Article Text
PDF
Statistics from Altmetric.com
Q In patients with osteoporosis, what is the accuracy of physical examination signs for diagnosing osteoporosis or spinal fracture?
Clinical impact ratings GP/FP/Primary care ★★★★★☆☆ Internal medicine ★★★★★☆☆ Gynaecology ★★★★☆☆☆ Rheumatology ★★★★★★☆ Endocrine ★★★★★★☆
METHODS
Data sources:
Medline (1966 to August 2004), hand searches of bibliographies of relevant articles including recent osteoporosis guidelines and 4 clinical skills textbooks, and contact with experts in the field.
Study selection and assessment:
studies that included data on the accuracy or precision of the history or physical examination signs for diagnosing osteoporosis, osteopenia, or spinal fracture and that used bone densitometry at any site or documented vertebral fracture as the gold standard. Studies with insufficient data to calculate likelihood ratios were excluded. Study quality was assessed (level 1 [highest] to level 3).
Outcomes:
sensitivity, specificity, and positive and negative likelihood ratios.
MAIN RESULTS
14 studies (13 815 patients) met the selection criteria. All but 1 study included women (mean age range 51–80 y). 11 studies (79%) had level 1 criteria, and 3 studies had level 3 criteria. No single physical examination sign ruled in osteoporosis or spinal fracture without needing further testing. Osteoporosis was best detected by weight <51 kg, kyphosis, self reported humped back, or <20 teeth; and spinal fracture was best detected by wall-occiput distance >0 cm, and rib-pelvis distance ⩽2 finger breadths (table). Individual studies that investigated such physical examination signs as height loss, armspan–height difference, grip strength, and hand skinfold thickness had variable or inconclusive results for detecting osteoporosis.
Physical examination signs for detecting osteoporosis or spinal fracture in postmenopausal women*
CONCLUSIONS
The most useful physical examination signs for detecting osteoporosis were weight <51 kg, kyphosis, self-reported humped back, and <20 teeth. Wall-occiput distance >0 cm and rib-pelvis distance ⩽2 finger breadths were the most useful physical examination signs for detecting spinal fractures.
Commentary
Green et al have tried to evaluate the accuracy of physical examination signs for diagnosing osteoporosis or spinal fracture. The authors concluded that no single manoeuvre is sufficient to rule out osteoporosis or spinal fractures without further testing. In patients who do not meet current bone mineral density (BMD) screening recommendations, several convenient examination findings (especially low weight) can substantially change the pretest probability of osteoporosis and suggest the need for earlier screening. Wall-occiput distance >0 cm and rib-pelvis distance <2 finger breadths suggest the presence of occult spinal fractures.
From an evidence-based perspective, none of these manoeuvres can, as a single manoeuvre, rule out osteoporosis or spinal fractures. The authors also provide the prevalence of osteoporosis and vertebral fractures for these physical examination findings, which is an important addition to relative risk estimates. More focus should be shifted to prevalence or the absolute risk of a fracture. It would be interesting to find out whether a combination of clinical risk factors and physical examination signs could enhance the predictive ability.
Other attempts to identify patients at high risk of osteoporosis or a high absolute risk of fractures include asking simple questions. For example, have glucocorticoids ever been used? Does the patient smoke, drink, have a history of fracture, or have a family history of fracture or secondary osteoporosis? In addition, existing tools such as the osteoporosis risk assessment instrument (ORAI) in the CaMos study can be used to identify women with low BMD and to predict osteoporosis based on clinical risk factors.1 Although the sensitivity of ORAI was high (93%), the specificity was low (46%), which is expected with most risk scores. All these can be used separately or together with physical examination signs to more accurately predict the absolute fracture risk for a patient and to determine an intervention or the need for a BMD measurement.
References
Footnotes
-
For correspondence: Dr C S Colón-Emeric, Duke University Medical Center, Durham, NC, USA. Colon001mc.duke.edu
-
Sources of funding: Bureau of Health Professions; National Institutes of Health; VA Medical Research Service.