Bisphosphonate use and increased incidence of subtrochanteric fracture in South Korea: results from the National Claim Registry
- First Online:
- Cite this article as:
- Lee, Y., Ha, Y., Park, C. et al. Osteoporos Int (2013) 24: 707. doi:10.1007/s00198-012-2016-8
- 567 Views
We evaluated trends in the incidences of typical and atypical hip fracture in relation to bisphosphonate use in Korea from 2006 to 2010, using nationwide data obtained from the Health Insurance Review and Assessment Service (HIRA).
Recently, atypical hip fractures in the subtrochanteric region have been reported among patients on bisphosphonate. However, the association between atypical hip fracture and bisphosphonate is controversial. We evaluated trends in the incidences of typical and atypical hip fracture in relation to bisphosphonate use in Korea from 2006 to 2010, using nationwide data obtained from the HIRA.
All new visits or admissions to clinics or hospitals for a typical and atypical hip fractures were recorded nationwide by HIRA using the ICD-10 code classification. Typical and atypical hip fractures were defined as femoral neck/intertrochanteric and subtrochanteric fracture, respectively. Bisphosphonate prescription data were also abstracted from the HIRA database.
The absolute number of typical and atypical hip fracture increased during the study period. Although age-adjusted incidence rates of typical hip fractures were stable in men and women, those of atypical hip fractures increased in women. Nationally, the annual numbers of prescriptions of bisphosphonate also increased during the study period.
The results of this study suggest a possible causal relationship between bisphosphonate use and the increased incidence of atypical hip fracture in Korea.
KeywordsBisphosphonateEpidemiologic studyHip fractureNationwide databaseSubtrochanteric fracture
Bisphosphonate is prescribed worldwide to prevent and treat osteoporosis, and its efficacy at reducing the risk of osteoporotic fractures has been proved and its safety demonstrated [1, 2]. But atypical hip fractures in the subtrochanteric region have recently been associated with the long-term use of bisphosphonate in the West [3–6], and in 2010, the Food and Drug Administration released a safety statement about the risk of atypical hip fracture in those on long-term bisphosphonate. However, the relation between bisphosphonate use and the risk of atypical hip fracture remains controversial [7–9]. In view of the low incidence of atypical hip fractures in the subtrochanteric region, an epidemiologic study based on nationwide database appears to provide the best means of securing sufficient statistical power, and one such study has already been conducted in the West .
Population aging is also a feature in East Asia, and is reflected by increases in the incidences of osteoporosis and osteoporotic fractures [11, 12]. Accordingly, the use of bisphosphonate to prevent and treat osteoporosis is increasing in East Asia . However, no epidemiologic study has been conducted on the association between the incidence of atypical hip fracture and bisphosphonate usage in East Asia. The purpose of this study was to evaluate trends in incidence of typical and atypical hip fractures and compare these with bisphosphonate use in the Korean population from 2006 to 2010, using data from the Health Insurance Review and Assessment Service (HIRA).
Materials and methods
We analyzed data from nationwide claims database of the HIRA of Korea between the years 2006 and 2010. In Korea, 97.0 % of the population is obliged to enroll in the Korean National Health Insurance Program. Patients pay around 30 % of total medical costs to clinics or hospitals, though some services are not covered by insurance, such as cosmetic surgery and some unproven therapies. Clinics and hospitals then submit claims for inpatient and outpatient care, including data on diagnoses [as determined by the International Classification of Diseases, 10th revision (ICD-10)], procedures, prescription records, demographic information, and direct medical costs, to obtain reimbursement for 70 % of total medical costs. The remaining 3 % of the population not insured by the Korean National Health Insurance Program are either covered by another Medical Aid Program or are temporary or illegal residents. These claims are also reviewed by HIRA, and thus, virtually all information about patients and their medical records is available from the Korean HIRA database, which has been used on several occasions for epidemiological studies [13–18].
We selected data from 2006 to 2010 to evaluate trends in the incidences of typical and atypical hip fractures. Hip fractures require hospitalizations and surgical intervention, and are recorded prospectively nationwide using ICD-10 codes, as described above. To identify hip fractures, selected ICD-10 codes and a minimum cutoff value of 50 years were used .
Typical hip fracture cases were defined as those requiring hospitalization with a primary diagnosis of femoral neck fractures (ICD-10S720) or intertrochanteric fracture (ICD-10 S721), and atypical hip fractures as those requiring hospitalization with a primary diagnosis of a subtrochanteric fracture (ICD-10 S722). We limited the cohort to those patients that underwent one of the following seven procedures: open reduction of fractured extremity–femur, closed pinning–femur, external fixation–pelvis/femur, closed reduction of fractured extremity–pelvis/femur, bone traction, skin traction, or hemiarthroplasty–hip.
Annual number of prescriptions of bisphosphonate was abstracted and collated. Because the data used are available to the public, only deidentified information was used in the present study. The study protocol was approved by the HIRA institutional review boards.
