closed reduction and society, to identify high-risk populations, and to help percutaneous fixation [N0991], total hip replacement health care decision-makers allocate scarce resources and [N0711], or hip hemiarthroplasty [N0715]) from January prioritize hip fractures along with other important clini- 1 to December 31, 2003 were identified from the NHI cal problems. claims database. The diagnosis and operation code for While the growth of the elderly population (>65 years hip fracture was selected based on previous epidemio- of age) in South Korea has been as marked as in other logic studies [1,14] and was confirmed by a panel of four Asian countries, rising from 3.8% in 1980 to 7.9% in 2002 orthopedic clinicians working in four different general [13], epidemiologic information about hip fractures hospitals in Korea. The cases having more than one claim among the elderly in South Korea is scarce. To provide a record that satisfied the inclusion criteria during 2003 national estimate on the incidence rate of hip fractures in were counted only once. this population, we need a result derived from a large, In general, not all operations are carried out during the population-based database for Korea. The Korean first fracture visit or admission. However, most of the National Health Insurance (NHI) claims database is such operations are performed within a month following the a database, containing all medical and prescription drug- first visit, according to a consultation of orthopedic clini- claim records for almost 100% of the Korean population. cians in Korea. Thus, we additionally defined the inci- South Korea operates a mandatory universal health dence cases as patients who did not have a record of a hip insurance system, with a centralized healthcare claims fracture-related operation in the claim record of the ini- database that provides a unique nationwide source of tial visit or admission, but had it within a month after the information on health care resource utilization. We used initial visit. this database to estimate the annual incidence and 1-year The 6-month period prior to 2003 (i.e., July - Decem- mortality of hip fractures in Koreans over the age of 50. ber, 2002) was set as a 'window period,' such that patients Furthermore, we assessed the patient and provider char- were defined as incident cases only if their first record of acteristics associated with post-fracture mortality in a fracture visit or admission was observed after this 6- order to help identify high-risk patients with poor health month period. Since most of the follow-up treatments for outcomes and develop effective treatment strategies after hip fracture are completed within 6 months after the ini- hip fracture incidence. The study results are expected to tial fracture, we assumed that the absence of any claims increase our understanding of the magnitude of the record with a diagnosis of hip fracture and hip-fracture- elderly population suffering from hip fractures and to related operation in the previous 6 months assured that urge public health policy makers to develop effective the fracture was a new case. national prevention strategies for osteoporosis. NHI claims data were merged with national mortality data provided by the National Statistical Office to deter- mine the survival status of individual patients at the 12 Methods Data sources and study population months following the incidence. Records from the Korean NHI claims database from 2002 Data analysis and 2004 were used to identify patients with hip fractures Age- and sex-specific annual incidence rates of hip frac- and to monitor their 1-year mortality. The incidence of tures per 100,000 inhabitants were calculated and pre- hip fractures was defined as patients having a claim sented in 5-year age intervals. The number of inhabitants record with a diagnosis of hip fracture and a hip fracture- in South Korea, as of 2003, was obtained from the Korean related operation. Since identifying hip fracture cases National Statistical Office. The age-standardized inci- solely based on diagnosis is likely to cause misclassifica- dence rate was computed using the Korea Census data of tion problems due to potential miscoding, we combined 2000. Age- and sex-specific 1-year mortality rates of hip the information from both diagnosis and surgical records. fractures were calculated and presented in 5-year age This conservative approach underestimates the real inci- intervals. To examine the difference in the incidence and dence rate of hip fractures in Korea, but improves the Kang et al. BMC Public Health 2010, 10:230 Page 3 of 9 http://www.biomedcentral.com/1471-2458/10/230 mortality of hip fractures according to gender, the gender affecting patient survival after hip fracture. As compared ratio of incidence and 1-year mortality rate was com- to the patients residing in Seoul, the capital city of South puted for each 5-year age stratum. Korea, those living in other parts of South Korea had a Cox's proportional hazards model was used to investi- significantly higher risk of death following fracture, gate the relationship between survival time and baseline which showed a hazard ratio ranging from 1.291 (95% patient and provider characteristics. The independent confidence interval (CI): 1.120-1.488) for those in metro- variables were age, gender, type of fracture, type of opera- politan cities to 1.656 (95% CI: 1.344-2.039) for those in tion, baseline health status, residence area, type of health non-city areas. insurance, provision of anti-osteoporosis medication To examine the impact of socioeconomic status on the therapy after fracture, and type of hospital providing mortality of hip fractures, we compared the mortality operation. between elderly enrolled in the two tiers of the Korean universal health insurance system: the Health Insurance (HI) and the Medical Aid (MA) program. The HI is a Results During the 1-year observation period from January wage-based, contributory insurance program covering through December 2003, there were 9,817 incidents of about 97% of the population, whereas the MA is a gov- hip fractures in persons aged 50 and older in South ernment-subsidized, public assistance program for the Korea: 6,892 females (70.2%) and 2,925 males (29.8%). poor and the medically indigent. The NHI claims data The mean (± SD) and the median age were 74.92 (± 9.13) include records for both the HI and the MA program and 76 years old, respectively. The mean age was younger populations. The survival analysis showed that patients in men (72.10 ± 9.47 years) than in women (76.11 ± 8.72 enrolled in the MA program had a 30% higher chance of years). The age-standardized annual incidence rate of hip death during the year following fracture than those in the fractures was 104.06 per 100,000; 146.38 per 100,000 HI program (hazard ratio: 1.300, 95% CI: 1.136-1.488). women and 61.72 per 100,000 men, yielding a female/ As a proxy measure for patients' baseline health status, male ratio of 2.37 (Table 1). a Charlson comorbidity index [15] was assigned to each The incidence rate of hip fractures increased with study subject based on the diagnosis shown in the NHI increasing age, from 6.74 per 100,000 for the youngest claims record during the study period prior to hip frac- group of 50-54 year olds to 686.82 per 100,000 for the ture. Charlson et al. developed a scoring system for oldest group of 85 years and older. This increase followed comorbid illness, such that each comorbid condition was an exponential relationship after age 65 for both genders assigned a whole number integer ranging from 1 to 6, (Figure 1). About 34.0% of all hip fractures occurred in which was proportional to the relative risk of 1-year mor- patients aged 80 years and older, although this age group tality associated with that disease [16]. Not surprisingly, comprised only 5.6% of the total population over 50. those who had had severe comorbid conditions before The gender ratio of the female to male incidence rate they had the fracture outbreak were 1.339 (for those with was computed for the 5-year age stratum. The ratio for Charlson Index of 1), 2.625 (Index of 2), or 5.693 (Index of the youngest 5-year group (50-54 years old) was smaller 3 or 6) times more likely to die as compared to those than 1, indicating a higher incidence rate among males without any of the Charlson Index conditions. than among females, probably due to occupation. How- The risk of mortality following fracture was not signifi- ever, this ratio increased with increasing age and was cantly different according to the type of hip fracture or greater than 1 for all groups older than or equal to 55 operation. The likelihood of dying among those with years old. claims records showing at least one reimbursed prescrip- Of 9,817 patients, 1,625 died during the first year post- tion for anti-osteoporosis drugs (alendronate, etidronate, fracture, yielding a mortality rate of 16.55% (Table 2). The raloxifen, risedronate, and salcatonin) following hip frac- mortality rate increased with increasing age from less ture was 42.4% lower than those without such claims than 10% of patients aged younger than 65 to more than records. 