Table 1.
Performance of OST among Asians.
| Study | Objective | Subject Description | Number of Subjects Recruited | Methods | Cutoff | Sensitivity (%) | Specificity (%) | AUC | Remarks |
|---|---|---|---|---|---|---|---|---|---|
| Koh et al. (2001) [11] |
To develop Osteoporosis Screening Tool for Asians (OSTA) | Postmenopausal women (mean age 62 years) recruited from 21 clinics in eight Asia countries. | 860 | DXA: 8 Hologic machines (3 Model 4500, 5 Model 2000), 4 Norland (2 XR-36, 1 S-26, 1 XR-26), 12 Lunar (3 DPX-IQ. 6 DPX-L, 3 Expert) machines |
OSTA < −1 T-score< −2.5 |
91 | 45 | 0.79 | |
| SOFSURF < 1.4 T-score < −2.5 |
90 | 46 | 0.77 | ||||||
| ORAI < 15 T-score < −2.5 |
84 | 52 | 0.76 | ||||||
| SCORE < 10 T-score < −2.5 |
90 | 33 | 0.77 | ||||||
| Park et al. (2003) [17] |
To validate the effectiveness of OSTA in identifying osteoporosis among Korean women | Postmenopausal women from a clinic in Korea and who were not on hormone replacement therapy (mean age: 59.1 ± 7.7 years) | 1101 | DXA GE Lunar model DPQ-IQ BMD at FN |
OSTA < −1 T-score ≤ −2.0 |
80 | 72 | 0.85 | Single-centered |
| OSTA < −1 T-score ≤ −2.5 |
87 | 67 | 0.873 | ||||||
| Geater et al. (2004) [18] |
To validate the performance of OSTA in predicting osteoporosis among Korean women | Thai post-menopausal women (mean age: 60.5 ± 9.7 years) without risk of osteoporosis | 388 | DXA Lunar, Madison BMD at FN and LS |
OSTA < −1 FN T-score < −2.5 |
93.5 | 60.8 | Value not mentioned | |
| OSTA < −1 LS T-score < −2.5 |
79.5 | 69.5 | Value not mentioned | ||||||
| OSTA < 0 FN T-score < −2.5 |
93.5 | 29.8 | Value not mentioned | ||||||
| OSTA < 0 LS T-score < −2.5 |
92.4 | 35.7 | Value not mentioned | ||||||
| Huang et al. (2015) [16] |
To determine the performance of OSTA among middle-aged and old women | Healthy women (age range: 40–96 years) from a hospital in Chengdu region, China | 15,752 | DXA (Lunar Prodigy- GE Healthcare, Madison, WI, USA) BMD at LS, FN, and TH |
OSTA < −1 LS T-score < −1 |
56.9 | 87.7 | 0.812 | |
| OSTA < −1 LS T-score < −2.5 |
77.3 | 73.5 | 0.812 | ||||||
| OSTA < −1 FN T-score < −1 |
56.2 | 89.8 | 0.822 | ||||||
| OSTA < −1 FN T-score < −2.5 |
88.1 | 69.3 | 0.822 | ||||||
| Yang et al. (2015) [21] |
To validate OSTA among elderly males to determine the risk of primary osteoporosis | Healthy males (mean age: 65.17± 9.29 years) | 245 | DXA (Hologic, Inc., Bedford, MA, USA) BMD at LS and LF |
OSTA < 1 FN T-score < −2.5 |
84 | 49 | 0.712 | |
| OSTA < 1 TH T-score < −2.5 |
Value not stated | Value not stated | 0.658 | ||||||
| OSTA < 1 LS T-score < −2.5 |
Value not stated | Value not stated | 0.535 | ||||||
| Oh et al. (2016) [22] |
To compare the effectiveness of Korean Osteoporosis Risk-Assessment Model for Men (KORAM-M) and OSTA | Men aged 50 and above from 2009 and 2010 Korean National Health and Nutrition Examination Survey | Development phase: 1340 Validation phase: 1110 |
DXA Hologic Discovery BMD at FN or LS |
Development: OSTA < −1 |
90.8 | 36.9 | 0.639 | |
| KORAM-M < −9 | 90.8 | 42.4 | 0.666 | ||||||
| Validation: OSTA < −1 |
92.3 | 33.2 | 0.627 | ||||||
| KORAM-M < −9 | 87.9 | 39.7 | 0.638 | ||||||