To determine trends in fracture incidences, the patients were categorized by age (subdivided into 5-year increments), gender, and fracture site (typical and atypical hip fracture). Age-adjusted and gender-specific incidence rates (per 100,000 person-years) were calculated using respective annual population. Numbers of men and women over 50 were obtained from the “Statistics Korea” web site (http://www.kosis.kr), the official web site of the Central Government Organization for Statistics . Annual Percentage Change for trends was used to determine whether fracture incidences changed from 2006 to 2010. (Joinpoint Regression Program, Version 3.5.2, Statistical Research and Applications Branch, National Cancer Institute, Bethesda, USA). Age-adjusted and gender-specific incidences from 2006 to 2010 were used in this analysis. To evaluate bisphosphonate use, annual numbers of prescriptions of bisphosphonate were evaluated. P values of less than 0.05 were considered significant.
Age-adjusted and gender-specific incidence (per 100,000 persons) of typical and atypical hip fractures in Korean over 50 years old
Typical hip fracture
Atypical hip fracture
Number of fracture
Incidence of fracture
Age-adjusted incidence of fracturea
Number of fracture
Incidence of fracture
Age-adjusted incidence of fracturea
Annual Percentage Change (95 % confidence interval) in age-adjusted and gender-specific incidence of typical and atypical hip fractures from 2006 to 2010
Typical hip fracture
–0.2 (–4.1 to 4.0)
1.1 (–3.6 to 6.1)
0.7 (–3.7 to 5.3)
Atypical hip fracture
–2.2 (–4.6 to 0.2)
4.1 (0.5 to 7.9)a
1.7 (–1.6 to 5.0)
This nationwide claim database study of the population aged 50 years or older demonstrates that incidence trends of typical and atypical hip fractures during the period 2006 to 2010 differ. Although the incidence of typical hip fracture was stable in women and men, increasing trends in absolute numbers and incidence rates were found for atypical hip fractures in women.
This finding concurs with the results of a recent national epidemiologic study conducted in the USA , in which it was demonstrated that bisphosphonate use and atypical hip fracture incidence significantly increased in postmenopausal women over 12 years, whereas typical hip fracture incidence declined. Although bisphosphonate was introduced in South Korea in 1988, the present study was conducted over 5 years because HIRA limited the study period to a maximum of 5 years. However, this is the first epidemiologic study to be conducted on the incidence of atypical hip fracture and bisphosphonate use in East Asia.
In the West, atypical hip fracture is a topic of concern, and it appears that this is also likely to occur in Asia because the prevalence of bisphosphonate is increasing due to societal aging. Actually, in a previous study using pooled data, it was reported that 46 % of all patients with an atypical hip fracture were Asian . However, some studies have found that no significant association exists between the incidence of atypical hip fractures and bisphosphonate use [7, 8]. Although no causal relation between atypical hip fracture and use of bisphosphonate can be demonstrated by epidemiologic study, our findings suggest that a potential relationship exists between observed increases in incidence of atypical hip fracture and bisphosphonate use.
The present study has the benefits of a large sample size and nationwide sampling, but it also has some limitations that warrant consideration. First, details of the characteristic radiographic findings of atypical hip fracture, which were described by task force of the American Society of Bone and Mineral Research , were not available because study design was based on National Claim Registry. Therefore, it was possible that fractures due to high-energy trauma were included in this study because distinction between high- and low-energy fractures could not be made by the ICD-10 coding system. However, the ICD-10 coding system, established in 1989 by WHO, included greater detail, standardized terminology, and expanded concepts for injuries, and other related factors [22, 23], and, therefore, it is more useful than previous version in an epidemiologic study using code. In addition, we used additional criteria of “aged 50 years or more” to minimize this bias. Second, because we used deidentified codes and the numbers of prescriptions, we could not determine the amounts of bisphosphonate prescribed at an individual level with atypical hip fracture, and did not consider compliance and persistence, although bisphosphonate treatment duration is likely to be associated with atypical hip fracture. However, the increased numbers of prescriptions of bisphosphonate concur with a previous study on burden of osteoporosis in Korea using the same HIRA database . Third, not all patients that experienced an atypical hip fracture were registered in the nationwide database because the National Claim Registry contains information on patients treated at a clinic or hospital. Furthermore, patients with asymptomatic insufficiency fracture in the subtrochanteric region may not have been included, and National Claim Registry-based incidences are likely to be somewhat underestimated. Fourth, we did not include patients with a femoral shaft fracture, although this type of fracture in elderly population could also be viewed as an atypical hip fracture [4, 8, 24]. This was done because we wanted to focus on subtrochanteric hip fractures, and we considered that femoral shaft fractures represent a different type in this regard. Fifth, our data on bisphosphonate may have included patients with a metabolic bone disorder other than osteoporosis. Finally, the incidences of typical hip fractures found in the present study are higher than those reported previously in HIRA database-based studies, although trends in incidence are consistent between studies [15, 16]. We believed that this was due to our use of deidentified data and the non-exclusion of duplication, whereas previous studies used protocols that avoided case duplication [15, 16].
Despite these limitations, the results of this study suggest a possible causal relationship between bisphosphonate use and the incidence of atypical hip fracture in a national Korean cohort. We suggest, because of increasing concerns and lack of knowledge regarding the pathophysiology of atypical hip fracture, that a well-designed study be conducted.
This study was supported by grant #11-2011-004 from the Seoul National University Bundang Hospital Research Fund. The authors wish to thank En-Sook Hwang for data collection.
Conflicts of interest