Bier J.1, Osterlo R.2,3, Koes B.1, Verhagen A.1
1Erasmus MC, General Health, Rotterdam, Netherlands, 2VU University Amsterdam, Epidemiology and Biostatistics, Amsterdam, Netherlands, 3VU University Amsterdam, Health Sciences, Faculty of Earth and Life Sciences, Amsterdam, Netherlands
Background: Subgrouping patients is becoming an increasingly popular method with potential for applying targeted treatment. For low back pain in primary care the STarT (Subgroup Targeted Treatment) Back screening Tool (SBT) is probably the best known for subgrouping back pain patients combined with a targeted treatment. Unfortunately it was not available for neck pain, witch is globally the 4th largest musculoskeletal disorder causing disability.
Purpose: To modified the Dutch Version of the STarT Back screening Tool (SBT) to fit patients with neck pain and evaluated the reliability and validity of this SBT-neck.
Methods: The SBT-neck was modified to be used in patients with neck pain. After finalizing the SBT-Neck general practitioners and physiotherapists included patients which completed a baseline questionnaire and follow-up measurements at three days weeks and three months. The construct validity was assessed by formulating hypotheses about the expected magnitude of the Pearsons correlation between SBT-neck and other questionnaires measuring pain, activity, kinesiophobia and catastrophizing. The reproducibility was assessed using the quadratic weighted kappa and the specific agreement. Predictive validity was assed using relative risk ratios (RR) for, amongst others, persisting disability at three months. Content validity was analysed using floor- and ceiling effects.
Results: In total 100 patients were included, 58.0%was categorized as low-risk for poor outcome, 37.0% as medium-risk and 5.0% as high-risk. For the construct validity we found, as expected, amoderate to high correlation for all questions except for pain question 1 and 2 (low correlation) and activity question 3 (lower than expected correlation) with their reference questionnaires. The reproducibility had a quadratic weighted kappa of 0.85 and the specific agreement of 90.9% for low-risk and 66.7% for medium-risk. The RR for persisting disability and persisting pain for medium-risk compared to low risk were 3.8 and 3.9 respectively. For high-risk, compared to low-risk were 5.9 and 3.7 respectively. We found no floor and ceiling effects.
Conclusion(s): The SBT-neck is successfully modified to predict persistingdisability and pain in patients with neck pain in Dutch primary care. The psychometric analysis showed sufficiently valid and reliable outcomes to predict persisting neck pain.
Implications: The SBT-neck can be used, in an early stage, to predict persisting disability and pain. Further research is needed to determine if stratified care can be added and will lead to a faster recovery.
Funding acknowledgements: CZ health care and the Dutch Artritis foundation
Topic: Musculoskeletal: spine
Ethics approval: Approved by the Medical ethics commition of the ErasmusMC, Rotterdam, The Netherlands, Number: METC-2014-256
All authors, affiliations and abstracts have been published as submitted.