20% of those over 80 years old. Male patients consistently As the level and the size of the hospital in which the showed a higher mortality rate than female patients in all patients received their operation decreased, so did the age categories except the youngest, with the ratio of chance of death. The hazard ratios of 1-year mortality, female to male mortality ranging from 0.48 to 0.72, com- with tertiary hospitals as a reference group, were 0.798 pared with 3.77 in the youngest group. (95% CI: 0.680-0.937) for hospitals and 0.746 for clinics Patient and provider characteristics available from the (95% CI: 0.573-0.972). NHI claims data were examined for their association with 1-year post-fracture mortality using the multivariate Discussion Cox-proportional hazard model (Table 3). While other This is the first nation-wide, population-based, epidemi- factors were held constant in the multivariate model, gen- ologic study of hip fractures in Korea and one of the few der and age remained statistically significant factors studies worldwide reporting the incidence of hip frac- Kang et al. BMC Public Health 2010, 10:230 Page 4 of 9 http://www.biomedcentral.com/1471-2458/10/230 Table 1: Sex- and age-specific annual incidence rates of hip fractures in South Korea in 2003 Total Female Male Age No. No. hip fx. Incidence No. popl No. hip fx. Incid. Per No. popl No. hip fx. Incid. Per F/M ratio (years) population per 2003 100,000 2003 100,000 of incid. 2003 100,000 50-54 2,567,149 173 6.74 1,274,375 69 5.41 1,292,774 104 8.04 0.67 55-59 2,104,791 403 19.15 1,061,917 205 19.30 1,042,874 198 18.99 1.02 60-64 1,935,661 824 42.57 1,023,028 449 43.89 912,633 375 41.09 1.07 65-69 1,594,318 1,311 82.23 882,721 817 92.55 711,597 494 69.42 1.33 70-74 1,099,247 1,790 162.84 665,698 1,245 187.02 433,549 545 125.71 1.49 75-79 683,904 1,982 289.81 441,732 1470 332.78 242,172 512 211.42 1.57 80-84 380,112 1,888 496.70 259,291 1456 561.53 120,821 432 357.55 1.57 85+ 210,537 1,446 686.82 159,330 1181 741.23 51,207 265 517.51 1.43 Overall 10,575,719 9,817 92.83 5,768,092 6,892 119.48 4,807,627 2,925 60.84 1.96 (50+) Age-standardized rates* 104.06 146.38 61.72 2.37 *Age-standardized incidence rate was computed using Korea Census data of 2000. tures based on national health insurance claims data. The Singapore (442 and 164 per 100,000), and Thailand (269 age-standardized annual incidence rate for the entire and 114 per 100,000) [9]. population of Korea over 50 years of age in 2003 was esti- The risk of 1-year mortality following hip fracture was mated as 104.06 per 100,000 inhabitants, which is about 16.55%, which is about 2.85 times higher than the mortal- 27.8% lower than that estimated in an earlier study con- ity rate of the general population in South Korea over 50 ducted on inhabitants residing in selected regions of years of age. Despite the heterogeneity of the study popu- Korea in 2001 (133 per 100,000) [17]. The difference in lation in terms of age and ethnicity, and the potential dis- the estimated incidence rates between the two studies crepancy in the quality of follow-up care after the can be explained by the difference in the dataset, meth- fracture, the post-fracture mortality at 1 year observed in ods used in each study to identify incidence cases, and other countries is similar to that in Korea. A recent U.S. the population characteristics in the study regions. In the study has reported that about 11.5% of 950 patients over earlier study, hip fracture cases admitted to selected hos- age 65 admitted to selected hospitals with a hip fracture pitals in Gwangju City and Chonnam Province were iden- between 1987 and 1997 died after 1 year [19]. The 1-year tified on the basis of medical records and radiographs. mortality rate was 15.7% among 541 fracture patients Among the identified incidence cases in that study, 88% over 60 years of age admitted to one university hospital in underwent surgery. By applying this operation rate to the Greece between 1989 and 1992 [20]. In a population- incidence rate estimated by our study, the total number of based, case-control study in Sweden, about 10.6% of incidence cases with and without operation was esti- 2,245 women aged 50-81 years admitted with a hip frac- mated as 118.3 per 100,000 (= 104.6/0.88), thereby reduc- ture died within 1 year [21]. The comparison with other ing the gap in the incidence estimates between the two Asian countries also showed similar results. The 1-year studies from 27.8% to 12.4%. mortality of patients with surgically treated hip fractures After standardization according to the age distribution at a hospital in Taiwan between 1998 and 2006 was about of the US white population in 1990 [18], the incidence 14.0% [22]. A Japanese study by Muraki et al. reported rates of hip fractures in Korea were 173 per 100,000 in that the 1-year mortality rate following hip fracture was women and 91 per 100,000 in men. Interestingly, the 11.5% [23]. standardized incidence rates of both genders were sub- Osteoporosis is recognized as a problem predominantly stantially lower than those of most of the Asian countries in elderly women [24,25]. Consequently, women have including Taiwan (505 per 100,000 in women and 225 per higher incidence rates of hip fracture than men. In Korea, 100,000 in men), Hong Kong (459 and 180 per 100,000), the female to male ratio of the age-standardized inci- Kang et al. BMC Public Health 2010, 10:230 Page 5 of 9 http://www.biomedcentral.com/1471-2458/10/230 that observed in other Asian countries, such as Japan (209.0 to 780 per 100,000), Hong Kong (404 to 1,639), Singapore (611), Malaysia (320), and Thailand (227 to 727) [4,6]. The increased incidence with increasing age in both genders confirms the equal vulnerability of both Female Male genders to the aging process. Thus, preventive strategies are important for both genders. It is interesting to observe higher post-fracture mortal- ity in Korean males than in females. Gender differences in post-hip fracture mortality have been reported in ear- 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Age lier studies showing that men were about twice as likely as women to die during the first years after hip fracture Figure 1 Comparison of age-specific incidence rates of hip frac- ture according to gender. [26-28]. The observed difference between men and women could be explained by under treatment of osteo- porosis [28] and higher infection rates in men following dence rates of hip fractures for those aged 50 and older is hip fracture [26,27]. 2.37. This ratio is comparable to western countries such Low socioeconomic status among the elderly popula- as Sweden (2.2, 1987-1991), Norway (2.3, 1983-1984), tion seems to be associated with an increased risk of frac- Australia (2.7, 1989-1990), and in U.S. whites (2.7, 1983- ture-caused mortality. The adjusted hazard ratio of post- 1984), and in some Asian countries such as Thailand (2.3, fracture mortality was significantly higher among those 1997-1998), Hong Kong (2.5, 1997-1998), and Malaysia enrolled in the MA program than among those in the HI (2.4, 1997-1998) [3]. Hip fracture rates were higher in program. Patients with low socioeconomic status, in gen- women of all age groups except for those under 55 years. eral, have poorer health status and therefore tend to have This pattern is commonly observed in populations of poorer health outcomes from the same conditions as other countries, although the starting age at which compared to those with higher socioeconomic status women show a higher hip fracture rate varies from coun- [29,30]. Thus, more effective public health strategies to try to county: 60 in Taiwan and Japan [4,9], but mostly 55 treat osteoporosis and to prevent fracture incidence or 50 in other countries such as Mexico, Argentina, Iran, should be implemented that target the indigent elderly and Morocco [1-3,5]. population in Korea. Although women showed consistently higher incidence The regional variation in fracture-related mortality rates of hip fractures than men, the high incidence rates among Korean elderly is noteworthy. Those living in among older men is worthy of attention. After age 75, places other than the capital city consistently showed a Korean men experienced hip fractures at a rate of 211.42 higher risk of death during the first year after hip fracture, to 517.51 per 100,000 inhabitants. This rate is similar to with a minimum of 21.4% or a maximum of 65.6% addi- Table 2: Sex- and age-specific post-hip fracture mortality rate at 1 year Total Female