Effects of an Online Training Program for Physiotherapists About Weight Management for People With Osteoarthritis

Summary

Participation Requirements

Description

The Centre for Health Exercise and Sports Medicine at the University of Melbourne have developed a 6-module online training program designed to up-skill physiotherapists in weight management for osteoarthritis patients. This is a randomised controlled trial evaluating the effectiveness of the traini...

The Centre for Health Exercise and Sports Medicine at the University of Melbourne have developed a 6-module online training program designed to up-skill physiotherapists in weight management for osteoarthritis patients. This is a randomised controlled trial evaluating the effectiveness of the training program on physiotherapist's confidence in weight management for osteoarthritis patients. After providing informed consent and baseline data, physiotherapists will be randomized to either a) an online training program or b) control group. Those randomized to the online training group will be provided access to the a 6-module weight management training program and have six weeks to complete all the modules. Those randomized to the control group will receive no intervention. At baseline and six weeks, primary and secondary outcome measures will be completed in both groups. The research question is: What are the effects of an online training program designed to up-skill physiotherapists in weight management for people with osteoarthritis? The primary objective is to evaluate the effects of an online training program on physiotherapist's self-reported confidence in knowledge about weight management for people with osteoarthritis. 80 participants will be recruited and randomization into the two study groups. To detect an effect size of 0.8 with 80% power and two sided significance level of 5%, 26 participants are required per group. Assuming a 35% loss to follow up, this will increase the same size to 40 per group (80 in total). The study will be conducted through The University of Melbourne. Participants will be able to access the training remotely in their own time and access all outcome measures online. A biostatistician will analyse blinded, de-identified data. Analyses will be conducted with participants in assigned groups, with multiple imputation to account for missing data if required. Changes in self-reported confidence scores (primary outcome) will be compared between groups at 6 weeks using linear regression (with group as a factor) adjusting for baseline scores and the stratifying variables. Similar analyses will be conducted for secondary outcomes measured at baseline and 6 weeks. Process measures will be described in terms of mean (SD)/median (IQ range) or number (frequency). Standard diagnostic plots will be used to verify model assumptions. If an imbalance is noted between groups for relevant baseline characteristics, sensitivity analyses will be conducted including these characteristics as covariates in the models assessing treatment effects. The biostatistician will also conduct planned exploratory analyses to investigate potential moderators that could influence the effect of the training at 6 weeks. Pre-identified potential moderators include practice location (predominantly private vs predominantly public), baseline self rated confidence in weight management and years of clinical experience. To assess the moderation of the effect of randomised treatment group by the binary potential moderator (practice location), an interaction term between randomised group and the potential moderator, as well as terms for the randomised group and the potential moderator, will be included in the outcome regression model. To assess the moderation of the effect of randomised treatment group by continuous potential moderators (self rated confidence in weight management and years of clinical experience), the multivariable fractional polynomial interaction approach will be applied. This approach allows for nonlinear functional forms of the continuous potential moderator to be included in the regression model for outcomes, with the potential for separate nonlinear functional forms in each treatment group.

Tracking Information