Activity interference was also recorded in the diaries daily using Item 5 from the 12-Item Short-Form Health Survey (Ware et al 1996), a 5-point scale anchored by ‘not at all’ through to ‘extreme interference’. To ensure completeness of follow-up, data from the diaries were collected by telephone interview at weekly intervals for the first four weeks, then monthly or until recovery for the subsequent eight selleck compound weeks (84 days in total). At

three months, a telephone exit interview was conducted at which the Neck Disability Index (Vernon and Mior 1991) was administered and pain scores were collected. Primary outcome: The primary outcome was the time taken from commencement of treatment to recovery from the episode of neck pain. The day of recovery from the episode of neck pain was defined as the first day of seven consecutive days on which the patient rated the intensity of their average daily neck pain as < 1 on the numerical rating scale from 0 to 10. Secondary outcomes: Secondary outcomes included time to recovery of normal activity as well as pain (numerical rating scale 0–10) and disability NU7441 concentration (Neck Disability Index scale 0–50) scores at

three months. Time to recovery of normal activity was defined as the first day of seven consecutive days in which the participant rated the degree of interference ‘not at all’. We examined 22 putative prognostic factors. Eight demographic variables were examined: age, gender, level of education, employment status, change of employment status due to neck pain, smoking habit, whether a compensation claim for neck pain had been lodged, and self-rated general health. Level of education was determined using items from the Australian Census 2001 (Trewin 2000). Employment status was determined using categories described by

Kenny et al (2000). Self-rated general health was measured using Item 1 of the 12-Item Short-Form Health Survey (SF-12). The 14 clinical variables examined were: pain intensity on the 0–10 numerical rating scale, duration of neck pain, disability measured by the Neck Disability Index from 0 (none) to 50 (worst), the physical (PCS) and mental health (MCS) component summary scales of the SF-12, presence of concomitant symptoms (upper limb pain, headache, upper back pain, lower back pain, dizziness and nausea), past history of neck pain, previous sick leave for GBA3 neck pain, and use of analgesics. The clinical course of the episode of neck pain was described using Kaplan-Meier survival curves and using descriptive statistics. Prognostic factors were evaluated using separate prognostic models for recovery from the episode of neck pain and disability at 3 months. The first stage involved examination of the univariate relationship between the outcome and each prognostic variable, using Cox regression (for time to recovery), and linear regression (for disability at 3 months). Variables with significant associations (p < 0.