Estimated risk for chronic pain determined using the generic STarT Back 5-item screening tool

Abstract

Hiroyuki Oka,1 Ko Matsudaira,1 Tomoko Fujii,1 Norimasa Kikuchi,2,3 Yuri Haga,2 Takayuki Sawada,2,3 Junji Katsuhira,4 Takahiko Yoshimoto,5 Kayo Kawamata,1 Juichi Tonosu,6 Masahiko Sumitani,7 Satoshi Kasahara,7 Sakae Tanaka,8 1Department of Medical Research and Management for Musculoskeletal Pain, Faculty of Medicine, 22nd Century Medical & Research Center, University of Tokyo, Tokyo, 2Clinical Study Support, Inc., 3Department of Public Health, Aichi Medical University School of Medicine, Aichi, 4Department of Prosthetics & Orthotics and Assistive Technology, Faculty of Medical Technology, Niigata University of Health and Welfare, Niigata, 5Department of Rehabilitation, Kameda Medical Center, Chiba, 6Department of Orthopaedic Surgery, Kanto Rosai Hospital, Kanagawa, 7Department of Pain and Palliative Medicine, Faculty of Medicine, 8Department of Orthopaedic Surgery, University of Tokyo, Tokyo, Japan Objective: The generic STarT Back 5-item screening tool (STarT-G) is used to manage chronic pain in the lower back and elsewhere. This study evaluated the validity of the Japanese version of this generic screening tool. Materials and methods: Japanese participants between the ages of 20 and 64 years completed online surveys regarding pain. Survey reliability was assessed with internal consistency, as calculated using Cronbach’s alpha coefficients. Spearman’s correlation coefficients were used to evaluate concurrent validity between the STarT-G score and standard reference questionnaires. Associations between STarT-G scores and the presence of a disability due to chronic pain (DCP) were analyzed using receiver operator characteristic (ROC) curves. Results: Analyses ultimately included data obtained from 52,842 Japanese participants (54.4% male) with a mean (standard deviation) age of 47.7 (9.4) years. Approximately 1.5% of participants had DCP, and the mean STarT-G score was 1.2 (1.4). The Cronbach’s alpha coefficient was 0.71, indicating an acceptable reliability. The STarT-G score moderately correlated with the pain numerical rating scale (NRS) score (Spearman’s correlation coefficient: r = 0.34). When the STarT-G threshold was set at 4, the sensitivity and specificity of the DCP predictive model were 65.8% and 82.4%, respectively, and the area under the ROC was 0.808. Conclusion: The STarT-G was internally consistent and was able to distinguish between subjects with and without a DCP. Therefore, the STarT-G can reliably be used in the Japanese population to identify patients with DCP. Keywords: chronic pain, disability, primary care, psychological factors, screening tool, somatic symptoms

Keywords

• The generic STarT Back 5-item screening tool chronic pain validity