Table 3 Characteristics of studies included for key questions 1a&b on the benefits and harms of screening versus no screening, and the comparative benefits and harms of different screening approaches
Study; design; country; funding; analysis | Population characteristics | Screening approach (n randomized) | Treatment threshold Risk in those meeting threshold | Above treatment threshold; Treatment uptake | Outcomes; follow-up |
|---|---|---|---|---|---|
KQ1a: benefits and harms of screening versus no screening | |||||
Merlijn 2019 (SALT) [4] RCT Netherlands Foundation, industry, academic Analysis: selected population (high risk) | 11,032 (20.5% of 53,794 in age-based sample) females aged 65 to 90 y with ≥1 clinical risk factor; 47% of original sample completed FRAX, but 56% of these were ineligible or did not have a risk factor. Mean (SD) 75.0 (6.7) y; 44% prior fracture (location NR); 1% type 1 diabetes | Screening (n=5575): 2-step - Completed FRAX-UK - BMD + VFA if ≥ 1 risk factor - 76% of eligible for BMD participated Usual care (n=5457): completed FRAX-UK; advised to visit GP if ≥ 1 risk factor; 6% underwent DXA and VFA | Treatment threshold: any of a) lumbar/thoracic fracture with vertebral height reduction, b) exceeding age-specific FRAX + BMD MOF risk threshold, or c) risk score ≥4 according to Dutch guidelinesa Mean (SD) FRAX + BMD: 10-y MOF risk: 23.9 (9.6%) 10-y hip fracture risk: 10.6 (10.1)% | Above treatment threshold: 1417/4228 (34%) who underwent screening; 25% for the screening group Self-report of any osteoporosis medication: 21% in screened (69% with treatment indication); 5% in usual care (mainly bisphosphonates) | ≥1 hip fractures: self-reported and verified ≥1 MOF (hip, clinical vertebral, wrist, humerus): self-reported and verified All-cause mortality: reported by relatives Follow-up: ≥36 months |
Rubin 2018 (ROSE) [5] RCT Denmark Government, academic Analysis: offer-to-screen; selected population (completed FRAX) | 34,229 (18,605 with FRAX; 54.4% of eligible) females aged 65 to 80 y Median (IQR) 71 (8) y; 10% prior fracture (location NR) in those with FRAX; diabetes NR | Screening (n=17,072; 9279 with FRAX): 2-step - Completed FRAX-Denmark - BMD + VFA when 10-y risk of MOF was ≥15% - 71% of eligible for BMD participated Usual care (n=17,157; 9326 with FRAX): completed FRAX, risk not calculated; 25% had DXA scan after the index date | Treatment threshold: BMD T-score at any site ≤2.5; vertebral fracture on VFA. Median (IQR) FRAX + BMD: NR in those meeting threshold. Screened group (n=5009) with DXA: 10-y MOF risk: 22 (15, 29)% 10-y hip fracture risk: 8.1 (5.6, 13)% | Above treatment threshold: 1236/9279 (13%) who completed FRAX; 7% for the screening group Any osteoporosis medication (pharmacy records): 23% in screened (80% with treatment indication); 18% in controls | ≥1 hip fracture: records (ICD-10 codes) ≥1 MOF (hip, clinical vertebral, wrist, humerus): records (ICD-10 codes) Follow-up: median (IQR) 5 (1.3) y |
Shepstone 2018 (SCOOP) [6, 91] RCT Government, foundation United Kingdom Analysis: selected population (completed FRAX) | 12,483 (32.3% of eligible) females aged 70 to 85 y Mean (SD) 75.5 (4.2) y; 24% prior fracture (location NR); diabetes NR | Screening (n=6233): 2-step - Completed FRAX - BMD when 10-y risk of hip fracture met high risk threshold based on age - 92% of eligible for BMD participated Usual care (n=6250): completed FRAX, fracture probability not calculated; GP received letter stating patient’s involvement | Treatment threshold: exceeding age-specific 10-y hip fracture risk (FRAX-BMD) threshold Mean (SD) FRAX (no BMD): 10-y MOF risk: 30.0 (10.7)% 10-y hip fracture risk: 17.9 (10.9)% | Above treatment threshold: 898/2790 (32%) who completed FRAX + BMD; 14% for the screening group Any osteoporosis prescription (GP records): 1486/6233 (24%) in screened (78% with treatment indication in first 6 months); 16% in controls | ≥1 hip fracture: self-report, records ≥1 osteoporosis-related fracture (not hands, feet, nose, skull, vertebrae): self-report; records All-cause mortality: registry data, family members, GPs Health-related quality of life: self-report via EuroQol-5D, Short-Form 12 Health Survey Serious AEs: GPs recorded serious AEs Follow-up: 5 y |
