This summary provides the key points from the document in 3 sentences: Screening older women for vertebral fractures using a systematic approach can improve case finding. The COSHIBA study found that using a simple screening tool to identify women at high risk of vertebral fractures led to more prescriptions of osteoporosis medication and fewer new fractures over 12 months compared to standard care. Vertebral fractures are clinically important to identify as they significantly increase future fracture risk independent of bone mineral density measurements.

1. Fracture Risk Assessment:
Vertebral Fractures
Dr Emma Clark
Consultant Senior Lecturer in Rheumatology
7th
November, 2016

2. Disclosures
Unrestricted educational grants
• Lilly
• Servier
• Amgen
Unrestricted research grant
• P&G (now WC)
Consultancy work
• Servier

3. • Why are vertebral fractures important?
• Clinical assessment: Might this person have a vertebral
fracture?
• Maximising case finding – a systematic approach
• Summary

4. • Why are vertebral fractures important?
• Clinical assessment: Might this person have a vertebral
fracture?
• Maximising case finding – a systematic approach
• Summary

5. Fracture risk assessment: importance of
vertebral fractures
• Clinical tools such as FRAX under-estimate future fracture risk in
someone with a vertebral fracture1
• In FRAX, fracture is a dichotomous variable, but….
[1] Blank RD et al (2011) J Clin Densitom 14(3):205-2011
[2] Klotzbuecher CM et al (2000) J Bone Miner Res 15(4):721-739
[3] Siris ES et al (2007) Osteop Int 18(6):761-770

7. Fracture risk assessment: importance of
vertebral fractures
• Clinical tools such as FRAX under-estimate future fracture risk in
someone with a vertebral fracture1
• In FRAX, fracture is a dichotomous variable, but….
• Prior vertebral fracture increases future fracture risk 4-fold,
whereas prior non-vertebral fracture doubles subsequent fracture
risk2
• And – there is an association between severity of prior vertebral
fractures and subsequent fracture risk3
, independent of BMD
[1] Blank RD et al (2011) J Clin Densitom 14(3):205-2011
[2] Klotzbuecher CM et al (2000) J Bone Miner Res 15(4):721-739
[3] Siris ES et al (2007) Osteop Int 18(6):761-770

8. Fracture risk assessment: importance of
vertebral fractures
• Presence of vertebral fracture improves
estimation of future fracture risk
• Understanding of severity of vertebral fracture
improves fracture risk assessment
• Presence of vertebral fracture may change
management

9. • Why are vertebral fractures important?
• Clinical assessment: Might this person have a vertebral
fracture?
• Maximising case finding – a systematic approach
• Summary

10. Should I X-ray this person?

11. Clinical features that might indicate
someone has a vertebral fracture
• History
• Examination

12. Clinical features that might indicate
someone has a vertebral fracture
• History
• Examination

13. Clinical features that might indicate
someone has a vertebral fracture
• History
– Pain
– Reported height loss
– Traditional risk factors for osteoporosis

14. Clinical features that might indicate
someone has a vertebral fracture
• History
– Pain
– Reported height loss
– Traditional risk factors for osteoporosis

15. Pain
• VFs may be clinically "silent“
[1] O'Neill et al, Osteop Int 2004; 15: 760-765
[2] Ismail et al, Osteop Int 1999; 9: 206-213

16. Pain
• New vertebral fractures occurring during a 4-year follow-
up that did not come to clinical attention, were
nonetheless associated with a two- to three-fold increase
in back pain and limitation
• Attitudes toward back pain in older women, and access to
health care
[1] Nevitt MC et al, Ann Int Med 1998; 128: 793-800
7223 white women older than
65 from SOF

17. Pain
• VFs may be clinically "silent"
• No more back pain than women without VFs1
• More VFs means more pain2
• Quality of back pain?
[1] O'Neill et al, Osteop Int 2004; 15: 760-765
[2] Ismail et al, Osteop Int 1999; 9: 206-213
•Site
•Intensity
•Specific quality e.g. stabbing, burning

18. Lateral back pain identifies prevalent vertebral fractures in
postmenopausal women: cross-sectional analysis of a
primary-care based cohort
EM Clark, AP Hutchinson, EV McCloskey, MD Stone,
JC Martin, AK Bhalla, JH Tobias (2009) Rheumatology 49:505-512
Thoracic area
Waist area
Lower back/buttock
area

19. • Lateral waist pain is associated with a 4.5 fold increased
risk of VFs (OR 4.48, 95%CI 2.02 to 9.94, P<0.001)
Lateral back pain identifies prevalent vertebral fractures in
postmenopausal women: cross-sectional analysis of a
primary-care based cohort
EM Clark, AP Hutchinson, EV McCloskey, MD Stone,
JC Martin, AK Bhalla, JH Tobias (2009) Rheumatology 49:505-512

