Individual participant data are presented in Table 3 (see eAddenda for Table 3). These risk differences show that ‘improvement’ occurred significantly more often among participants in the

experimental group (Table 2). The ‘worst case’ analysis indicates that for every three patients treated, one more patient would achieve ‘improvement’ than would otherwise occur (95% CI 1.7 to 6.5). The ‘complete case’ analysis indicates that for every two patients treated, one more patient would achieve ‘improvement’ than would otherwise occur (95% CI 1.5 to 3.3). Although nearly 60% of the experimental group were using medication at baseline, there was no relationship between medication use and CHIR-99021 cost improvement in this group (RR 1.02, 95% CI 0.56 to 1.84). Analyses of follow-up scores for pain and activity limitations added medication use and duration of symptoms as covariates FDA-approved Drug Library cell line to account for baseline differences between groups. Therefore, Patient-Specific Functional Scale change scores were analysed with an ANCOVA rather than an unpaired t-test. The experimental group had better follow-up scores for pain and activity limitations with ‘moderate’ standardised mean differences (≥0.6 but < 1.2) (Hopkins 2011) (Table 4). NNT values show that

substantially greater proportions of participants in the experimental group achieved clinically important change scores for neck pain, arm pain, Neck Disability Index, and Patient-Specific Functional Scale

(Table 5). Individual participant data for these outcomes are again presented in Table 3 (see eAddenda for Table 3). There was no evidence to suggest that neural crotamiton tissue management was harmful. ‘Worst case’ intention-to-treat and ‘complete case’ analyses showed no difference in the prevalence of worsening between groups (Table 2). Additionally, no participants had to stop neural tissue management early because of an exacerbation and associated development of two or more abnormal neurological findings that they and the physiotherapist related to treatment. Sixteen participants (42%) reported an adverse event that they related to neural tissue management after 29 of the 151 treatments (19%). Questionnaires were returned for 25 of the 29 adverse events. The characteristics of these adverse events are summarised in Table 6. On average, an adverse event consisted of three to four unpleasant sensations (82 unpleasant sensations over 25 adverse events). Aggravation of neck or arm pain and headache were most common. Nearly all (95%) unpleasant sensations started within 24 hours of the previous treatment session and approximately 80% lasted < 24 hours. Importantly, no additional treatments were needed for any unpleasant sensation and 88% of unpleasant sensations had little or no impact on participants’ daily activities.