Table 2.
Performance of OST for non-Asians.
| Study | Objective | Subject Description | Number of Subjects Recruited | Methods | Cutoff | Sensitivity (%) | Specificity (%) | AUC | Remarks |
|---|---|---|---|---|---|---|---|---|---|
| Richy et al. 2004 [10] |
To validate and compare the performance of OST with other osteoporosis risk indices | Postmenopausal White women (mean age: 61.5 ± 8.8 years) without Paget’s disease or advanced osteoarthritis | 4035 | DXA: Hologic QDR 2000 BMD at any site |
OST < 2 T-score ≤ −2.5 T-score ≤ −2 |
86 82 |
40 44 |
0.726 0.713 |
Subjects were either referred or came spontaneously for osteoporosis evaluation and may differ in some ways from the general population |
| SCORE > 7 T-score ≤ −2.5 T-score ≤ −2 |
86 78 |
40 46 |
0.708 0.700 |
||||||
| ORAI > 8 T-score ≤ −2.5 T-score ≤ −2 |
76 73 |
48 51 |
0.670 0.668 |
||||||
| OSIRIS < 1 T-score ≤ −2.5 T-score ≤ −2 |
64 58 |
69 73 |
0.730 0.717 |
||||||
| Cadarette et al. 2004 [38] |
To validate the performance of osteoporosis risk indices to determine women at high risk of osteoporosis | Women (mean age: 62.4 years) with age range of 45–90 years | 644 | DXA BMD at FN and LS |
ORAI > 8 T-score < −2.5 |
92.5 | 38.7 | 0.80 | The study included data from women who have been selected for BMD testing |
| OST chart <2 T-score < −2.5 |
91.5 | 45.7 | 0.82 | ||||||
| OST equation < 2 T-score < −2.5 |
95.3 | 39.6 | 0.82 | ||||||
| Body weight criterion < 70 kg | 93.4 | 34.6 | 0.73 | ||||||
| Adler et al. 2003 [40] |
To assess the performance of OST in men | American men (mean age: 64.3 ± 12.3 years) recruited from pulmonary and rheumatology clinic | 181 | Hologic QDR 4500 (Hologic, Inc., Bedford, MA, USA) BMD at LS, FN and TH |
OST = 3 T-score ≤ −2.5 |
93 | 66 | 0.836 | The study was not designed specifically to validate OST Small sample size |
| OST= 3 T-score ≤ −2.0 |
74 | 72 | 0.815 | ||||||
| Ghazi et al. (2007) [41] |
To evaluate the performance of OST in predicting men with low BMD | White men (age range: 50–85 years) from a hospital in Morocco | 229 | DXA Lunar Prodigy Vision machine (GE) BMD at TH and LS |
OST = 2 TH T-score ≤ −2.5 |
87.5 | 58.2 | 0.787 | |
| OST = 2 LS T-score ≤ −2.5 |
63.6 | 59.5 | 0.660 | ||||||
| OST = 2 T-score ≤ −2.5 at any site |
64 | 60.3 | 0.667 | ||||||
| Lynn et al. (2008) [43] |
To evaluate the use of OST, Male Osteoporosis Screening Tool (MOST) and Quantitative Ultrasound Index (QUI) and body weight as osteoporosis screening tools | Caucasian and Hong Kong Chinese men, aged ≥ 65 years and community-dwelling from Osteoporotic Fractures in Men (MrOS) Study | 4658 Caucasian men 1914 Hong Kong Chinese men |
DXA Hologic QDR 4500 W (Hologic Inc.) BMD at LS and PF |
Caucasian: | ||||
| OST ≤1 T-score at any site ≤ −2.5 |
79.3 | 48.5 | 0.714 | ||||||
| OST ≤2 T-score at any site ≤ −2.5 |
87.6 | 36.1 | |||||||
