Introduction Policy makers, clinicians and patients need robust evidence of the effectiveness of healthcare interventions to inform decision-making. Highly aggressive cancer surgery has major implications for patients and requires substantial healthcare resources, but presents considerable challenges for evaluation. Although a randomised controlled trial is the optimal study design to investigate therapeutic effectiveness, randomisation is not possible where there are major, irreversible, differences in the treatments in intervention and control groups. Furthermore, with novel surgical techniques, as experience is gained there may be changes in patient selection, referral patterns and outcomes over time, introducing additional potential sources of bias. These methodological issues will be explored using data from a ongoing prospective, multi-centre evaluation of pelvic exenteration, a highly radical procedure for pelvic cancer where the alternative is palliative care.
Methods Baseline demographic, clinical and quality of life data are collected during the period of treatment decision-making, with clinical and quality of life outcomes assessed every 3 months thereafter. Trends in referral patterns, treatment choices, patient characteristics and outcomes will be compared between groups.
Results Since 2008, 144 patients have been assessed, with 111 (77%) undergoing pelvic exenteration. Although the absolute rate of patient referral has increased over time, there is no significant trend in the proportion undergoing surgery (χ21=0.46, p=0.5). Trends in group characteristics and outcomes to June 2011 will be presented.
Conclusion Non-randomised evaluations of complex surgery are challenging but necessary. Ensuring the comparability of treatment groups is critical for robust evaluation.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.