20. Using self-reports of pain and other variables to distinguish
between older women with back pain due to vertebral fractures
and those with back pain due to degenerative changes
EM Clark, R Gooberman-Hill, TJ Peters (2016) Osteop Int 27:1459-1467
• Women aged >65 who had a thoracic spinal radiograph in the
previous 3 months were recruited
• Used the McGill Pain Questionnaire along with other questions
about back pain

21. Using self-reports of pain and other variables to distinguish
between older women with back pain due to vertebral fractures
and those with back pain due to degenerative changes
EM Clark, R Gooberman-Hill, TJ Peters (2016) Osteop Int 27:1459-1467

22. Using self-reports of pain and other variables to distinguish
between older women with back pain due to vertebral fractures
and those with back pain due to degenerative changes
EM Clark, R Gooberman-Hill, TJ Peters (2016) Osteop Int 27:1459-1467
With vertebral fracture Without vertebral fracture
▪ Pain for a few days/weeks ▪ Pain for months to years
▪ Brief or momentary pain ▪ Other severe pain experiences
▪ Improvement of pain on lying ▪ Negative effect of weather
▪ Pain described as crushing ▪ Pain radiating down legs
No difference in pain severity, or bothersomeness of back pain

23. Using self-reports of pain and other variables to distinguish
between older women with back pain due to vertebral fractures
and those with back pain due to degenerative changes
EM Clark, R Gooberman-Hill, TJ Peters (2016) Osteop Int 27:1459-1467
Independent predictors:
•Age
•History of previous fracture
•Pain described as crushing
•Pain improving on lying down
•Pain not spreading down legs
AUC 0.85 (95 % CI 0.79 to 0.92)

24. Clinical features that might indicate
someone has a vertebral fracture
• History
– Pain
– Reported height loss
– Traditional risk factors for osteoporosis

25. History of height loss
• Reported height loss
= reported height at aged 25 minus measured height now
– is associated with presence of vertebral fractures1,2
• Trousers or skirts now too long
• Can’t reach up to cupboards that could reach before
[1] Nicholson PHF et al, Osteop Int 1993; 3: 300-307
[2] Tobias JH et al, Osteop Int 2007; 18: 35-43

26. Clinical features that might indicate
someone has a vertebral fracture
• History
– Pain
– Reported height loss
– Traditional risk factors for osteoporosis

27. Risk factors for osteoporosis
• In addition to age and gender
• Late menarche associated with increased risk1,2
• Current smoking is associated with increased risk2,3
• Steroid usage particularly important in children4
and men5
,
but probably not in inflammatory disease6
[1] Roy DK et al, Osteop Int 2003; 14: 19-26
[2] van der Klift M et al, JBMR 2004; 19: 1172-1180
[3] Jaramillo JD et al, Annals Am Thor Soc 2015; 12(5): 648-656
[4] LeBlanc CM et al, JBMR 2015; 30(9): 1667-1675
[5] Sugiyama T et al, Int Med 2011; 50(8): 817-824
[6] Ghazi M et al, Osteop Int 2012; 23(2): 581-587

28. Clinical features that might indicate
someone has a vertebral fracture
• History
• Examination

29. Clinical features that might indicate
someone has a vertebral fracture
• History
• Examination
– Increased thoracic kyphosis
– Rib to pelvis distance

30. Clinical features that might indicate
someone has a vertebral fracture
• History
• Examination
– Increased thoracic kyphosis
– Rib to pelvis distance

31. Increased thoracic kyphosis
• Kyphosis measured by video rasterstereograph1
or
radiographic angle measurement2
can predict women with
vertebral fractures
• The majority of men and women with the most
exaggerated kyphoses have no evidence of vertebral
fracture or osteoporosis3
– Degenerative disc disease was the most common finding
[1] Tan B-K et al, J Rheum 2008; 35(2):327-334
[2] Ensrud KE et al, J Am Ger Soc 1997; 45(6):682-687
[3] Schneider DL et al, J Rheum 2004; 31(4):747-752
1407 people aged 50-96 from the Rancho
Bernado study in the US

32. Clinical features that might indicate
someone has a vertebral fracture
• History
• Examination
– Increased thoracic kyphosis
– Rib to pelvis distance

