Problem elicitation to assess patient priorities in ankylosing spondylitis and fibromyalgia

To elicit patient priorities as outcome measures in ankylosing spondylitis (AS) and fibromyalgia (FM); to relate these measures to other outcomes; to assess construct validity and sensitivity to change of the problem elicitation technique (PET) questionnaire.

One hundred thirty-four patients with AS were randomly allocated to weekly sessions of group physical therapy or daily exercises at home, whereas 73 patients with FM were randomized into one of 3 groups (low impact fitness, biofeedback, controls). The PET questionnaire was applied by trained interviewers at baseline and at 6 (FM) and 9 (AS) month followup. A PET score was calculated at each assessment. Construct validity of the PET was assessed by correlation and multiple regression of baseline values with other disease outcomes (pain, stiffness, patient's global assessment, Sickness Impact Profile (SIP), Health Assessment Questionnaire (HAQ), Arthritis Impact Measurement Scale (AIMS), patient utilities). Sensitivity to change of PET was assessed against changes in these outcomes and by comparing the efficiency of the PET with other outcomes.

Patients with FM identified more problems (mean 6.8) than patients with AS (mean 4.4). Moreover, more patients with AS than with FM were unable to identify any problem at baseline (10% compared to 1%). The PET score improved from 14.9 to 11.3 (p = 0.0001) in patients with AS but did not change from 21.8 to 21.1 (p = 0.24) in patients with FM. Construct validity testing of the PET score showed statistically significant (p < 0.05) correlations with AIMS, utilities, SIP, HAQ, pain, stiffness, and patient's global health in both groups of patients (r varying from 0.22 to 0.66). By multiple regression pain explained 29% of the variance in PET scores among patients with AS. In FM patient global assessment accounted for 39% of total variance of PET scores, whereas pain explained another 15%. Changes in PET scores correlated significantly (p < 0.05) with changes in AIMS, utilities, pain, stiffness, and patient global health in both AS and FM (r varying from 0.22 to 0.51). Some 6% of the variance in changes in PET scores was explained by changes in pain in patients with AS and 35% by changes in pain and subjective health in patients with FM. Assessment of sensitivity to change revealed that efficiency of the PET score was 0.6 in patients with AS and 0.09 in those with FM. Compared to other outcomes this was reasonable in patients with AS but low in those with FM.

Obtaining patient priorities was generally feasible. In both groups of patients construct validity of the PET questionnaire was satisfactory. The PET was much more sensitive to change in patients with AS than in patients with FM.