Reliability of the McKenzie spinal pain classification using patient assessment forms

Objectives To determine whether case studies presented on a completed McKenzie assessment form, commonly used in the McKenzie educational training programme, contain sufficient information to permit a therapist to reach a classification of the patient.

Design An inter-rater reliability study of patient classifications made based upon inspection of McKenzie assessment forms. The assessment forms of 50 patients with spinal pain (25 with low back pain and 25 with neck pain) were examined and classified into McKenzie treatment syndromes and sub-syndromes.

Setting Private physiotherapy clinics.

Participants Fifty physiotherapists, all with similar training in the McKenzie method.

Main outcome measures Classification into McKenzie treatment syndromes and sub-syndromes. The reliability of the judgements was expressed using multi-rater kappa (k) and percentage agreement.

Results The reliability of syndrome classification was k = 0.56 (95% confidence interval 0.46–0.66) with a percentage agreement of 91%. The reliability of sub-syndrome classification was k = 0.68 (95% confidence interval 0.67–0.69) with a percentage agreement of 76%.

Conclusions The reliability analysis suggests that the patient assessment forms evaluated in this study provided an adequate, but not ideal, clinical simulation because the reliability was less than that obtained from inspection of real patients. Future research comparing various types of clinical simulation is indicated to determine which provides the closest replication of the dynamics of the real patient encounter.