| Huang et al. (2017) [23] |
To assess the effectiveness of OSTA using various cutoffs | Healthy men aged 40–96 years recruited from a hospital in Chengdu region, China | 11,039 | DXA (GE Lunar, Madison, WI, USA) BMD at LS and FN | OSTA < −1 LS T-score < −1 |
27.6 | 89.2 | value not stated | |
| OSTA < −1 LS T-score ≤ −2.5 |
57.3 | 86.7 | |||||||
| OSTA < −1 FN T-score < −1 |
28.5 | 92.7 | |||||||
| OSTA < −1 FN T-score ≤ −2.5 |
65.9 | 87.0 | |||||||
| Bhat et al. (2017) [24] |
To evaluate the performance of OSTA in predicting OP among Indian men | Indian men above 50 years and without apparent risk of OP | 257 | DXA (QDR 4500 A, Hologic Inc., Bedford, MA, USA) BMD at LS, TH and FN |
OSTA ≤ 2 T-score at any sites ≤ −2.5 |
95.7 | 33.6 | 0.702 | |
| Zha et al. (2014) [4] |
To validate OSTA and QUS and their combination in predicting OP among the high-risk population | Chinese men (mean age: 78.0 years) | 472 | DXA (Discovery A, Hologic, USA) QUS (Sahara clinical bone sonometer- Hologic) BMD at LS and LH |
OSTA < −3.5 FN T-score < −2.5 |
65.5 | 74.8 | 0.724 | Small sample size Sample recruited from a single centre |
| OSTA < −3.5 TH T-score < −2.5 |
81.8 | 72.7 | 0.787 | ||||||
| OSTA < −3.5 LS T-score < −2.5 |
45.4 | 74.7 | 0.652 | ||||||
| OSTA < −3.5 T-score at any site < −2.5 |
47.3 | 76.8 | 0.676 | ||||||
| QUS < −1.15 FN T-score < −2.5 |
88.9 | 47.4 | 0.762 | ||||||
| QUS < −2.15 TH T-score < −2.5 |
82.4 | 86.6 | 0.883 | ||||||
| QUS < −1.25 LS T-score < −2.5 |
82.7 | 57.9 | 0.750 | ||||||
| QUS < −1.25 T-score at any site < −2.5 |
80.4 | 59.7 | 0.762 | ||||||
| Chang & Yang (2016) [25] |
To conduct a cutoff study among males by using OST, BMI, age and body weight | Retrospective data of Northern Taiwan males with mean age of 71.9 ± 13.3 years | 834 | DXA BMD at FN |
OST < −1.86 T-score ≤ −2.5 |
69.2 | 63 | 0.70 | Subjects were patients referred to BMD test by orthopaedic surgeons |
| BMI < 23 kg/m2 T-score ≤ −2.5 |
60.4 | 61.6 | 0.63 | ||||||
| Weight < 58.8 kg T-score ≤ −2.5 |
43.9 | 78.2 | 0.66 | ||||||
| Kung et al. (2003) [28] |
To develop OSTA for Asian men | Community-dwelling Chinese men (age: 50–93 years) | 420 | Development followed by validation in 356 men DXA: QDR 2000 Plus Hologic, Waltham, MA, USA BMD at LS and LF QUS: Sahara Hologic, Waltham, MA, USA |
Development: OSTA < −1 T-score ≤ −2.5 |
73 | 68 | 0.790 | Subjects were not selected randomly |
| Validation: OSTA < −1 T-score ≤ −2.5 |
71 | 68 | 0.780 | ||||||
| Validation: QUI < −1.2 T-score ≤ −2.5 |
76 | 72 | 0.80 | ||||||
| Either OSTA <−1 or QUI < −2.5 T-score ≤ −2.5 |
88 | 64 | 0.82 | ||||||
| Chan et al. (2006) [30] |
To compare the validity of various OP risk indices in elderly Chinese females | Community-dwelling postmenopausal women (age ≥55) | 135 | DXA (Hologic QDR 4500 W) BMD at FN and LS |
OSTA (cutoff ≤ −2 FN T-score ≤ −2.5 LS T-score ≤ −2.5 |
90.9 | 58.8 | 0.82 | Small sample size |
| 91.9 | 42.9 | 0.73 | |||||||
| SCORE (cutoff ≥ 8) FN T-score ≤ −2.5 LS T-score ≤ −2.5 |
93.9 | 60.8 | 0.80 | ||||||
| 86.5 | 60.2 | 0.72 | |||||||
| ORAI (cutoff ≥ 20) FN T-score ≤ −2.5 LS T-score ≤ −2.5 |
75.8 | 66.7 | 0.76 | ||||||
