Moving beyond validity and predictive values: Use of clinical utility indices to identify optimum cut offs of body mass indices to rule in and rule out obesity

Abstract

Background: Considering the limitations of validity and predictive values as stand-alone indicators of evaluating a screening/diagnostic test, this study assesses a composite index - Clinical Utility Index- of Body Mass Index (BMI) for assessment of obesity. Methods: The secondary data of bus drivers of an urban transportation facility regarding weight, height, age and body fat percentage collected during a health camp were used to calculate the validity, predictive values and clinical utility indices for different cut off values of BMI. Positive Clinical Utility Index (+CUI) was calculated by multiplying sensitivity and posited predictive value. Negative Clinical Utility Index (-CUI) was calculated by multiplying specificity and negative predictive value. Results: A BMI cut off value of 25kg/m2 had the maximum positive clinical utility(+CUI=0.75) and 26 kg/m2 had the highest negative clinical utility (-CUI=0.82). The BMI cut off of 25kg/m2 had good clinical utility across all age groups. Conclusions: BMI cut off value of 25kg/m2 has good clinical utility to diagnose obesity and BMI cut off of 26 kg/m2 has excellent utility for ruling out obesity in this high prevalent setting.

Keywords

• obesity
• validity
• predictive values
• clinical utility index