Quantitative ultrasound and risk factor enquiry as predictors of postmenopausal osteoporosis: comparative study in primary care

Introduction

The current recommendation for primary care physicians to identify women at high risk of osteoporosis relies on the assessment of clinical risk factors as a selection method for referral for dual energy x ray absorptiometry (DXA).1 DXA remains the “gold standard” diagnostic investigation for osteoporosis, but the restrictions of cost and availability necessitate an effective selection process. Little evidence exists about the value of enquiring about risk factors in primary care as a selection method, but it has been reported to be a poor predictor of low bone mass.2 Quantitative ultrasound scanning can be used to predict risk of osteoporotic fracture.3 Preliminary findings indicate that ultrasound scanning is as good as clinical risk factors for prediction of osteoporosis, but its role in primary care has yet to be clarified.4 We compared these selection methods in postmenopausal women in a primary care setting.

Participants, methods, and results

We assessed 200 consecutive women aged 60–69 years attending a primary care clinic between April 2000 and July 2002. Seven general practices in South Warwickshire referred women because of perceived risk (48%) or interest (52%). An experienced practice nurse completed a risk factor questionnaire, calculated body mass index, and did a heel ultrasound scan (Sahara densitometer). One general practitioner interviewed the women to clarify details and referred the women for DXA scanning of the hip and lumbar spine at a local hospital.

We deemed risk factor status to be positive if at least one criterion for referral for DXA according to the 1999 Royal College of Physicians' guidelines was present.5 We expressed quantitative ultrasound measurement as a T score and chose the level defining the lowest quarter of readings to assess sensitivity and specificity, as no agreed cut-off point for referral for DXA exists.

We obtained complete data for 190 women, of whom 31 (16.3%) had osteoporosis on DXA scan. We classified 113 (59.5%) women as risk factor positive—body mass index < 19 kg/m² (5), height loss > 2 inches (5 cm) or kyphosis (5), maternal hip fracture (20), early menopause or hysterectomy < 45 years (40), secondary amenorrhoea > 1 year (5), prednisolone 7.5 mg > 6 months (10), fracture after age 50 (43), x ray osteopenia (28), medical condition associated with increased risk of osteoporosis (13). Forty nine (25.8%) women had an ultrasound reading below T = - 1.7.

Risk factor enquiry was a poor predictor. Only 19% of women with risk factors had osteoporosis, and this method failed to identify one third of the osteoporotic women (table). However, ultrasound scanning with a cut-off point of T = - 1.7 almost doubled specificity compared with risk factors (McNemar's test P=0.006, 95% confidence interval 31% to 49%) for roughly the same sensitivity. Adding an ultrasound scan to risk factors improved sensitivity by 22% (P=0.015, 8% to 37%) and reduced specificity by 4% (P=0.015, - 8% to - 1%). This enabled identification of 90% of the women with osteoporosis and increased prediction for women without the condition.

Comment

The combination of quantitative ultrasound scanning and enquiry about risk factors detected more cases of osteoporosis and had slightly better predictive values than risk factors alone. Ultrasound alone had much better specificity than risk factors alone. However, good clinical practice requires an overall assessment of risk for osteoporosis rather than ultrasound scanning as a stand alone procedure. Ultrasound scanning is a simple, quick, non-ionising, portable, and inexpensive investigation. It provides general practitioners with an opportunity to improve on the current method of identification of risk of osteoporosis and selection for DXA and is an ideal test for practice nurses to perform in a nurse led osteoporosis clinic. Further evaluation of quantitative ultrasound scanning and assessment of its cost effectiveness are warranted.

We thank practice nurses Sue Mills and Debbie Gray at Bridge House for performing clinic assessments and Anthony Mander, gynaecologist at the University of Manchester, for his helpful comments on the paper.