| MOST ≤26 T-score at any site ≤ −2.5 |
88.5 | 50 | 0.799 | ||||||
| MOST ≤27 T-score at any site ≤ −2.5 |
94.7 | 37.8 | |||||||
| Chinese: | |||||||||
| OST ≤−2 T-score at any site ≤ −2.5 |
81.8 | 56.2 | 0.759 | ||||||
| OST ≤−1 T-score at any site ≤ −2.5 |
91.9 | 36.4 | |||||||
| MOST ≤21 T-score at any site ≤ −2.5 |
86.8 | 59.3 | 0.831 | ||||||
| MOST ≤22 T-score at any site ≤ −2.5 |
94.2 | 42.3 | |||||||
| Gourlay et al. (2005) [39] |
To compare the performance of three osteoporosis risk indices in two different age groups. | Postmenopausal women aged 45–96 years | 4035 | DXA: Hologic QDR 1000, 2000 and 4500 (Hologic Inc., Waltham, MA, USA) BMD at FN |
OST ≤ 1 Ages 45−64 years |
89.2 | 45 | 0.768 | Subjects from a single centre |
| OST ≤ −1 Ages ≥ 65 years |
84.6 | 47.5 | 0.762 | ||||||
| ORAI ≥ 8 Ages 45−64 years |
88.5 | 46.2 | 0.750 | ||||||
| ORAI ≥ 13 Ages ≥ 65 years |
89.2 | 44.7 | 0.747 | ||||||
| SCORE ≥ 7 Ages 45−64 years |
88.5 | 39.8 | 0.757 | ||||||
| SCORE ≥ 11 Ages ≥ 65 years |
88.8 | 42.3 | 0.745 | ||||||
| Sinnott et al. (2006) [42] |
To assess the performance of QUS, OST, WBC and BMI to predict low BMD in African American | African American men (age: 35 and above) recruited from clinics |
128 | DXA: GE Lunar (General Electric, Madison, WI, USA) BMD at LS and non-dominant hip QUS Achilles Plus System (Lunar, Madison, WI, USA) |
QUS ≤ −1 T-score ≤ −2.0 |
83 | 71 | 0.80 | Small sample size |
| OST < 4 T-score ≤ −2.0 |
83 | 57 | 0.83 | ||||||
| WBC < 85 kg | 74 | 50 | 0.70 | ||||||
| BMI ≥ 30 | 83 | 43 | 0.70 | ||||||
| Machado et al. (2009) [44] |
To compare three different OP risk indices at different cutoffs in determining individuals who are at risk of OP | Portuguese men age 50 and above (mean age: 63.77 ± 8.22 years) | 202 | DXA: Hologic QDR4500/c BMD at LS and PF |
OST < 1 OST < 2 OST < 3 OST < 4 |
47.1 61.8 75.5 85.3 |
72.6 63.7 50.0 32.7 |
0.598 0.627 0.632 0.590 |
|
| OSTA < 1 OSTA < 2 OSTA < 3 OSTA < 4 |
38.2 55.9 73.5 76.5 |
82.1 67.9 58.3 42.9 |
0.602 0.619 0.659 0.597 |
||||||
| BWC < 65 kg BWC < 70 kg BWC < 75 kg BWC < 80 kg |
26.5 47.1 73.5 82.4 |
89.3 77.4 61.3 35.7 |
0.579 0.622 0.674 0.590 |
||||||
| Richards et al. (2014) [57] |
To determine the performance of OST in predicting osteoporosis in males. | Male US veterans above 50 years recruited from VA Medical Centers | 518 | DXA: Hologic (Bedford, MA, USA) BMD at TH, LS or DF |
OST ≤ 6 T-score ≤ −2.5 |
82.6 | 33.6 | 0.67 | DXA machines from differed manufacturers were used and the results were not standardized. |
| Crandall et al. [54] [47] |
To compare the performance of USPSTF (FRAX) with OST and SCORE to predict osteoporosis | Women aged 50–64 years who participated Women’s Health Initiative Observational Study and Clinical Trials at three of the 40 clinical centres | 5165 | DXA Hologic QDR2000 or QDR4500 (Bedford, MA, USA) BMD at hip or LS |