RCT United Kingdom Foundation, industry Analysis: offer-to-screen; acceptors of screening (completed BMD) | 4800 (3128 attended / had complete follow-up; 65% of eligible); peri-menopausal females aged 45 to 54 y Mean (SD) 58.4 (3.7) y; prior fractures and diabetes NR | Screening (n=2400; 1764 attended): 1-step invitation to be screened by BMD via DXA Usual care (n=2400; 1364 with complete follow-up): not invited to be screened | Treatment threshold: BMD at any site within the lowest quartile of first 1000 women screened Baseline risk: NR | Above treatment threshold: NR; lowest quartile Self-reported uptake of any osteoporosis medication >3 months (bisphosphonates, raloxifene, hormone replacement therapy): 69% in screened (% with treatment indication NR); 59% in controls | ≥1 hip fracture: self-reported and verified ≥1 MOF (hip, wrist, vertebrae, humerus): self-report and verified General health: self-reported Health status (2-y follow-up): self-reported on Short-Form 36 Survey All-cause mortality: NR Follow-up: median 9.1 y in screened, 8.8 y in controls |
Kern 2005 [68] CCT (non-random allocation based on availability of screening) United States Government, foundation Analysis: selected population (enrolled in another study) | 3107 adults ≥65 years (87% of eligible study participants offered screening) Mean (SD) 76.2 (4.9) y; 56% female; <0.1% with radius or ulna fracture in past 5 y, other fractures NR; 1% diabetes | Screening (n=1422): 1-step offer to be screened by BMD via DXA; 97% completed scans Usual care (n=1685): not offered BMD scan | Risk definition: BMD below age-matched mean of densitometer manufacturer’s reference group Risk in those meeting threshold: NR | Above treatment threshold: 33% of those completing a DXA scan (392 females, 69 males); 32% in the screening group Any bone-enhancing medication (includes calcium, multi-vitamins, estrogen, calcitonin, bisphosphonates): 27% in screened (31% with treatment indication); NR in controls | Total number of hip fractures: hospital records (ICD-9 codes); verified against Medicare claims data All-cause mortality: surveillance of hospital records and verified against Medicare claims data Follow-up: mean 4.9 y |
KQ1b: comparative benefits and harms of different screening approaches | |||||
LaCroix 2005 (OPRA) [93] RCT (3-arm) United States Industry Analysis: offer-to-screen | 9268 (3167 [34%] participated) females aged 60 to 80 y Mean (SD) 70.0 (5.6) y; 17% prior fracture; diabetes NR | Universal screen (n=1986; 415 participated): 1-step invitation to be screened by BMD via DXA SCORE-based screen (n=1940; 576 participated): 2-step - All completed SCORE - BMD offered if score ≥7 - 74% were eligible for BMD SOF-based screen (n=5342; 2176 participated): - All completed SCORE - BMD offered if ≥5 clinical risk factors - 7% were eligible for BMD | Risk definition: ≥5 fracture risk factors and/or BMD T-score <−2.5 for 60–64 y or z-score <-0.43 for ≥65 y; prior fracture after age 50 (SOF-based group only) Risk in those meeting threshold: NR | Above treatment threshold: 28% of those screened in the universal group (6% of allocated); 32% of those completing the SCORE-based tool (7% of allocated); 18% of completing the SOF-based tool (7% of allocated) Any dispensed prescription for osteoporosis medication (includes alendronate, hormone replacement therapy, calcitonin, raloxifene): 13% in universal screening (21% of screened), 14% in SCORE-based (20% of screened), 13% in SOF-based (17% of screened) group | Total number of hip fractures; all non-pathologic (osteoporotic) fractures: hospitalization and outpatient visit records (ICD-9 codes) Follow-up: mean (range) 28 (24–33) months |