Male Age No. hip No. Mortality No. hip No. Mortality No. hip No. Mortality F/M ratio of (years) Fx. 2003 mortality rate (%) Fx. 2003 mortality rate (%) Fx. 2003 mortality rate (%) mortality rate 50-54 173 14 8.09 69 10 14.49 104 4 3.85 3.77 55-59 403 21 5.21 205 7 3.41 198 14 7.07 0.48 60-64 824 76 9.22 449 32 7.13 375 44 11.73 0.61 65-69 1,311 150 11.44 817 78 9.55 494 72 14.57 0.66 70-74 1,790 216 12.07 1,245 125 10.04 545 91 16.70 0.60 75-79 1,982 347 17.51 1470 224 15.24 512 123 24.02 0.63 80-84 1,888 413 21.88 1456 293 20.12 432 120 27.78 0.72 85+ 1,446 388 26.83 1181 296 25.06 265 92 34.72 0.72 Overall 9,817 1,625 16.55 6,892 1,065 15.45 2,925 560 19.15 0.81 (50+) Incidence per 100,000 Kang et al. BMC Public Health 2010, 10:230 Page 6 of 9 http://www.biomedcentral.com/1471-2458/10/230 Table 3: Multivariate survival analysis: Patient and provider characteristics associated with 1-year mortality after hip fracture Variables N (%) Hazard ratio 95% confidence interval Gender Female 6,892 (70.2) 1.000 Male 2,925 (29.8) 1.400 (1.260-1.557) Age (years) 50-59 576 (5.9) 1.000 60-69 2,135 (21.7) 1.787 (1.250-2.553) 70-79 3,772 (38.4) 2.879 (2.041-4.062) 80-89 2,923 (29.8) 4.742 (3.361-6.691) ≥ 90 411 (4.2) 9.273 (6.387-13.463) Residence Seoul (Capital city) 2,388 (24.3) 1.000 Metropolitan city 2,584 (26.3) 1.291 (1.120-1.488) Small city 4,133 (42.1) 1.214 (1.058-1.392) Others (none-city area) 712 (7.3) 1.656 (1.344-2.039) Health National Health Insurance program 8,514 (86.7) 1.000 insurance National Medical Aid program 1,303 (13.3) 1.300 (1.136-1.488) Baseline 0 4,742 (48.3) 1.000 Charlson 1 4,038 (41.1) 1.339 (1.198-1.496) Index (before hip 2 938 (9.6) 2.625 (2.264-3.042) fracture) 3, 6 99 (1.0) 5.693 (4.249-7.627) Type of Neck fracture 5,435 (55.4) 1.000 hip fracture Petrochanteric fracture. 3,733 (38.0) 1.090 (0.966-1.230) Neck + Petrochanteric fracture. 649 (6.6) 1.031 (0.843-1.261) th Type of operation Total hip replacement (ICD 10 4,574 (46.6) 1.000 procedure code: N0711) Open reduction/internal fixation 4,592 (46.8) 0.963 (0.855-1.084) (N0601) Closed reduction/percutaneous 397 (4.0) 0.958 (0.720-1.274) fixation (N0991) Hip hemiarthroplasty (N0715) 216 (2.2) 0.793 (0.536-1.174) Total hip replacement & Open 38 (0.4) 1.213 (0.575-2.561) reduction (N0711 + N0601) Treatment* No 7,761 (79.1) 1.000 Yes 2,056 (20.9) 0.576 (0.499-0.666) Type of hospital providing Tertiary hospital 2,714 (27.6) 1.000 operation General hospital 4,649 (47.4) 0.900 (0.796-1.018) Hospital 1,901 (19.4) 0.798 (0.680-0.937) Clinic 553 (5.6) 0.746 (0.573-0.972) * Treatment was defined as whether the patient took at least one prescription of anti-osteoporosis medicine after hip fracture. Kang et al. BMC Public Health 2010, 10:230 Page 7 of 9 http://www.biomedcentral.com/1471-2458/10/230 tional chance of death. This discrepancy may be due to only, we will underestimate incidence rates by excluding the lack of access to health care providers or the poor less severe cases that are initiated with office visits. To quality of health care provided in non-city or small city overcome this problem, the present study defined inci- areas. Further investigation is necessary to figure out the dent cases either by initial admission or a visit containing reasons for this regional discrepancy and therefore to a diagnosis of hip fracture. However, while most hospital reduce this discrepancy. admissions coincide with an acute new episode of illness, From our analysis, patients receiving medication to it is difficult for outpatient visits to determine whether treat osteoporosis after fracture were 42.4% less likely to the case is a new episode or part of a post-episode course. die within the year following fracture. The association In addition, the accuracy of diagnostic codes among the between reduced mortality and post-fracture use of anti- claims data for office visits is lower than that of hospital osteoporosis drugs in elderly hip fracture patients has admission [38,39]. Thus, to improve the accuracy of iden- been addressed in earlier studies. A prospective analysis tifying incidence cases based on initial fracture visits, we from Finland has reported a 43% reduction in deaths at excluded patients with a claims record with a hip fracture 36 months following hip fracture in females who used during the 6 months prior to the initial visit. Further- prescribed calcium plus vitamin D supplementation con- more, to confirm the hip fracture cases, hip fracture- comitantly with anti-osteoporosis drugs [31]. Also, an related operation codes were used in addition to the diag- observational study from Canada has shown that mortal- nostic codes. ity is significantly lower in the group treated with anti- Several methodological issues arise from the studies osteoporosis drugs than in the untreated group following using administrative claims data. First, not all patients hip fracture [32]. Because elderly people with prior frac- with fractures have access to hospitals. The incidence tures are at a higher risk for future fractures [33], clinical based on insurance claims records would be underesti- practice guidelines recommend the initiation of pharma- mated if many hip fractures in Korea were not diagnosed cologic treatment of osteoporosis after the first fracture or treated in health care institutions. However, due to the [34]. However, under-treatment of osteoporosis [35,36] emergent character of hip fractures, it is believed that vir- and low adherence to oral anti-osteoporosis treatment tually all hip fractures present to health services [3]. following hip fracture [37] is commonly observed. In our Thus, the incidence of hip fractures identified from study, only 20.9% of the fractured patients received anti- claims records is considered to be close to the actual inci- osteoporosis drugs following hip fracture. Thus, a more dence. aggressive practice of diagnosing and providing medica- Second, the reliance on ICD-10 diagnostic codes to tion therapy for osteoporosis seems to be an urgent need identify incident fractures may cause misclassification of in Korea. incident hip fractures due to the inherent nature of claims Interestingly, patients that underwent an operation for data such as voluntary or non-voluntary miscoding hip fracture from hospitals or clinics tend to have a better behavior. However, a recent validation study for the diag- chance of survival than those from tertiary hospitals. nostic codes of the NHI claims database in Korea has alle- Although our multivariate model tried to include as many viated this concern. It has been revealed that about 70% patient characteristic variables as possible from the NHI of primary, secondary, or tertiary diagnosis codes from data, we could not comprehensively control for hip frac- NHI claims records coincide with those from medical ture severity and patient health conditions due to the lack records. In addition, the accuracy of diagnosis codes of clinical information available from the claims records. tended to be higher for claims from hospital admissions Thus, it cannot be concluded that the quality of care and compared with office visits, and for claims of severe con- patient outcomes after hip fracture treatment are better ditions compared with mild conditions [38,39]. among hospitals and clinics than among tertiary hospitals in Korea. It might be possible that patients with more Conclusion severe conditions underwent hip fracture repairs from In conclusion, our results indicate a high incidence and larger and more advanced hospitals (i.e., tertiary or gen- mortality rate of hip fractures for both men and women eral hospitals) and therefore the prognosis after surgery living in South Korea. These high incidence rates for both tended to be poorer. A more thorough adjustment for genders are sentinel signals of the impact of hip fractures patient baseline disease severity is necessary to make a in Korea, in terms of high mortality, morbidity, and concrete conclusion about this finding. health care costs for treatment. It is our hope that this Since a hip fracture is a severe condition, most of the national epidemiological study will help raise the aware- care episodes are initiated with hospital admission and ness of hip fractures among the elderly population and most prior studies have relied on hospitalized hip frac- support the practice of aggressive diagnosis and the tures to identify incident cases [14,17]. However, if we adoption of effective treatment options for osteoporosis. restrict the study sample to include hospitalized fractures Kang et al. BMC Public Health 2010, 10:230 Page 8 of 9 http://www.biomedcentral.com/1471-2458/10/230 16. Charlson ME, Pompei P, Ales KL, MacKenzie RC: A new method of List of abbreviations classifying prognostic comorbidity in longitudinal studies: CI: confidence interval; HI: Health Insurance; MA: Medi- development and validation. J Chronic Dis 1987, 40:373-383. cal Aid; NHI: National Health Insurance. 