USPSTF (FRAX ≥ 9.3%) FN T-score ≤ −2.5 |
34.1 | 85.8 | 0.60 | |
| OST <2 FN T-score ≤ −2.5 |
79.8 | 66.3 | 0.73 | ||||||
| SCORE >7 FN T-score ≤ −2.5 |
74 | 70.8 | 0.72 | ||||||
| Geusens et al (2002) [51] |
To compare the performance of 4 osteoporosis risk indices in determining postmenopausal women with low BMD | Women (45 years and above) from US clinic, Rotterdam Study (55 years and above), women screened for a clinical trial (55 to 81 years old) and women from the general clinic (50 to 80 years) | 1102 women from US clinic 3374 women from Rotterdam Study 23,833 women screened for a clinical trial 4204 women from the general clinic |
DXA Hologic (Waltham, MA, USA); Norland (Fort Atkinson, WI, USA); and Lunar (Madison, WI, USA) BMD at FN or LS |
OST <2 T-score ≤−2.5 |
88 | 52 | Value not mentioned | Large sample size Selection bias may occur |
| ORAI >8 T-score ≤−2.5 |
90 | 52 | |||||||
| SCORE >7 T-score ≤−2.5 |
89 | 58 | |||||||
| SOFSURF >−1 T-score ≤−2.5 |
92 | 37 | |||||||
| Wallace et al. (2004) [48] |
To compare the performance of five osteoporosis risk indices in determining postmenopausal African-American women with low BMD | Women (mean age: 59.4 ± 12.5 years) from an osteoporosis study | 174 | DXA Hologic QDR 2000 BMD at FN |
ABONE ≥ 2 T-score ≤ −2.5 |
73.0 | 59.6 | Value not mentioned | Small sample size |
| ORAI ≥ 9 T-score ≤ −2.5 |
65.6 | 78.9 | |||||||
| OST < 2 T-score ≤ −2.5 |
75.4 | 75.0 | |||||||
| SCORE ≥ 6 T-score ≤ −2.5 |
83.6 | 53.9 | |||||||
| Weight Criterion < 70 kg T-score ≤ −2.5 |
68.9 | 69.2 | |||||||
| Zimering et al. (2007) [55] |
To compare a novel osteoporosis screening tool with OST in predicting low BMD | Development phase: Caucasian men (mean age: 68.4 ± 10.2 years) Validation phase: Caucasian men (mean age: 68.4 ± 10.2 years) African American men (mean age: 60.9 ± 13 years) |
Development: 639 Caucasian men Validation: 197 Caucasian men 134 African American |
DXA Hologic QDR 4500 SL machine (Waltham, MA, USA) BMD at FN, TH and LS |
Caucasian Mscore (cutoff = 9) FN T-score ≤ −2.5 |
88 | 57 | 0.84 | Mscore is the first validated risk assessment tool developed in men |
| OST (cutoff= 4) FN T-score ≤ −2.5 |
85 | 51 | 0.81 | ||||||
| M score age-weight (cutoff = 9) FN T-score ≤ −2.5 |
85 | 58 | 0.81 | ||||||
|
African American Mscore = 9 FN T-score ≤ −2.5 |
NT | NT | NT | ||||||
| OST (cutoff = 4) FN T-score ≤ −2.5 |
100 | 72 | 0.99 | ||||||
| Mscore age-weight (cutoff = 9) FN T-score ≤ −2.5 |
100 | 73 | 0.99 | ||||||
| Jiang et al. (2016) [52] |
To compare the performance of screening tools with BMI alone in identifying early postmenopausal women with OP | Postmenopausal women (mean age: 57 ± 4.2 years) | 445 | DXA | BMI < 28 | 95 | 38 | 0.73 | Small sample size Low statistical power of detecting the difference in AUCs |
| OST < 2 T-score ≤ −2.5 |
79 | 56 | 0.73 | ||||||