33. Rib to pelvis distance
[1] Siminoski K et al, Am J Med 2003; 115:233-236

34. Rib to pelvis distance
[1] Tobias et al, Osteop Int 2007; 18:35-43

35. Rib to pelvis distance
[1] Tobias et al, Osteop Int 2007; 18:35-43
3.2%
12.2%

36. Examination for recent onset
vertebral fractures
• New osteoporotic vertebral fractures are tender to gentle
percussion whereas degenerative spinal disease is not1
[1] Langdon J et al (2010) Annals Royal Col Surg Eng 92(2):163-166

37. Summary of potentially useful features in
history and examination
• Traditional risk factors for osteoporosis
– Females
– Older age
– Previous fracture
– Smoking
– Steroids in men and children
• Back pain
– Lateral waist pain
– Back pain improving on lying down
• Reported height loss of >4cm
• Rib-to-pelvis distance of 1 finger

38. • Why are vertebral fractures important?
• Clinical assessment: Might this person have a vertebral
fracture?
• Maximising case finding – a systematic approach
• Summary

39. Maximising case finding: A systematic
approach for vertebral fractures
• Spinal radiographs
• VFAs
• Raising awareness in radiology

40. Maximising case finding: A systematic
approach for vertebral fractures
• Spinal radiographs
– In unselected older women from primary care
– In older women with back pain from primary care

41. Maximising case finding: A systematic
approach for vertebral fractures
• Spinal radiographs
– In unselected older women from primary care
– In older women with back pain from primary care
COSHIBA = RCT1
of a clinical tool for identifying which older
women should have spinal radiographs
[1] Clark EM et al (2012) JBMR 27(3):664-671

42. • Lateral waist pain is associated with a 4.5 fold increased
risk of VFs (OR 4.48, 95%CI 2.02 to 9.94, P<0.001)
Lateral back pain identifies prevalent vertebral fractures in
postmenopausal women: cross-sectional analysis of a
primary-care based cohort
EM Clark, AP Hutchinson, EV McCloskey, MD Stone,
JC Martin, AK Bhalla, JH Tobias (2009) Rheumatology 49:505-512

43. • Lateral waist pain is associated with a 4.5 fold increased
risk of VFs (OR 4.48, 95%CI 2.02 to 9.94, P<0.001)
Yes No
Yes 13 45
No 16 248
VF on X-ray
Presence of
lateral waist
pain
Sensitivity: 44.8%
Specificity: 84.6%
Lateral back pain identifies prevalent vertebral fractures in
postmenopausal women: cross-sectional analysis of a
primary-care based cohort
EM Clark, AP Hutchinson, EV McCloskey, MD Stone,
JC Martin, AK Bhalla, JH Tobias (2009) Rheumatology 49:505-512

44. COSHIBA
• Simple screening tool (based on a pilot study1
)
– History of previous fracture at any age
– Reported height loss
– Margolis back pain score
– Rib-pelvis distance
• Use of score with predetermined threshold
– Predicts all those with 2 or more VFs, and half of those with
1VF
– AUC 0.88 (0.80 to 0.97)
• Women were randomised to screening or standard
approach
[1] Tobias et al, 2007 Osteop Int 18: 35-43.
COSHIBA
Cohort for Skeletal
Health in Bristol and
Avon

45. Results
primary and secondary outcomes:
Control arm Screening arm
n (%) n (%) OR (95%CI), P value
NEW OSTEOPOROSIS MEDICATION PRESCRIPTION
Within 6 months of joining the study n=2921
yes
no
17 (0.9)
1925 (99.1)
19 (1.9)
960 (98.1)
2.24 (1.16, 4.33), P=0.016
Between 6 and 12 months of joining the study n=2710
yes
no
18 (1.0)
1788 (99.0)
9 (1.0)
895 (99.0)
0.99 (0.45, 2.23), P=0.998
NEW FRACTURES
Within 6 months of joining the study n=2921
yes
no
34 (1.8)
1908 (98.2)
15 (1.5)
964 (98.5)
0.87 (0.47, 1.61), P=0.664
Between 6 and 12 months of joining the study n=2703
yes
no
41 (2.3)
1752 (97.7)
6 (0.7)
904 (99.3)
0.28 (0.12, 0.67), P=0.004
[1] Clark et al, 2012 J Bone Miner Res 27:664-671
COSHIBA
Cohort for Skeletal
Health in Bristol and
Avon

