Text this: Reliability of a pressure pain threshold protocol: secondary analysis of a longitudinal trial with cluster randomization