| 62 | 62 | 0.68 | |||||||
| ABONE (cutoff = 3) FN T-score ≤ −2.5 LS T-score ≤ −2.5 |
81.8 | 55.9 | 0.70 | ||||||
| 73 | 54.1 | 0.66 | |||||||
| SCORE (cutoff ≥ 8) FN T-score ≤ |
67.9 | 77.5 | 0.78 | ||||||
| 62.2 | 76.5 | 0.73 | |||||||
| Chaovisitsaree et al. (2007) [19] |
To compare OSTA with DXA in determining osteopenia and osteoporosis menopausal women | Thai menopausal women (age range: 45–87 years) from Menopause Clinic in Chiang Mai University | 315 | DXA BMD at FN, LS and radius |
OSTA < −1 LS T-score ≤−1 FN T-score ≤−1 Radius T-score ≤−1 |
36.2 | 71.4 | Value not mentioned | |
| 40.6 | 72.0 | ||||||||
| 48.3 | 75.1 | ||||||||
| OSTA < −1 LS T-score ≤−2.5 FN T-score ≤−2.5 Radius T-score ≤−2.5 |
45.8 | 68.9 | |||||||
| 75.0 | 67.8 | ||||||||
| 60 | 68.5 | ||||||||
| Chen et al. (2016) [27] |
To compare the performance of different screening tools to predict fracture or OP risk among older people | Community-dwelling older people aged 60 and above (mean age: 67.4 ± 6,4 years) recruited from Tanzi District, Taiwan | 553 | DXA Hologic Discovery Wi Bone Densitometer BMD at FN QUS GE Lunar, Madison, WI |
QUS FN T-score ≤ −2.5 |
20 (M) 59 (F) |
86 (M) 75 (F) |
0.72(M) 0.77(F) |
|
| ABONE ≥ 2 | 100 (M) 100 (F) |
28 (M) 10 (F) |
0.78(M) 0.70(F) |
||||||
| BWC < 70 kg | 100 (M) 100 (F) |
36 (M) 7 (F) |
0.92(M) 0.80(F) |
||||||
| FRAX Hip fracture (>3%) |
80 (M) 83 (F) |
71 (M) 54 (F) |
0.86(M) 0.75(F) |
||||||
| MOF (>20%) | 0 (M) 17 (F) |
99 (M) 96 (F) |
0.77(M) 0.71(F) |
||||||
| GARVAN Hip fracture (>3%) |
60 (M) 28 (F) |
79 (M) 95 (F) |
0.72(M) 0.80(F) |
||||||
| Any osteoporotic fracture (>20%) | 20 (M) 55 (F) |
96 (M) 73 (F) |
0.72(M) 0.75(F) |
||||||
| ORAI ≥ 9 | 100 (M) 100 (F) |
19 (M) 5 (F) |
0.87(M) 0.77(F) |
||||||
| OSIRIS ≤ 1 | 100 (M) 100 (F) |
29 (M) 6 (F) |
0.94(M) 0.83(F) |
||||||
| OSTA ≤ −1 | 100 (M) 100 (F) |
58 (M) 27 (F) |
0.94(M) 0.83(F) |
||||||
| SCORE ≥ 6 | 100 (M) 100 (F) |
45 (M) 15 (F) |
0.91(M) 0.80(F) |
||||||
| Chen et al. (2017) [35] |
To establish a prediction model to identify osteopenia risk in women aged 40–55 years | Taiwanese women recruited from a health checkup centre | 1350 | DXA (DPX-L; GE Lunar Health Care, Madison, WI, USA) BMD at LS |
OSTA ≤ 1 | 78 | 47 | 0.69 | Novel algorithm to predict osteopenia |
| OPAT ≥ 1 −1 ≥ T-score > −2.5 at LS |
87 | 42 | 0.77 | ||||||
| Panichyawat & Tanmahasamut (2012) [31] |
To compare the performance of OSTA and Khon Kaen Osteoporosis Study (KKOS) scoring system to predict OP among postmenopausal women in Thailand | Postmenopausal women (mean age: 55.8 ± 5.9 years) from menopause clinic | 441 | DXA BMD at FN and TH |
OSTA = −1 T-score at any site ≤ −2.5 |
51.7 | 77.4 | 0.65 | Subjects from a single centre |
| OSTA = 0 T-score at any site ≤ −2.5 |
66.7 | 57.1 | 0.62 | ||||||
| KKOS = −1 T-score at any site ≤ −2.5 |
56.3 | 71.8 | 0.64 | ||||||
| KKOS = 0 T-score at any site ≤ −2.5 |
57.5 | 67.2 | 0.62 | ||||||
| Oh et al. (2013) [36] |
To develop Korean Osteoporosis Risk-Assessment Model (KORAM) and compare its performance with OSTA | Postmenopausal women who participated in the 2009 and 2010 Korean National Health and Nutrition Examination Survey | Development: 1209 Validation: 1046 |