17. Rowe SM, Song EK, Kim JS, Lee JY, Park YB, Bae BH, Hur CI: Rising incidence of hip fracture in Gwangju City and Chonnam Province, Competing interests Korea. J Korean Med Sci 2005, 20:655-658. The authors declare that they have no competing interests. 18. US Census Bureau: Population Estimates. [http://www.census.gov/ popest/arvhives/1990s/nat_age_sex.html]. Accessed March 5, 2010 Authors' contributions 19. Richmond J, Aharonoff GB, Zuckerman JD, Koval KJ: Mortality risk after HYK developed the study design, interpretted the data, and prepared the man- hip fracture. Journal of Orthopaedic Trauma 2003, 17(1):53-56. uscript draft. KHY and SHM consulted clinical matters regarding osteoporosis, 20. Karagiannis A, Papakitsou E, Dretakis K, Galanos A, Megas P, Lambiris E, hip fractures, and their treatment, and helped to develop the implication of the Lyritis GP: Mortality rates of patients with a hip fracture in a study findings. YNK and WJC performed the statistical analysis. DRK partici- southwestern district of Greece: Ten-year follow-up with reference to pated in its design and provided statistical advice. SEP helped to draft the man- the type of fracture. Calcified Tissue International 2006, 78:72-77. uscript and performed the necessary reference searches. All authors have read 21. Farahmand BY, Michaelsson K, Ahlbom A, Ljunghall S, Baron JA: Survival and approved the final manuscript. after hip fracture. Osteoporos Int 2005, 16:1583-1590. 22. Ho CA, Li CY, Hsieh KS, Chen HF: Factors determining the 1-year survival Author Details after operated hip fracture: a hospital-based analysis. J Orthop Sci 2010, Graduate School of Public Health, Institute of Health Services Research, Yonsei 15:30-37. University, Seoul, South Korea, Department of Orthopedics, College of 23. Muraki S, Yamamoto S, Ishibashi H, Nakamura K: Factors associated with Medicine, Yonsei Universtiy, Seoul, South Korea, Graduate School of Public mortality following hip fracture in Japan. J Bone Miner Metab 2006, Health, Yonsei University, Seoul, South Korea and Health Insurance Review & 24(2):100-104. Assessment Service, Seoul, South Korea 24. McClung MR, Geusens P, Miller PD, Zippel H, Bensen WG, Roux C, et al.: Effect of risedronate on the risk of hip fracture in elderly women. New Received: 8 September 2009 Accepted: 4 May 2010 Eng J Med 2001, 344(5):333-340. Published: 4 May 2010 25. Harris ST, Watts NB, Genant HK, McKeever CD, Hangartner T, Keller M, et al.: T © T Bh h M 2 i is s C 0 arti P i1 s0 an u K b clic l a O e n i He p g se avai et n al a A th lcce lable ; 2010, lics esn f arti s ro 10 ee m cle B :230 : h i d ottp: Med istri //bu w Ce w te n w d tr .bi u aln Lt od m e d. e r th dce e te ntral.co rms o m f th /1e 471- Cre2458/ ative C 10/ om 230 mons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis. JAMA 1999, References 282(14):1344-1352. 1. Clark P, Lavielle P, Franco-Marina F, Ramírez E, Salmerón J, Kanis JA, 26. Roche JJW, Wenn RT, Sahota O, Moran CG: Effect of comorbitidies and Cummings SR: Incidence rates and life-time risk of hip fractures in postoperative complications on mortality after hip fracture in elderly Mexicans over 50 years of age: a population-based study. Osteoporos people: prospective observational cohort study. BMJ 2005, Int 2005, 16:2025-2030. 331(7529):1374. 2. Morosano M, Masoni A, Sanchez A: Incidence of hip fractures in the city 27. Wehren LE, Hawkes WG, Orwig DL, Hebel JR, Zimmerman SI, Magaziner J: of Rosario, Argentina. Osteoporos Int 2005, 16:1339-1344. Gender difference in mortality after hip fracture: the role of infection. J 3. Soveid M, Serati AR, Masoompoor M: Incidence of hip fracture in Shiraz, Bone Miner Res 2003, 18(12):2231-2237. Iran. Osteoporos Int 2005, 16:1412-1416. 28. Kiebzak GM, Beinart GA, Perser K, Ambrose CG, Siff SJ, Heggeness MH: 4. Hagino H, Katagiri H, Okano T, Yamamoto K, Teshima R: Increasing Undertreatment of osteoporosis in men with hip fracture. Arch Intern incidence of hip fracture in Tottori Prefecture, Japan: Trend from 1986 Med 2002, 162(19):2217-2222. to 2001. Osteoporos Int 2005, 16:1963-1968. 29. Chen E, Martin AD, Matthews KA: Understanding health disparities: the 5. Maghraoui AE, Koumba BA, Jroundi I, Achemlal L, Bezza A, Tazi MA: role of race and socioeconomic status in children's health. American Epidemiology of hip fractures in 2002 in Rabat, Morocco. Osteoporos Journal of Public Health 2006, 96(4):702-708. Int 2005, 16:597-602. 30. Sekine M, Chandola T, Martikainen P, Marmot M, Kagamimori S: 6. Lau EM, Lee JK, Suriwongpaisal P, Saw SM, Das De S, Khir A, Sambrook P: Socioeconomic inequalities in physical and mental functioning of The incidence of hip fracture in four Asian countries: the Asian Japanese civil servants: explanations from work and family Osteoporosis Study (AOS). Osteoporos Int 2001, 12:239-243. characteristics. Social Science & Medicine 2006, 63(2):430-445. 7. Ross PD, Huang C: Hip fracture incidence among Caucasians in Hawaii 31. Nurmi-Luthje I, Luthje P, Kaukonen JP, et al.: Post-fracture prescribed is similar to Japanese. A population-based study. Aging (Milano) 2000, calcium and vitamin D supplements alone or, in females, with 12:356-359. concomitant anti-osteoporotic drugs is associated with lower 8. Yan L, Zhou B, Prentice A, Wang X, Golden MHN: Epidemiological study mortality in elderly hip fracture patients: a prospective analysis. Drug of hip fracture in Shenyang, People's Republic of China. Bone 1999, Aging 2009, 26(5):409-421. 24:151-155. 32. Cree MW, Juby AG, Carriere KC: Mortality and morbidity associated with 9. Chie WC, Yang RS, Liu JP, Tsai KS: High incidence rate of hip fracture in osteoporosis drug treatment following hip fracture. Osteoporos Int Taiwan: estimated from a nationwide health insurance database. 2003, 14:722-727. Osteoporos Int 2004, 15:998-1002. 33. Feldstein AC, Nichols GA, Elmer PJ, Smith DH, Aickin M, Herson M: Older 10. Koh LKH, Saw SM, Lee JJM, et al.: Hip fracture incidence rates in women with fractures: patients falling through the cracks of guideline- Singapore 1991-1998. Osteoporos Int 2001, 12:311-318. recommended osteoporosis screening and treatment. Journal of Bone 11. Gullberg B, Johnell O, Kanis JA: World-wide projections for hip fracture. and Joint Surgery 2003, 85(12):2294-2302. Osteoporos Int 1997, 7:407-413. 34. American Association of Clinical Endocrinologists: AACE clinical practice 12. Zebaze RMD, Seeman E: Epidemiology of hip and wrist fracture in guidelines for the prevention and treatment of postmenopausal Cameroon, Africa. Osteoporos Int 2003, 14:301-305. osteoporosis. J Fla Med Assoc 1996, 83:552-566. 13. Korea National Statistical Office: National Estimated Population Data, 35. Luthje P, Nurmi-Luthje I, Kaukonen JP, Kuurne S, Naboulsi H, Kataja M: 1980, 2002. . Undertreatment of osteoporois following hip fracture in the elderly. 14. Maravic M, Le Bihan C, Landais P, Fardellone P: Incidence and cost of Arch Gerontol Geriatr 2009, 49(1):153-157. osteoporotic fractures in France during 2001: A methodological 36. Petrella RJ, Jones TJ: Do patients receive recommended treatment of approach by the national hospital database. Osteoporos Int 2005, osteoporosis following hip fracture in primary care? BMC Fam Pract 16:1475-1480. 2006, 7:31. 15. Murray SB, Bates DW, Ngo L, Ufberg JW, Shapiro NI: Charlson index is 37. Lo JC, Pressman AR, Omar MA, et al.: Persistence with weekly associated with one-year mortality in emergency department patients alendronate therapy among postmenopausal women. Osteopors Int with suspected infection. Academic Emergency Medicine 2006, 2006, 17(6):922-928. 13:530-536. Kang et al. BMC Public Health 2010, 10:230 Page 9 of 9 http://www.biomedcentral.com/1471-2458/10/230 38. Kim JY: Strategies to enhance the use of National Health Insurance claims database in generating health statistics. Technical report by Health Insurance Review and Assessment Services 2005. in Korean 39. Park BJ, Sung JH, Park KD, Seo SW, Kim SW: Strategies to improve the validity of diagnostic codes of National Health Insurance claims data. Technical report by Health Insurance Review and Assessment Services 2002. in Korean Pre-publication history The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1471-2458/10/230/prepub doi: 10.1186/1471-2458-10-230 Cite this article as: Kang et al., Incidence and mortality of hip fracture among the elderly population in South Korea: a population-based study using the National Health Insurance claims data BMC Public Health 2010, 10:230 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Public Health Springer Journals

Incidence and mortality of hip fracture among the elderly population in South Korea: a population-based study using the National Health Insurance claims data

Loading next page...