| ORAI ≥ 9 T-score ≤ −2.5 |
74 | 60 | 0.69 | ||||||
| SCORE ≥ 6 T-score ≤ −2.5 |
92 | 34 | 0.75 | ||||||
| USPSTF ≥ 9.3% | 24 | 83 | 0.62 | ||||||
| RF ≥ 1 risk factors | 66 | 62 | 0.64 | ||||||
| Pecina et al. (2016) [53] |
To compare the effectiveness of risk tools to predict OP in women aged 50–64 | Retrospective data of women (mean age: 56.6 ± 3.4) who underwent DXA scan in a clinic | 290 | DXA BMD at hip/LS |
USPSTF FRAX ≥ 9.3% | 36 | 73 | 0.55 | |
| SCORE ≥ 6 | 74 | 42 | 0.58 | ||||||
| OST < 2 | 56 | 69 | 0.63 | ||||||
| ORAI ≥ 9 | 52 | 67 | 0.60 | ||||||
| Hawker et al. (2012) [56] |
To develop a screening tool to guide bone density testing in healthy mid-life women | Healthy women (age range 40–60) receiving their first BMD in an urban teaching hospital | 944 | DXA Lunar Prodigy (GE Healthcare, Madison WI, USA) BMD at FN, TH and LS |
New tool T-score ≤ −2.0 |
93 | 36 | 0.75 | Only Caucasian population is involved |
| OST ≤1 T-score ≤ −2.0 |
47 | Value not mentioned | 0.69 | ||||||
| Cook et al. (2005) [49] |
To assess the performance of various osteoporosis screening tools and quantitative ultrasound in relation to DXA scan | Postmenopausal women (age range: 29–87 years) recruited from DXA scanning clinics | 208 | DXA Hologic QDR 4500 C (Hologic Inc., Bedford, MA, USA) BMD at LS and PF |
OST < −1 T-score ≤ −2.5 |
0.52 | 0.82 | 0.716 | |
| SCORE T-score ≤ −2.5 |
0.5 | 0.83 | 0.720 | ||||||
| ORAI T-score ≤ −2.5 |
0.43 | 0.86 | 0.664 | ||||||
| QUS BUA calcaneus T-score ≤ −2.5 |
0.56 | 0.92 | 0.766 | ||||||
| VOS calcaneus T-score ≤ −2.5 |
0.61 | 0.72 | 0.723 | ||||||
| Perez-Castrillon et al. (2007) [46] |
To identify if the combination of OST and calcaneal DXA improves the diagnosis of OP | Males with a mean age of 47 ± 13 years and females with mean age of 66 ± 8 years recruited from two university hospitals | 67 males 94 females |
DXA: Pixi-Lunar, DPXL Lunar (Madison, WI, USA) and Hologic QDR-4500; Hologic Inc. (Bedford, MD, USA) BMD at right calcaneal and hip |
Men OST≤3 T-score < −2.5 |
39 | 86 | 0.623 | Small sample size |
| Women OST ≤ 2 T-score < −2.5 |
94 | 59 | 0.762 | ||||||
| Richards et al. (2009) [47] |
To evaluate the performance of OST in predicting low BMD in male patients with rheumatoid arthritis | Males (mean age: 65.4 ± 10.5 years) recruited from a multicenter registry of rheumatoid arthritis | 795 | DXA Hologic Inc. (Bedford, MA, USA) BMD at Femur and LS |
OST ≤ 4 | 64 | 54 | Not mentioned | Low lean body mass in RA could limit the utility of the OST in this population |
Abbreviation: AP, anteroposterior; AUC, area under curve; BMD, bone mineral density; LS, lumbar spine; FN, femoral neck; TH, total hip; NT, not tested; OP, osteoporosis; PF, proximal femur; QUS, quantitative ultrasound; RF, Risk Factor-Based Approach; USPSTF, the U.S. Preventive Services Task Force; WBC, Weight-based Criterion.