DXA QDR Discovery, Hologic BMD at TF, FN and LS |
Development: OSTA < 0 FN or LS T-score < −2.5 |
96.8 | 28.3 | 0.626 | |
| OSTA < 0 FN or LS T-score < −2.0 |
93.7 | 34.6 | 0.641 | ||||||
| KORAM < −9 FN or LS T-score < −2.5 |
91.2 | 50.6 | 0.709 | ||||||
| KORAM < −9 FN or LS T-score < −2.0 |
85.2 | 60.1 | 0.726 | ||||||
| Validation: OSTA <0 FN or LS T-score < −2.5 |
94.2 | 29.2 | 0.617 | ||||||
| OSTA < 0 FN or LS T-score < −2.0 |
90.9 | 35.0 | 0.629 | ||||||
| KORAM < −9 FN or LS T-score < −2.5 |
84.8 | 51.6 | 0.682 | ||||||
| KORAM < −9 FN or LS T-score < −2.0 |
79.2 | 60.2 | 0.697 | ||||||
| Lim et al. (2011) [37] |
To develop and validate Malaysian Osteoporosis Screening Tool (MOST) to detect low BMD in Malaysia | Healthy women (mean age: 51.3 ± 5.4 years) from a residential area | Development: 514 Validation: 72 |
DXA Norland XR-36 BMD at FN and LS |
OST < 2 FN T-score ≤−2.5 |
88 | 52 | Value not mentioned | |
| ORAI > 8 FN T-score ≤−2.5 |
90 | 52 | |||||||
| SCORE > 7 FN T-score ≤−2.5 |
89 | 58 | |||||||
| SOFSURF > −1 FN T-score ≤−2.5 |
92 | 37 | |||||||
| MOST ≥ 4 FN/LS T-score ≤−2.5 |
Development: 80.2 Validation: 100 |
Development: 55.5 Validation: 67.6 |
|||||||
| Ma et al. (2016) [33] |
To compare the performance of OSTA and BFH in determining osteoporosis among postmenopausal Han Chinese women | Community-dwelling Han Chinese postmenopausal women with age range of 40–89 years (mean age: 60.71 ± 8.47 years) | 1721 | DXA Hologic Discovery QDR Wi BMD at LS, FN and TH |
OSTA < −1 T-score at any sites <−2.5 |
65.28 | 77.15 | 0.782 | Subjects from a single centre |
| BFH-OST < −9.1 T-score at any sites <−2.5 |
73.58 | 72.66 | 0.797 | ||||||
| Lin et al. (2017) [32] |
To assess the performance new screening tool to determine osteoporosis | Development phase: Community-dwelling Han Chinese males aged 50 and above (mean age: 65.42 ± 8.8) Validation phase: Hospital-dwelling Han Chinese men |
Development: 1870 Validation: 574 |
DXA Discovery Wi, QDR series, Hologic, Waltham, MA, USA BMD at hip and LS |
Development: BFH-OSTM ≤ 70 T-score < −2.5 |
84.96 | 53.49 | 0.763 | |
| Validation: OSTA ≤ −1 T-score < −2.5 |
50.42 | 82.20 | 0.732 | ||||||
| BFH-OSTM ≤ 70 T-score < −2.5 |
89.92 | 48.57 | 0.795 | ||||||
| Satyaraddi et al. (2017) [34] |
To evaluate the performance of OSTA and Male Osteoporosis Risk Estimation Score (MORES) in predicting OP among South Indian rural elderly men |
Indian men aged 65 and above (mean age: 71.9 ± 5.2 years) recruited by cluster random sampling | 512 | DXA Hologic QDR4500 Discovery A BMD at LS and FN |
OSTA ≤ 2 LS T-score ≤ −2.5 |
94 | 17 | 0.716 | Further validation study is needed for a larger cohort of subjects |
| FN T-score ≤ −2.5 | 99 | 18 | 0.778 | ||||||
| MORES ≥ 6 LS T-score ≤ −2.5 |
98 | 15 | 0.855 | ||||||
| FN T-score ≤ −2.5 | 98 | 13 | 0.760 |
Abbreviation: AP, anteroposterior; AUC, area under curve; BMD, bone mineral density; BWC, body weight criteria; LS, lumbar spine; FN, femoral neck; TH, total hip; MOF, Major osteoporotic fracture; OP, osteoporosis; PF, proximal femur.