Validity and reliability of the Dutch STarT MSK tool in patients with musculoskeletal pain in primary care physiotherapy

• Published in: PloS one 2021-03, Vol.16 (3), p.e0248616-e0248616
• Main Authors: van den Broek, Anke G, Kloek, Corelien J J, Pisters, Martijn F, Veenhof, Cindy
• Format: Article
• Language: English
• Subjects: Arthritis; Back pain; Biomedical materials; Cognitive ability; Data analysis; Demographic aspects; Diagnosis; Editing; Evaluation; Health care; Health care facilities; Health risks; Human motion; Inflammation; Innovations; Low back pain; Medical prognosis; Medical research; Medicine and Health Sciences; Methodology; Musculoskeletal diseases; Neurosciences; Osteoarthritis; Pain; Patients; People and Places; Physical therapy; Physiological aspects; Polymyalgia rheumatica; Primary care; Quality of life; Questionnaires; Rehabilitation; Reliability analysis; Research and Analysis Methods; Rheumatoid arthritis; Spinal cord; Spondyloarthropathy; Sports; Subgroups
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0248616

To evaluate the validity and reliability of the Dutch STarT MSK tool in patients with musculoskeletal pain in primary care physiotherapy. Physiotherapists included patients with musculoskeletal pain, aged 18 years or older. Patients completed a questionnaire at baseline and follow-up at 5 days and 3 months, respectively. Construct validity was assessed by comparing scores of STarT MSK items with reference questionnaires. Pearson's correlation coefficients were calculated to test predefined hypotheses. Test-retest reliability was evaluated by calculating quadratic-weighted kappa coefficients for overall STarT MSK tool scores (range 0-12) and prognostic subgroups (low, medium and high risk). Predictive validity was assessed by calculating relative risk ratios for moderate risk and high risk, both compared with low risk, in their ability to predict persisting disability at 3 months. In total, 142 patients were included in the analysis. At baseline, 74 patients (52.1%) were categorised as low risk, 64 (45.1%) as medium risk and 4 (2.8%) as high risk. For construct validity, nine of the eleven predefined hypotheses were confirmed. For test-retest reliability, kappa coefficients for the overall tool scores and prognostic subgroups were 0.71 and 0.65, respectively. For predictive validity, relative risk ratios for persisting disability were 2.19 (95% CI: 1.10-4.38) for the medium-risk group and 7.30 (95% CI: 4.11-12.98) for the high-risk group. The Dutch STarT MSK tool showed a sufficient to good validity and reliability in patients with musculoskeletal pain in primary care physiotherapy. The sample size for high-risk patients was small (n = 4), which may limit the generalisability of findings for this group. An external validation study with a larger sample of high-risk patients (≥50) is recommended.