Baseline pain characteristics predict pain reduction after physical therapy in women with chronic pelvic pain. Secondary analysis of data from a randomized controlled trial.

Background and aims Women with chronic pelvic pain represent a heterogeneous group, and it is suggested that the existence of sub-groups can explain varying results and inconclusiveness in clinical trials. Some predictors of treatment outcome are suggested, but the evidence is limited. The primary aim of this study was to explore if selected pre-treatment characteristics of the participants in a recently conducted randomized controlled trial were associated with treatment outcome. Methods In this study secondary analysis of data collected in a randomized trial were conducted. The participants were women with chronic pelvic pain randomized to two different physical therapy treatments. Analyses in this study were performed for the whole group as a cohort. The primary outcome measure was change in pain intensity from baseline to 12 months, measured with the numeric rating scale (0-10). The women were asked to rate their mean pelvic pain intensity during the last 7 days. Based on previous research and on available variables from the randomized controlled trial four potential predictive factors were derived from the baseline data and assessed one by one in a linear regression model, adjusted for age and treatment group. The variables with strongest association (p < 0.10) with the primary outcome were further included in a multivariable linear regression model with backward selection, adjusted for age and treatment group. Results Fifty women (mean age 38.1, SD = 12.2) were included in the analysis. For these women the mean change in pain intensity was -1.2 points (95% CI -1.8 to -0.7) from baseline to 12 months. The multivariable regression model showed that pelvic pain duration of 6 years or more was associated with less decrease in pain intensity with a regression coefficient of 1.3 (95% CI 0.3-2.4). Baseline pain intensity was associated with higher pain reduction after PT treatment with a regression coefficient per SD increase in baseline pain of -0.6 (95% CI -1.1 to -0.1). None of the women with main pain site other places than in the pelvis reported any pain reduction after physical therapy treatment, but due to the small numbers the predictor was not included in the regression analysis. Conclusions We identified that pelvic pain duration of 6 years or more was associated with less pain reduction, and that higher baseline pain intensity was associated with higher pain reduction after physical therapy treatment in this sample of women with chronic pelvic pain. For the variable main pain site other places than the pelvis the results are unsure due to small numbers. Implications Based on our finding of long pain duration as a negative predictor for pain reduction, we emphasize that early intervention is important. Many of the participants in our RCT reported pelvic surgeries or other treatments prior to referral for PT, and we suggest that referral to a non-invasive intervention such as PT should be considered at an earlier stage. In order to tailor interventions to the individual women's needs, thorough baseline assessments, preferably in a multidisciplinary setting, should be performed.