 
/lp/springer-journals/incidence-and-mortality-of-hip-fracture-among-the-elderly-population-MuL0o6iVqq

References (87)

Publisher
Springer Journals
Copyright
Copyright © 2010 by Kang et al; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Public Health; Medicine/Public Health, general; Epidemiology; Environmental Health; Biostatistics; Vaccine
eISSN
1471-2458
DOI
10.1186/1471-2458-10-230
pmid
20438644
Publisher site
See Article on Publisher Site

Abstract

Background: The lack of epidemiologic information on osteoporotic hip fractures hampers the development of preventive or curative measures against osteoporosis in South Korea. We conducted a population-based study to estimate the annual incidence of hip fractures. Also, we examined factors associated with post-fracture mortality among Korean elderly to evaluate the impact of osteoporosis on our society and to identify high-risk populations. Methods: The Korean National Health Insurance (NHI) claims database was used to identify the incidence of hip fractures, defined as patients having a claim record with a diagnosis of hip fracture and a hip fracture-related operation during 2003. The 6-month period prior to 2003 was set as a 'window period,' such that patients were defined as incident cases only if their first record of fracture was observed after the window period. Cox's proportional hazards model was used to investigate the relationship between survival time and baseline patient and provider characteristics available from the NHI data. Results: The age-standardized annual incidence rate of hip fractures requiring operation over 50 years of age was 146.38 per 100,000 women and 61.72 per 100,000 men, yielding a female to male ratio of 2.37. The 1-year mortality was 16.55%, which is 2.85 times higher than the mortality rate for the general population (5.8%) in this age group. The risk of post-fracture mortality at one year is significantly higher for males and for persons having lower socioeconomic status, living in places other than the capital city, not taking anti-osteoporosis pharmacologic therapy following fracture, or receiving fracture-associated operations from more advanced hospitals such as general or tertiary hospitals. Conclusion: This national epidemiological study will help raise awareness of osteoporotic hip fractures among the elderly population and hopefully motivate public health policy makers to develop effective national prevention strategies against osteoporosis to prevent hip fractures. Background society in terms of premature death, long-term disability, Osteoporosis has become a significant public health restricted social functioning, cost of care, lost productiv- problem in recent years, especially with the growth of the ity, and informal caregiver time. Previous studies have elderly population [1]. Hip fracture is the most valid epi- shown that the incidence rates of hip fractures for Asian demiologic marker of osteoporosis, a major cause of mor- people were lower than those for Whites living in North- bidity and mortality in the aging population [1-3], and ern Europe and North America [4-8]. However, recent imposes a considerable burden on patients, payers, and reports have shown an increase in the incidence of hip fractures in Asian populations [4,9,10], and it is projected that 45-70% of global hip fractures will occur in Asia by * Correspondence: hykang@yuhs.ac Graduate School of Public Health, Institute of Health Services Research, 2050 due to the growth of the elderly population [1,11]. Yonsei University, Seoul, South Korea Full list of author information is available at the end of the article © 2010 Kang et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in BioMed Central any medium, provided the original work is properly cited. Kang et al. BMC Public Health 2010, 10:230 Page 2 of 9 http://www.biomedcentral.com/1471-2458/10/230 To reduce the cost of fracture care, it is important to validity of the incidence cases identified from insurance adopt effective prevention strategies to reduce osteopo- claims data. rosis, including dietary supplements such as calcium and All claims records of outpatient visits or hospital vitamin D and pharmacologic therapy. However, the lack admissions of patients aged 50 or older containing a diag- of detailed data on hip fracture epidemiology, particularly nosis of femur fracture (fracture of femur [International in developing countries, hampers the development of Classification of Diseases (ICD)-10 diagnostic code: S72], preventive or curative measures against osteoporosis fracture of the neck of the femur [S72.0, S72.00], pertro- [12]. Thus, knowledge regarding the incidence of hip chanteric fracture [S72.1, S72.10]) and hip fracture- fractures and their outcomes, including mortality, is related operation (open reduction & internal fixation needed to evaluate the impact of osteoporosis on our [ICD-10 procedure code: N0601], closed reduction and society, to identify high-risk populations, and to help percutaneous fixation [N0991], total hip replacement health care decision-makers allocate scarce resources and [N0711], or hip hemiarthroplasty [N0715]) from January prioritize hip fractures along with other important clini- 1 to December 31, 2003 were identified from the NHI cal problems. claims database. The diagnosis and operation code for While the growth of the elderly population (>65 years hip fracture was selected based on previous epidemio- of age) in South Korea has been as marked as in other logic studies [1,14] and was confirmed by a panel of four Asian countries, rising from 3.8% in 1980 to 7.9% in 2002 orthopedic clinicians working in four different general [13], epidemiologic information about hip fractures hospitals in Korea. The cases having more than one claim among the elderly in South Korea is scarce. To provide a record that satisfied the inclusion criteria during 2003 national estimate on the incidence rate of hip fractures in were counted only once. this population, we need a result derived from a large, In general, not all operations are carried out during the population-based database for Korea. The Korean first fracture visit or admission. However, most of the National Health Insurance (NHI) claims database is such operations are performed within a month following the a database, containing all medical and prescription drug- first visit, according to a consultation of orthopedic clini- claim records for almost 100% of the Korean population. cians in Korea. Thus, we additionally defined the inci- South Korea operates a mandatory universal health dence cases as patients who did not have a record of a hip insurance system, with a centralized healthcare claims fracture-related operation in the claim record of the ini- database that provides a unique nationwide source of tial visit or admission, but had it within a month after the information on health care resource utilization. We used initial visit. this database to estimate the annual incidence and 1-year The 6-month period prior to 2003 (i.e., July - Decem- mortality of hip fractures in Koreans over the age of 50. ber, 2002) was set as a 'window period,' such that patients Furthermore, we assessed the patient and provider char- were defined as incident cases only if their first record of acteristics associated with post-fracture mortality in a fracture visit or admission was observed after this 6- order to help identify high-risk patients with poor health month period. Since most of the follow-up treatments for outcomes and develop effective treatment strategies after hip fracture are completed within 6 months after the ini- hip fracture incidence. The study results are expected to tial fracture, we assumed that the absence of any claims increase our understanding of the magnitude of the record with a diagnosis of hip fracture and hip-fracture- elderly population suffering from hip fractures and to related operation in the previous 6 months assured that urge public health policy makers to develop effective the fracture was a new case. national prevention strategies for osteoporosis. NHI claims data were merged with national mortality data provided by the National Statistical Office to deter- mine the survival status of individual patients at the 12 Methods Data sources and study population months following the incidence. Records from the Korean NHI claims database from 2002 Data analysis and 2004 were used to identify patients with hip fractures Age- and sex-specific annual incidence rates of hip frac- and to monitor their 1-year mortality. The incidence of tures per 100,000 inhabitants were calculated and pre- hip fractures was defined as patients having a claim sented in 5-year age intervals. The number of inhabitants record with a diagnosis of hip fracture and a hip fracture- in South Korea, as of 2003, was obtained from the Korean related operation. Since identifying hip fracture cases National Statistical Office. The age-standardized inci- solely based on diagnosis is likely to cause misclassifica- dence rate was computed using the Korea Census data of tion problems due to potential miscoding, we combined 2000. Age- and sex-specific 1-year mortality rates of hip the information from both diagnosis and surgical records. fractures were calculated and presented in 5-year age This conservative approach underestimates the real inci- intervals. To examine the difference in the incidence and dence rate of hip fractures in Korea, but improves the Kang et al. BMC Public Health 2010, 10:230 Page 3 of 9 http://www.biomedcentral.com/1471-2458/10/230 mortality of hip fractures according to gender, the gender affecting patient survival after hip fracture. As compared ratio of incidence and 1-year mortality rate was com- to the patients residing in Seoul, the capital city of South puted for each 5-year age stratum. Korea, those living in other parts of South Korea had a Cox's proportional hazards model was used to investi- significantly higher risk of death following fracture, gate the relationship between survival time and baseline