Considering aging society, osteoporosis and its related fractures have become a growing health problem worldwide.[12] Osteoporosis occasionally results in osteoporotic fracture in hip, spine, humerus, and wrist.[345] In Korea, the annual incidences of osteoporotic fractures were 1,614 per 100,000 person-years in people aged 50 years or more in 2008.[67]

It is obvious that securing an appropriate level of post-fracture management for patients with osteoporosis may significantly reduce the risk of osteoporotic fracture. Especially, patients with previous osteoporotic fracture have higher risk of a subsequent fracture than those without previous fracture.[8]

Thus, post-fracture management for osteoporosis is highly recommended to prevent the occurrence of new fragility fractures.[91011]

However, even high-risk patients with previous fracture often do not receive preventive management worldwidely. [812131415] Korea is not an exception. Only 52.2% were aware of their diagnosis and 58.4% received pharmacological treatment among those with osteoporotic fractures.[16]

Bone mineral density (BMD) measurement is the first important step to investigate and manage patients with osteoporosis.[17] That is the important opportunity to initiate secondary prevention in patients with previous osteoporotic fracture.[18] In fact, a few empirical studies have already dealt with this issue in Korea.

There was lack of studies on the rate of BMD measurement after osteoporotic fracture in Korea, and what factors are associated with BMD measurement after fractures.

Therefore, our purposes were 1) to evaluate the rate of BMD measurement after osteoporotic fracture in Korean population; and 2) to determine associated factors with BMD measurement after fractures in Korea.

We used data from the nationwide claims database of Health Insurance Review & Assessment Service (HIRA). Almost 97% of the Korean populations have been currently covered with this national insurance system. In other words, the medical claims data include demographic information (age and gender), diagnoses using the International Classification of Diseases, Tenth Revision (ICD-10) codes and procedures for diagnosis and treatment using codes in both of inpatients and outpatients care. Thus, it is firmly certain that all information about health care utilization is available from the HIRA database. Several epidemiologic studies have used this national claim database.[192021] We analyzed patients aged over 50 years who were diagnosed with osteoporotic fracture by physician at 2010.

We identified patients with hip, spine, humerus and wrist fractures diagnosed in 2010. To identify patients with these fractures, we adopted the diagnostic codes using the ICD-10 (hip, S720 and S721; spine, M484, M485, S220, S221, and S320; humerus, S422 and S423; wrist fractures, S525 and S526) and the procedure codes according to each anatomic site.[372223]

If an individual with fracture had more than one outpatient visits or admissions within the time period of six months, the cases were not counted separately, as below.[2425]

Double recording was avoided by counting only one record in the case that a person had more than one record in the HIRA database. If a patient had both spine and wrist fractures, only the first episode was counted.

The data based on the HIRA came from the patients who had experienced a hip, spine, humerus or wrist fracture and had undergone BMD examinations within 6 months before and after osteoporotic fractures. The procedure codes (HC 341-HC 344) for these examinations included dual x-ray absorptiometry scans (single site, HC 341; multiple sites, HC 342), quantitative computed tomography scans (HC 343), and other methods, including ultrasound (HC 344).

The rates of BMD examinations were estimated within 6 months after osteoporotic fractures.

We evaluated gender, age group (10-year incremental), type of insurance, residency area (rural vs. urban), type of medical institute, department, history of depression, rheumatoid arthritis, medical history suggestive of secondary osteoporosis, osteoporosis-induced drug, and number of family members as potential associated factors.

The significance of differences was determined with use of a χ² test. Statistical analyses were performed using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA).

We analyzed a total of 192,556 patients who were diagnosed of osteoporotic fracture in 2010. Of these patients, only 103,785 (53.9%) had been measured with BMD within 6-months post fracture (Table 1).

Indeed, the number of female showed higher frequency to have BMD measurement as compared with men (57.9% vs. 37.9%). And, older patients were more likely to have BMD measurement, especially in population aged from 70 to 79 (66.7%). Patients in medical benefit system indicating almost free-ride medical aid program in Korea tended to have BMD measurement more frequently than those with medical care insurance (61.7% vs. 53.1%). Patients living in urban area were less likely to have BMD measurement than those living in rural area (51.6% vs. 97.6%). Patients who utilized public health care center showed highest likelihood (80.3%) of being taken BMD measurement than those who utilized another type of medical institute. When patients visited at neurosurgery department after fracture, BMD was measured in the most proportion (79.3%). As compared with those who had no medications during the baseline period, patients who took glucocorticoids, were taken BMD measurement more frequently (53.4% vs. 71.9%). Patients who had depression history were less likely to have BMD measurement than those with no depression history (31.2% vs. 54.0%). Patients who had rheumatoid arthritis were more likely to have BMD measurement than those with no rheumatoid arthritis (70.3% vs. 53.5%). Patients who had medical history suggestive of secondary osteoporosis were more likely to have BMD measurement than those with no medical history osteoporosis (62.0% vs. 52.4%). Patients who took osteoporosis-induced drug were more likely to have BMD measurement than those with no medical history osteoporosis (59.3% vs. 43.2%). Patients who had family members of more than four were more likely to have BMD measurement than those with less than four (58.4% vs. 50.7%).

This study examined BMD measurement in patients after the first fracture in Korea by using nationwide retrospective data. We found that only 37.9% of men and 57.9% of women underwent BMD measurement within 6 months after a fracture.

The low rate of BMD measurement was associated with men, younger age, urban residency, and depression history, in terms of patients' factor. Our results show that postfracture management including BMD measurement is particularly poor among younger men, which was evidently in the line with the previous studies.[2627]

In addition, physician in orthopedic department made the lowest level of BMD measurement after fracture (45.7%). Given that the most of fractures have been treated in orthopedic department, it is important that orthopedic surgeon have awareness of necessity of BMD measurement. [4] In our analysis, patients who utilized larger hospital showed less likelihood of being taken BMD measurement than those who utilized another smaller medical institute, which is likely to be related to the possibility of more coordinated care between the different departments in small hospital than in large hospital. Our findings indicate that more coordinated service model is required in Korea to reduce the care gap for secondary fracture prevention.

According to our results, many of the factors affected the BMD measurement significantly. Increasing age, use of glucocorticoid use, osteoporosis-inducing comorbid disease including rheumatologic disease, and osteoporosisinduced drug user were associated with higher likelihood of BMD measurement, which is in line with prior studies. [828] Patients who utilized public health care center showed highest likelihood (80.3%) of being taken BMD measurement than those who utilized another type of medical institute. This means that public health care center play an important role to manage osteoporosis in Korea.

There were several limitations. First, we assumed that all patients with osteoporotic-fracture were eligible for BMD measurement. Second, we could not include patients who had BMD measurement after fracture, when patients with fracture took routine health care examination, because it was not included claim database. Therefore, there is possibility for us to under estimate the real gap in the post-fracture BMD measurement. Third, we could not evaluate whether the patients with BMD measurement took anti-osteoporosis treatment after BMD measurement. Finally, we could not perform multivariable analysis, because we could not link age, gender, disease or procedure code, type of insurance and so on using joint key of each individual.

Given the drawbacks raised above, our findings indicate that lack of BMD measurement after fracture remains a problem in Korea, especially among young men with osteoporotic fracture. Besides, this study provided the basic information to optimize management after fracture, such as giving education patient and physicians about the importance of post-fracture management.