Centralization and directional preference: a systematic review.

Centralization is a symptom response to repeated movements that can be used to classify patients into sub-groups, determine appropriate management strategies, and prognosis. The aim of this study was to systematically review the literature relating to centralization and directional preference, and specifically report on prevalence, prognostic validity, reliability, loading strategies, and diagnostic implications. Search was conducted to June 2011; multiple study designs were considered. 62 studies were included in the review; 54 related to centralization and 8 to directional preference. The prevalence of centralization was 44.4% (range 11%-89%) in 4745 patients with back and neck pain in 29 studies; it was more prevalent in acute (74%) than sub-acute or chronic (42%) symptoms. The prevalence of directional preference was 70% (range 60%-78%) in 2368 patients with back or neck pain in 5 studies. Twenty-one of 23 studies supported the prognostic validity of centralization, including 3 high quality studies and 4 of moderate quality; whereas 2 moderate quality studies showed evidence that did not support the prognostic validity of centralization. Data on the prognostic validity of directional preference was limited to one study. Centralization and directional preference appear to be useful treatment effect modifiers in 7 out of 8 studies. Levels of reliability were very variable (kappa 0.15-0.9) in 5 studies. Findings of centralization or directional preference at baseline would appear to be useful indicators of management strategies and prognosis, and therefore warrant further investigation.