which showed a hazard ratio ranging from 1.291 (95% patient and provider characteristics. The independent confidence interval (CI): 1.120-1.488) for those in metro- variables were age, gender, type of fracture, type of opera- politan cities to 1.656 (95% CI: 1.344-2.039) for those in tion, baseline health status, residence area, type of health non-city areas. insurance, provision of anti-osteoporosis medication To examine the impact of socioeconomic status on the therapy after fracture, and type of hospital providing mortality of hip fractures, we compared the mortality operation. between elderly enrolled in the two tiers of the Korean universal health insurance system: the Health Insurance (HI) and the Medical Aid (MA) program. The HI is a Results During the 1-year observation period from January wage-based, contributory insurance program covering through December 2003, there were 9,817 incidents of about 97% of the population, whereas the MA is a gov- hip fractures in persons aged 50 and older in South ernment-subsidized, public assistance program for the Korea: 6,892 females (70.2%) and 2,925 males (29.8%). poor and the medically indigent. The NHI claims data The mean (± SD) and the median age were 74.92 (± 9.13) include records for both the HI and the MA program and 76 years old, respectively. The mean age was younger populations. The survival analysis showed that patients in men (72.10 ± 9.47 years) than in women (76.11 ± 8.72 enrolled in the MA program had a 30% higher chance of years). The age-standardized annual incidence rate of hip death during the year following fracture than those in the fractures was 104.06 per 100,000; 146.38 per 100,000 HI program (hazard ratio: 1.300, 95% CI: 1.136-1.488). women and 61.72 per 100,000 men, yielding a female/ As a proxy measure for patients' baseline health status, male ratio of 2.37 (Table 1). a Charlson comorbidity index [15] was assigned to each The incidence rate of hip fractures increased with study subject based on the diagnosis shown in the NHI increasing age, from 6.74 per 100,000 for the youngest claims record during the study period prior to hip frac- group of 50-54 year olds to 686.82 per 100,000 for the ture. Charlson et al. developed a scoring system for oldest group of 85 years and older. This increase followed comorbid illness, such that each comorbid condition was an exponential relationship after age 65 for both genders assigned a whole number integer ranging from 1 to 6, (Figure 1). About 34.0% of all hip fractures occurred in which was proportional to the relative risk of 1-year mor- patients aged 80 years and older, although this age group tality associated with that disease [16]. Not surprisingly, comprised only 5.6% of the total population over 50. those who had had severe comorbid conditions before The gender ratio of the female to male incidence rate they had the fracture outbreak were 1.339 (for those with was computed for the 5-year age stratum. The ratio for Charlson Index of 1), 2.625 (Index of 2), or 5.693 (Index of the youngest 5-year group (50-54 years old) was smaller 3 or 6) times more likely to die as compared to those than 1, indicating a higher incidence rate among males without any of the Charlson Index conditions. than among females, probably due to occupation. How- The risk of mortality following fracture was not signifi- ever, this ratio increased with increasing age and was cantly different according to the type of hip fracture or greater than 1 for all groups older than or equal to 55 operation. The likelihood of dying among those with years old. claims records showing at least one reimbursed prescrip- Of 9,817 patients, 1,625 died during the first year post- tion for anti-osteoporosis drugs (alendronate, etidronate, fracture, yielding a mortality rate of 16.55% (Table 2). The raloxifen, risedronate, and salcatonin) following hip frac- mortality rate increased with increasing age from less ture was 42.4% lower than those without such claims than 10% of patients aged younger than 65 to more than records. 20% of those over 80 years old. Male patients consistently As the level and the size of the hospital in which the showed a higher mortality rate than female patients in all patients received their operation decreased, so did the age categories except the youngest, with the ratio of chance of death. The hazard ratios of 1-year mortality, female to male mortality ranging from 0.48 to 0.72, com- with tertiary hospitals as a reference group, were 0.798 pared with 3.77 in the youngest group. (95% CI: 0.680-0.937) for hospitals and 0.746 for clinics Patient and provider characteristics available from the (95% CI: 0.573-0.972). NHI claims data were examined for their association with 1-year post-fracture mortality using the multivariate Discussion Cox-proportional hazard model (Table 3). While other This is the first nation-wide, population-based, epidemi- factors were held constant in the multivariate model, gen- ologic study of hip fractures in Korea and one of the few der and age remained statistically significant factors studies worldwide reporting the incidence of hip frac- Kang et al. BMC Public Health 2010, 10:230 Page 4 of 9 http://www.biomedcentral.com/1471-2458/10/230 Table 1: Sex- and age-specific annual incidence rates of hip fractures in South Korea in 2003 Total Female Male Age No. No. hip fx. Incidence No. popl No. hip fx. Incid. Per No. popl No. hip fx. Incid. Per F/M ratio (years) population per 2003 100,000 2003 100,000 of incid. 2003 100,000 50-54 2,567,149 173 6.74 1,274,375 69 5.41 1,292,774 104 8.04 0.67 55-59 2,104,791 403 19.15 1,061,917 205 19.30 1,042,874 198 18.99 1.02 60-64 1,935,661 824 42.57 1,023,028 449 43.89 912,633 375 41.09 1.07 65-69 1,594,318 1,311 82.23 882,721 817 92.55 711,597 494 69.42 1.33 70-74 1,099,247 1,790 162.84 665,698 1,245 187.02 433,549 545 125.71 1.49 75-79 683,904 1,982 289.81 441,732 1470 332.78 242,172 512 211.42 1.57 80-84 380,112 1,888 496.70 259,291 1456 561.53 120,821 432 357.55 1.57 85+ 210,537 1,446 686.82 159,330 1181 741.23 51,207 265 517.51 1.43 Overall 10,575,719 9,817 92.83 5,768,092 6,892 119.48 4,807,627 2,925 60.84 1.96 (50+) Age-standardized rates* 104.06 146.38 61.72 2.37 *Age-standardized incidence rate was computed using Korea Census data of 2000. tures based on national health insurance claims data. The Singapore (442 and 164 per 100,000), and Thailand (269 age-standardized annual incidence rate for the entire and 114 per 100,000) [9]. population of Korea over 50 years of age in 2003 was esti- The risk of 1-year mortality following hip fracture was mated as 104.06 per 100,000 inhabitants, which is about 16.55%, which is about 2.85 times higher than the mortal- 27.8% lower than that estimated in an earlier study con- ity rate of the general population in South Korea over 50 ducted on inhabitants residing in selected regions of years of age. Despite the heterogeneity of the study popu- Korea in 2001 (133 per 100,000) [17]. The difference in lation in terms of age and ethnicity, and the potential dis- the estimated incidence rates between the two studies crepancy in the quality of follow-up care after the can be explained by the difference in the dataset, meth- fracture, the post-fracture mortality at 1 year observed in ods used in each study to identify incidence cases, and other countries is similar to that in Korea. A recent U.S. the population characteristics in the study regions. In the study has reported that about 11.5% of 950 patients over earlier study, hip fracture cases admitted to selected hos- age 65 admitted to selected hospitals with a hip fracture pitals in Gwangju City and Chonnam Province were iden- between 1987 and 1997 died after 1 year [19]. The 1-year tified on the basis of medical records and radiographs. mortality rate was 15.7% among 541 fracture patients Among the identified incidence cases in that study, 88% over 60 years of age admitted to one university hospital in underwent surgery. By applying this operation rate to the Greece between 1989 and 1992 [20]. In a population- incidence rate estimated by our study, the total number of based, case-control study in Sweden, about 10.6% of incidence cases with and without operation was esti- 2,245 women aged 50-81 years admitted with a hip frac- mated as 118.3 per 100,000 (= 104.6/0.88), thereby reduc- ture died within 1 year [21]. The comparison with other ing the gap in the incidence estimates between the two Asian countries also showed similar results. The 1-year studies from 27.8% to 12.4%. mortality of patients with surgically treated hip fractures After standardization according to the age distribution at a hospital in Taiwan between 1998 and 2006 was about of the US white population in 1990 [18], the incidence 14.0% [22]. A Japanese study by Muraki et al. reported rates of hip fractures in Korea were 173 per 100,000 in that the 1-year mortality rate following hip fracture was women and 91 per 100,000 in men. Interestingly, the 11.5% [23]. standardized incidence rates of both genders were sub- Osteoporosis is recognized as a problem predominantly stantially lower than those of most of the Asian countries in elderly women [24,25]. Consequently, women have including Taiwan (505 per 100,000 in women and 225 per higher incidence rates of hip fracture than men. In Korea, 100,000 in men), Hong Kong (459 and 180 per 100,000), the female to male ratio of the age-standardized inci- Kang et al. BMC Public Health 2010, 10:230 Page 5 of 9 http://www.biomedcentral.com/1471-2458/10/230 that observed in other Asian countries, such as Japan (209.0 to 780 per 100,000), Hong Kong (404 to 1,639), Singapore (611), Malaysia (320), and Thailand (227 to 727) [4,6]. The increased incidence with increasing age in both genders confirms the equal vulnerability of both Female Male genders to the aging process. Thus, preventive strategies are important for both genders. It is interesting to observe higher post-fracture mortal- ity in Korean males than in females. Gender differences in post-hip fracture mortality have been reported in ear- 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Age lier studies showing that men were about twice as likely as women to die during the first years after hip fracture Figure 1 Comparison of age-specific incidence rates of hip frac- ture according to gender. [26-28]. The observed difference between men and women could be explained by under treatment of osteo- porosis [28] and higher infection rates in men following dence rates of hip fractures for those aged 50 and older is hip fracture [26,27]. 