Clinical prediction tools to identify patients at highest risk of myeloma in primary care

Koshiaris et al presented an equation for predicting 495 patients with myeloma within 2 years, who were aged ≥40 years.¹ Older age, male sex, back, chest, and rib pain, nosebleeds, low haemoglobin, platelets, white cell count, raised mean corpuscular volume, calcium, and erythrocyte sedimentation rate were selected as significant predictors. By using full blood count, an area under the curve (AUC) (95% confidence interval [CI]) was 0.84 (0.81 to 0.87), and sensitivity (95% CI) at the highest risk decile was 62% (55% to 68%). By using the all-test model, the AUC (95% CI) was 0.87 (0.84 to 0.90) and sensitivity (95% CI) at the highest risk decile was 72% (66% to 78%). Regarding the prediction model of myeloma, I understand that the independent variables may be limited for general physicians, and an interval period between medical check and diagnosis of myeloma may be important for the prediction model.

On this point, Blair et al conducted a 16-year follow-up study, and reported the significance of anthropometry for contributing diagnosis of myeloma in postmenopausal women.² In an age-adjusted model, weight and waist circumference significantly contributed to the risk of myeloma. In contrast, body mass index (BMI) did not relate to the risk of myeloma. This information was partly confirmed by reports by Hagström et al.³ During a median follow-up of 20 years, waist circumference and waist–hip ratio were significant predictors for myeloma, and BMI did not significantly become a predictor of myeloma. Body composition may be a good predictor for long-term risk of myeloma.