Osteoporosis screening in people with airways disease

We tested associations between risk factors and bone mineral density in airways disease subjects, and developed a clinical screening tool to identify people who could benefit from bone mineral density testing. Subjects were recruited through hospital outpatients and pharmacies (Newcastle, n = 172). With survey refinement, we then tested a revised tool in a second sample (Adelaide, n = 317). Study factors included oral/inhaled corticosteroid use, asthma severity, respiratory admissions, physical activity, percent predicted forced expiratory volume in one second (FEV1), body mass index, and smoking history. Outcomes were bone mineral density of lumbar vertebra (L2-4) and total (or neck of) femur. Analysis was logistic regression with generation of a simple screening algorithm based upon coefficients. Scoring algorithm risk factors for T-score of < - 2.0: age > or = 68 = 10 points, bone mineral density < 20 = 25, weight < 60 kg = 20, 60-69 kg = 10, > or = 80 cigarette pack years = 15, low-level leisure activity = 5, area under receiver operator curve 0.83. For a cut-off score of 10, sensitivity was 91.2%, specificity 53.9%, positive and negative predictive values 52.3 and 91.7%, and 67.2% were correctly classified. In conclusions, our model has acceptable sensitivity, although limited specificity. Use of this tool may reduce unnecessary referrals for bone mineral density measurement.