2.37. This ratio is comparable to western countries such Low socioeconomic status among the elderly popula- as Sweden (2.2, 1987-1991), Norway (2.3, 1983-1984), tion seems to be associated with an increased risk of frac- Australia (2.7, 1989-1990), and in U.S. whites (2.7, 1983- ture-caused mortality. The adjusted hazard ratio of post- 1984), and in some Asian countries such as Thailand (2.3, fracture mortality was significantly higher among those 1997-1998), Hong Kong (2.5, 1997-1998), and Malaysia enrolled in the MA program than among those in the HI (2.4, 1997-1998) [3]. Hip fracture rates were higher in program. Patients with low socioeconomic status, in gen- women of all age groups except for those under 55 years. eral, have poorer health status and therefore tend to have This pattern is commonly observed in populations of poorer health outcomes from the same conditions as other countries, although the starting age at which compared to those with higher socioeconomic status women show a higher hip fracture rate varies from coun- [29,30]. Thus, more effective public health strategies to try to county: 60 in Taiwan and Japan [4,9], but mostly 55 treat osteoporosis and to prevent fracture incidence or 50 in other countries such as Mexico, Argentina, Iran, should be implemented that target the indigent elderly and Morocco [1-3,5]. population in Korea. Although women showed consistently higher incidence The regional variation in fracture-related mortality rates of hip fractures than men, the high incidence rates among Korean elderly is noteworthy. Those living in among older men is worthy of attention. After age 75, places other than the capital city consistently showed a Korean men experienced hip fractures at a rate of 211.42 higher risk of death during the first year after hip fracture, to 517.51 per 100,000 inhabitants. This rate is similar to with a minimum of 21.4% or a maximum of 65.6% addi- Table 2: Sex- and age-specific post-hip fracture mortality rate at 1 year Total Female Male Age No. hip No. Mortality No. hip No. Mortality No. hip No. Mortality F/M ratio of (years) Fx. 2003 mortality rate (%) Fx. 2003 mortality rate (%) Fx. 2003 mortality rate (%) mortality rate 50-54 173 14 8.09 69 10 14.49 104 4 3.85 3.77 55-59 403 21 5.21 205 7 3.41 198 14 7.07 0.48 60-64 824 76 9.22 449 32 7.13 375 44 11.73 0.61 65-69 1,311 150 11.44 817 78 9.55 494 72 14.57 0.66 70-74 1,790 216 12.07 1,245 125 10.04 545 91 16.70 0.60 75-79 1,982 347 17.51 1470 224 15.24 512 123 24.02 0.63 80-84 1,888 413 21.88 1456 293 20.12 432 120 27.78 0.72 85+ 1,446 388 26.83 1181 296 25.06 265 92 34.72 0.72 Overall 9,817 1,625 16.55 6,892 1,065 15.45 2,925 560 19.15 0.81 (50+) Incidence per 100,000 Kang et al. BMC Public Health 2010, 10:230 Page 6 of 9 http://www.biomedcentral.com/1471-2458/10/230 Table 3: Multivariate survival analysis: Patient and provider characteristics associated with 1-year mortality after hip fracture Variables N (%) Hazard ratio 95% confidence interval Gender Female 6,892 (70.2) 1.000 Male 2,925 (29.8) 1.400 (1.260-1.557) Age (years) 50-59 576 (5.9) 1.000 60-69 2,135 (21.7) 1.787 (1.250-2.553) 70-79 3,772 (38.4) 2.879 (2.041-4.062) 80-89 2,923 (29.8) 4.742 (3.361-6.691) ≥ 90 411 (4.2) 9.273 (6.387-13.463) Residence Seoul (Capital city) 2,388 (24.3) 1.000 Metropolitan city 2,584 (26.3) 1.291 (1.120-1.488) Small city 4,133 (42.1) 1.214 (1.058-1.392) Others (none-city area) 712 (7.3) 1.656 (1.344-2.039) Health National Health Insurance program 8,514 (86.7) 1.000 insurance National Medical Aid program 1,303 (13.3) 1.300 (1.136-1.488) Baseline 0 4,742 (48.3) 1.000 Charlson 1 4,038 (41.1) 1.339 (1.198-1.496) Index (before hip 2 938 (9.6) 2.625 (2.264-3.042) fracture) 3, 6 99 (1.0) 5.693 (4.249-7.627) Type of Neck fracture 5,435 (55.4) 1.000 hip fracture Petrochanteric fracture. 3,733 (38.0) 1.090 (0.966-1.230) Neck + Petrochanteric fracture. 649 (6.6) 1.031 (0.843-1.261) th Type of operation Total hip replacement (ICD 10 4,574 (46.6) 1.000 procedure code: N0711) Open reduction/internal fixation 4,592 (46.8) 0.963 (0.855-1.084) (N0601) Closed reduction/percutaneous 397 (4.0) 0.958 (0.720-1.274) fixation (N0991) Hip hemiarthroplasty (N0715) 216 (2.2) 0.793 (0.536-1.174) Total hip replacement & Open 38 (0.4) 1.213 (0.575-2.561) reduction (N0711 + N0601) Treatment* No 7,761 (79.1) 1.000 Yes 2,056 (20.9) 0.576 (0.499-0.666) Type of hospital providing Tertiary hospital 2,714 (27.6) 1.000 operation General hospital 4,649 (47.4) 0.900 (0.796-1.018) Hospital 1,901 (19.4) 0.798 (0.680-0.937) Clinic 553 (5.6) 0.746 (0.573-0.972) * Treatment was defined as whether the patient took at least one prescription of anti-osteoporosis medicine after hip fracture. Kang et al. BMC Public Health 2010, 10:230 Page 7 of 9 http://www.biomedcentral.com/1471-2458/10/230 tional chance of death. This discrepancy may be due to only, we will underestimate incidence rates by excluding the lack of access to health care providers or the poor less severe cases that are initiated with office visits. To quality of health care provided in non-city or small city overcome this problem, the present study defined inci- areas. Further investigation is necessary to figure out the dent cases either by initial admission or a visit containing reasons for this regional discrepancy and therefore to a diagnosis of hip fracture. However, while most hospital reduce this discrepancy. admissions coincide with an acute new episode of illness, From our analysis, patients receiving medication to it is difficult for outpatient visits to determine whether treat osteoporosis after fracture were 42.4% less likely to the case is a new episode or part of a post-episode course. die within the year following fracture. The association In addition, the accuracy of diagnostic codes among the between reduced mortality and post-fracture use of anti- claims data for office visits is lower than that of hospital osteoporosis drugs in elderly hip fracture patients has admission [38,39]. Thus, to improve the accuracy of iden- been addressed in earlier studies. A prospective analysis tifying incidence cases based on initial fracture visits, we from Finland has reported a 43% reduction in deaths at excluded patients with a claims record with a hip fracture 36 months following hip fracture in females who used during the 6 months prior to the initial visit. Further- prescribed calcium plus vitamin D supplementation con- more, to confirm the hip fracture cases, hip fracture- comitantly with anti-osteoporosis drugs [31]. Also, an related operation codes were used in addition to the diag- observational study from Canada has shown that mortal- nostic codes. ity is significantly lower in the group treated with anti- Several methodological issues arise from the studies osteoporosis drugs than in the untreated group following using administrative claims data. First, not all patients hip fracture [32]. Because elderly people with prior frac- with fractures have access to hospitals. The incidence tures are at a higher risk for future fractures [33], clinical based on insurance claims records would be underesti- practice guidelines recommend the initiation of pharma- mated if many hip fractures in Korea were not diagnosed cologic treatment of osteoporosis after the first fracture or treated in health care institutions. However, due to the [34]. However, under-treatment of osteoporosis [35,36] emergent character of hip fractures, it is believed that vir- and low adherence to oral anti-osteoporosis treatment tually all hip fractures present to health services [3]. following hip fracture [37] is commonly observed. In our Thus, the incidence of hip fractures identified from study, only 20.9% of the fractured patients received anti- claims records is considered to be close to the actual inci- osteoporosis drugs following hip fracture. Thus, a more dence. aggressive practice of diagnosing and providing medica- Second, the reliance on ICD-10 diagnostic codes to tion therapy for osteoporosis seems to be an urgent need identify incident fractures may cause misclassification of in Korea. incident hip fractures due to the inherent nature of claims Interestingly, patients that underwent an operation for data such as voluntary or non-voluntary miscoding hip fracture from hospitals or clinics tend to have a better behavior. However, a recent validation study for the diag- chance of survival than those from tertiary hospitals. nostic codes of the NHI claims database in Korea has alle- Although our multivariate model tried to include as many viated this concern. It has been revealed that about 70% patient characteristic variables as possible from the NHI of primary, secondary, or tertiary diagnosis codes from data, we could not comprehensively control for hip frac- NHI claims records coincide with those from medical ture severity and patient health conditions due to the lack records. In addition, the accuracy of diagnosis codes of clinical information available from the claims records. tended to be higher for claims from hospital admissions Thus, it cannot be concluded that the quality of care and compared with office visits, and for claims of severe con- patient outcomes after hip fracture treatment are better ditions compared with mild conditions [38,39]. among hospitals and clinics than among tertiary hospitals in Korea. It might be possible that patients with more Conclusion severe conditions underwent hip fracture repairs from In conclusion, our results indicate a high incidence and larger and more advanced hospitals (i.e., tertiary or gen- mortality rate of hip fractures for both men and women eral hospitals) and therefore the prognosis after surgery living in South Korea. These high incidence rates for both tended to be poorer. A more thorough adjustment for genders are sentinel signals of the impact of hip fractures patient baseline disease severity is necessary to make a in Korea, in terms of high mortality, morbidity, and concrete conclusion about this finding. health care costs for treatment. It is our hope that this Since a hip fracture is a severe condition, most of the national epidemiological study will help raise the aware- care episodes are initiated with hospital admission and ness of hip fractures among the elderly population and most prior studies have relied on hospitalized hip frac- support the practice of aggressive diagnosis and the tures to identify incident cases [14,17]. However, if we adoption of effective treatment options for osteoporosis. restrict the study sample to include hospitalized fractures Kang et al. BMC Public Health 2010, 10:230 Page 8 of 9 http://www.biomedcentral.com/1471-2458/10/230 16. Charlson ME, Pompei P, Ales KL, MacKenzie RC: A new method of List of abbreviations classifying prognostic comorbidity in longitudinal studies: CI: confidence interval; HI: Health Insurance; MA: Medi- development and validation. J Chronic Dis 1987, 40:373-383. cal Aid; NHI: National Health Insurance. 