Discriminative validity and responsiveness of the Oswestry Disability Index among Japanese outpatients with lumbar conditions

The Oswestry Disability Index (ODI) is one of the most used assessment scales for patients with spine conditions, and translations into several languages have already been available. However, the scale's discriminative validity and responsiveness to the clinical change was somewhat understudied in these translated versions of the ODI. In this study, we independently developed a Japanese version of the ODI, and tested its discriminative and responsive performances among outpatients with various spinal conditions. We recruited 167 outpatients from seven participating clinics, and concurrently measured the translated ODI and MOS Short Form 36 (SF36) as a reference scale. We also obtained from medical records clinical information such as diagnoses, the past history of surgery, and existence of subjective symptoms and clinical signs. For testing discriminative validity, scores were compared by the number of symptoms and signs, with the trend test. Receiver operating characteristics (ROC) analysis was also conducted to compare ODI and SF36 in their performance to discriminate the existence of signs/symptoms, by chi-square test on the area under ROC curve (AUC). For 35 patients (17 clinically stable, 18 undergoing surgery and clinically significantly changed), the two scales were repeatedly administered after 3-6 months to compare responsiveness by using ROC analysis. The translated ODI and the SF36 Physical Function (PF) subscale showed a significant trend increase as the numbers of symptoms/signs increased. They also showed comparable performance in discriminating the existence of signs/symptoms (AUC=0.70-0.76 for ODI, 0.69-0.70 for SF36 PF, P=0.15-0.81), and clinical status change over time (AUC=0.82 for ODI, 0.72 for SF36 PF, P=0.31). Our results showed that the translated Japanese ODI showed fair discriminative validity and responsiveness as the original English scale showed.