46. Results
primary and secondary outcomes:
Control arm Screening arm
n (%) n (%) OR (95%CI), P value
NEW OSTEOPOROSIS MEDICATION PRESCRIPTION
Within 6 months of joining the study n=2921
yes
no
17 (0.9)
1925 (99.1)
19 (1.9)
960 (98.1)
2.24 (1.16, 4.33), P=0.016
Between 6 and 12 months of joining the study n=2710
yes
no
18 (1.0)
1788 (99.0)
9 (1.0)
895 (99.0)
0.99 (0.45, 2.23), P=0.998
NEW FRACTURES
Within 6 months of joining the study n=2921
yes
no
34 (1.8)
1908 (98.2)
15 (1.5)
964 (98.5)
0.87 (0.47, 1.61), P=0.664
Between 6 and 12 months of joining the study n=2703
yes
no
41 (2.3)
1752 (97.7)
6 (0.7)
904 (99.3)
0.28 (0.12, 0.67), P=0.004
[1] Clark et al, 2012 J Bone Miner Res 27:664-671
COSHIBA
Cohort for Skeletal
Health in Bristol and
Avon

47. COSHIBA Discussion
• COSHIBA is unlikely to be cost effective from an NHS perspective
• Rough estimate of cost per QALY of £30,000
• Need to target any systematic screening to a higher risk subgroup
[1] Clark et al, 2012 J Bone Miner Res 27:664-671
COSHIBA
Cohort for Skeletal
Health in Bristol and
Avon

48. Maximising case finding: A systematic
approach for vertebral fractures
• Spinal radiographs
– In unselected older women from primary care
– In older women with back pain from primary care

49. Using self-reports of pain and other variables to distinguish
between older women with back pain due to vertebral fractures
and those with back pain due to degenerative changes
EM Clark, R Gooberman-Hill, TJ Peters (2016) Osteop Int 27:1459-1467
Independent predictors:
•Age
•History of previous fracture
•Pain described as crushing
•Pain improving on lying down
•Pain not spreading down legs
•A cut-off of 0.39 gives a sensitivity of 0.77 and a specificity of 0.78
•Applying this cut-off to a group of older women with back pain would
identify 77% of those with a vertebral fracture whilst reducing the
number of radiographs performed by 60 %

50. Maximising case finding: A systematic
approach for vertebral fractures
• Spinal radiographs
• VFAs
• Raising awareness in radiology

51. VFAs as a routine part of traditional DXA
scans
• Vertebral Fracture Assessment

52. VFA: vertebral fracture assessment

54. VFAs as a routine part of traditional DXA
scans
• Vertebral Fracture Assessment
• Take extra time
– To carry out the scan - particularly on older machines
– To report
• Unselected VFAs vs a targeted approach

55. VFAs as a routine part of traditional DXA
scans
• Vertebral Fracture Assessment
• Take extra time
– To carry out the scan - particularly on older machines
– To report
• Unselected VFAs vs a targeted approach
• Need to have an impact on patient care i.e. change
management in enough to justify the extra resources

56. Our experience in Bristol1
Eligibility criteria for VFA
•Women aged >65 and men aged >70
– Osteopaenia with intermediate FRAX scores
– Very low T scores <-3.5
We report VFAs as
•No vertebral fracture seen (give level which can be interpreted)
•Definite vertebral fracture
•Suspicious for vertebral fracture and recommend a spinal
radiograph
[1] P33 Impact of VFA on management within the real-life setting of a busy NHS DXA service

58. Our experience in Bristol1
Practicalities
•Patient completes pre-DXA questionnaire
•DXA performed and answers to questions put into FRAX
•Whilst patient is still on the table, technician decides whether a
VFA should be performed
Impact over an 8-month period
•170 VFAs performed
•Management was changed in 18 (10.6%)
WARNING: do not rely on the manufacturers software without
human interpretation
[1] P33 Impact of VFA on management within the real-life setting of a busy NHS DXA service

60. Maximising case finding: A systematic
approach for vertebral fractures
• Spinal radiographs
• VFAs
• Raising awareness in radiology

61. Systematic approaches for radiology
departments
• Use the ‘F’ word
– Vertebral Fracture Initiative by IOF
www.iofbonehealth.org/

64. Systematic approaches for radiology
departments
• Use the ‘F’ word
– Vertebral Fracture Initiative by IOF
www.iofbonehealth.org/
– In all images that show a vertebral fracture e.g. CXR, spinal
radiographs, pelvis radiographs
• Sagittal realignment of all
CT chest abdo pelvis scans

65. • Why are vertebral fractures important?
• Clinical assessment: Might this person have a vertebral
fracture?
• Maximising case finding – a systematic approach
• Summary

66. Fracture Risk Assessment:
Vertebral Fractures
• Presence of vertebral fracture improves fracture risk
assessment, and may change management
• For individual patients, when assessing future fracture risk,
consider spinal radiographs if they have features in the history
and examination – more research needed
• For a systematic approach, routine VFAs on a selected
subgroup, and full engagement with radiology departments
seem promising

67. Thank you