Aims
To quantify patient preferences for the mode of radiation therapy treatment for prostate cancer.
Methods
A cross-sectional survey containing health preferences methods of best-worst scaling (BWS) and discrete choice experiment (DCE) was conducted. Patients attending a regional tertiary cancer centre within the last 5 years were invited. The BWS included attributes of treatment environment (standard and magnetic resonance linear accelerator), time on bed, wait time for treatment, bladder/bowel preparations; and the DCE added the attributes of number of treatments, requirements for gel insertion, and out-of-pocket costs. Multinomial logit modelling (MNL) analysed trade-offs, and relative attribute importance and marginal willingness-to-pay (mWTP) calculated. Latent class analysis (LCA) analysed heterogeneity in preference behaviours, and the probability of uptake of various clinical scenarios examined.
Results
146 respondents completed both BWS and DCE components, with the majority between ages 71-80, married and retired. The most important BWS attributes were time on bed and wait time for treatment, however with the added DCE attributes, cost, gel insertion and daily time were considered most important (p < 0.01). There was a positive preference for fewer treatments. Respondents were willing to pay up to AU$132 for the least number of treatments, and AU$328 to avoid gel insertion. LCA revealed two classes, the first (58%) focussed on cost, gel insertion, and bladder/bowel preparation and the second on number of treatments, wait time and time on bed. The highest predicted uptake (58.3%, IQR:56.4-60.1)) included standard linear accelerator, 15 minutes daily time, 3 weeks wait time, 2 fractions over 2 weeks, with no gel insert or cost.
Conclusions
Patients prefer shorter treatment times, without gel insertion and at lower costs. These findings have implications in planning future service delivery, particularly with online adaptive strategies for prostate radiation therapy.