17. Rowe SM, Song EK, Kim JS, Lee JY, Park YB, Bae BH, Hur CI: Rising incidence of hip fracture in Gwangju City and Chonnam Province, Competing interests Korea. J Korean Med Sci 2005, 20:655-658. The authors declare that they have no competing interests. 18. US Census Bureau: Population Estimates. [http://www.census.gov/ popest/arvhives/1990s/nat_age_sex.html]. Accessed March 5, 2010 Authors' contributions 19. Richmond J, Aharonoff GB, Zuckerman JD, Koval KJ: Mortality risk after HYK developed the study design, interpretted the data, and prepared the man- hip fracture. Journal of Orthopaedic Trauma 2003, 17(1):53-56. uscript draft. KHY and SHM consulted clinical matters regarding osteoporosis, 20. Karagiannis A, Papakitsou E, Dretakis K, Galanos A, Megas P, Lambiris E, hip fractures, and their treatment, and helped to develop the implication of the Lyritis GP: Mortality rates of patients with a hip fracture in a study findings. YNK and WJC performed the statistical analysis. DRK partici- southwestern district of Greece: Ten-year follow-up with reference to pated in its design and provided statistical advice. SEP helped to draft the man- the type of fracture. Calcified Tissue International 2006, 78:72-77. uscript and performed the necessary reference searches. All authors have read 21. Farahmand BY, Michaelsson K, Ahlbom A, Ljunghall S, Baron JA: Survival and approved the final manuscript. after hip fracture. Osteoporos Int 2005, 16:1583-1590. 22. Ho CA, Li CY, Hsieh KS, Chen HF: Factors determining the 1-year survival Author Details after operated hip fracture: a hospital-based analysis. J Orthop Sci 2010, Graduate School of Public Health, Institute of Health Services Research, Yonsei 15:30-37. University, Seoul, South Korea, Department of Orthopedics, College of 23. Muraki S, Yamamoto S, Ishibashi H, Nakamura K: Factors associated with Medicine, Yonsei Universtiy, Seoul, South Korea, Graduate School of Public mortality following hip fracture in Japan. J Bone Miner Metab 2006, Health, Yonsei University, Seoul, South Korea and Health Insurance Review & 24(2):100-104. Assessment Service, Seoul, South Korea 24. McClung MR, Geusens P, Miller PD, Zippel H, Bensen WG, Roux C, et al.: Effect of risedronate on the risk of hip fracture in elderly women. New Received: 8 September 2009 Accepted: 4 May 2010 Eng J Med 2001, 344(5):333-340. Published: 4 May 2010 25. Harris ST, Watts NB, Genant HK, McKeever CD, Hangartner T, Keller M, et al.: T © T Bh h M 2 i is s C 0 arti P i1 s0 an u K b clic l a O e n i He p g se avai et n al a A th lcce lable ; 2010, lics esn f arti s ro 10 ee m cle B :230 : h i d ottp: Med istri //bu w Ce w te n w d tr .bi u aln Lt od m e d. e r th dce e te ntral.co rms o m f th /1e 471- Cre2458/ ative C 10/ om 230 mons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis. JAMA 1999, References 282(14):1344-1352. 1. Clark P, Lavielle P, Franco-Marina F, Ramírez E, Salmerón J, Kanis JA, 26. Roche JJW, Wenn RT, Sahota O, Moran CG: Effect of comorbitidies and Cummings SR: Incidence rates and life-time risk of hip fractures in postoperative complications on mortality after hip fracture in elderly Mexicans over 50 years of age: a population-based study. Osteoporos people: prospective observational cohort study. BMJ 2005, Int 2005, 16:2025-2030. 331(7529):1374. 2. Morosano M, Masoni A, Sanchez A: Incidence of hip fractures in the city 27. Wehren LE, Hawkes WG, Orwig DL, Hebel JR, Zimmerman SI, Magaziner J: of Rosario, Argentina. Osteoporos Int 2005, 16:1339-1344. Gender difference in mortality after hip fracture: the role of infection. J 3. Soveid M, Serati AR, Masoompoor M: Incidence of hip fracture in Shiraz, Bone Miner Res 2003, 18(12):2231-2237. Iran. Osteoporos Int 2005, 16:1412-1416. 28. Kiebzak GM, Beinart GA, Perser K, Ambrose CG, Siff SJ, Heggeness MH: 4. Hagino H, Katagiri H, Okano T, Yamamoto K, Teshima R: Increasing Undertreatment of osteoporosis in men with hip fracture. Arch Intern incidence of hip fracture in Tottori Prefecture, Japan: Trend from 1986 Med 2002, 162(19):2217-2222. to 2001. Osteoporos Int 2005, 16:1963-1968. 29. Chen E, Martin AD, Matthews KA: Understanding health disparities: the 5. Maghraoui AE, Koumba BA, Jroundi I, Achemlal L, Bezza A, Tazi MA: role of race and socioeconomic status in children's health. American Epidemiology of hip fractures in 2002 in Rabat, Morocco. Osteoporos Journal of Public Health 2006, 96(4):702-708. Int 2005, 16:597-602. 30. Sekine M, Chandola T, Martikainen P, Marmot M, Kagamimori S: 6. Lau EM, Lee JK, Suriwongpaisal P, Saw SM, Das De S, Khir A, Sambrook P: Socioeconomic inequalities in physical and mental functioning of The incidence of hip fracture in four Asian countries: the Asian Japanese civil servants: explanations from work and family Osteoporosis Study (AOS). Osteoporos Int 2001, 12:239-243. characteristics. Social Science & Medicine 2006, 63(2):430-445. 7. Ross PD, Huang C: Hip fracture incidence among Caucasians in Hawaii 31. Nurmi-Luthje I, Luthje P, Kaukonen JP, et al.: Post-fracture prescribed is similar to Japanese. A population-based study. Aging (Milano) 2000, calcium and vitamin D supplements alone or, in females, with 12:356-359. concomitant anti-osteoporotic drugs is associated with lower 8. Yan L, Zhou B, Prentice A, Wang X, Golden MHN: Epidemiological study mortality in elderly hip fracture patients: a prospective analysis. Drug of hip fracture in Shenyang, People's Republic of China. Bone 1999, Aging 2009, 26(5):409-421. 24:151-155. 32. Cree MW, Juby AG, Carriere KC: Mortality and morbidity associated with 9. Chie WC, Yang RS, Liu JP, Tsai KS: High incidence rate of hip fracture in osteoporosis drug treatment following hip fracture. Osteoporos Int Taiwan: estimated from a nationwide health insurance database. 2003, 14:722-727. Osteoporos Int 2004, 15:998-1002. 33. Feldstein AC, Nichols GA, Elmer PJ, Smith DH, Aickin M, Herson M: Older 10. Koh LKH, Saw SM, Lee JJM, et al.: Hip fracture incidence rates in women with fractures: patients falling through the cracks of guideline- Singapore 1991-1998. Osteoporos Int 2001, 12:311-318. recommended osteoporosis screening and treatment. Journal of Bone 11. Gullberg B, Johnell O, Kanis JA: World-wide projections for hip fracture. and Joint Surgery 2003, 85(12):2294-2302. Osteoporos Int 1997, 7:407-413. 34. American Association of Clinical Endocrinologists: AACE clinical practice 12. Zebaze RMD, Seeman E: Epidemiology of hip and wrist fracture in guidelines for the prevention and treatment of postmenopausal Cameroon, Africa. Osteoporos Int 2003, 14:301-305. osteoporosis. J Fla Med Assoc 1996, 83:552-566. 13. Korea National Statistical Office: National Estimated Population Data, 35. Luthje P, Nurmi-Luthje I, Kaukonen JP, Kuurne S, Naboulsi H, Kataja M: 1980, 2002. . Undertreatment of osteoporois following hip fracture in the elderly. 14. Maravic M, Le Bihan C, Landais P, Fardellone P: Incidence and cost of Arch Gerontol Geriatr 2009, 49(1):153-157. osteoporotic fractures in France during 2001: A methodological 36. Petrella RJ, Jones TJ: Do patients receive recommended treatment of approach by the national hospital database. Osteoporos Int 2005, osteoporosis following hip fracture in primary care? BMC Fam Pract 16:1475-1480. 2006, 7:31. 15. Murray SB, Bates DW, Ngo L, Ufberg JW, Shapiro NI: Charlson index is 37. Lo JC, Pressman AR, Omar MA, et al.: Persistence with weekly associated with one-year mortality in emergency department patients alendronate therapy among postmenopausal women. Osteopors Int with suspected infection. Academic Emergency Medicine 2006, 2006, 17(6):922-928. 13:530-536. Kang et al. BMC Public Health 2010, 10:230 Page 9 of 9 http://www.biomedcentral.com/1471-2458/10/230 38. Kim JY: Strategies to enhance the use of National Health Insurance claims database in generating health statistics. Technical report by Health Insurance Review and Assessment Services 2005. in Korean 39. Park BJ, Sung JH, Park KD, Seo SW, Kim SW: Strategies to improve the validity of diagnostic codes of National Health Insurance claims data. Technical report by Health Insurance Review and Assessment Services 2002. in Korean Pre-publication history The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1471-2458/10/230/prepub doi: 10.1186/1471-2458-10-230 Cite this article as: Kang et al., Incidence and mortality of hip fracture among the elderly population in South Korea: a population-based study using the National Health Insurance claims data BMC Public Health 2010, 10:230

Journal

BMC Public HealthSpringer Journals

Published: May 4, 2010

There are no references for this article.