Expert guidelines recommend clinical set up and weekly portal images, not daily image guidance (IGRT), for palliative radiation (RT); however, unnecessary daily IGRT is often still used. Default order options leverage insights from behavioral economics to improve quality of care by introducing a treatment standard indicative of practice guidelines while preserving physician choice with the ability to opt out. We conducted a cluster randomized stepped-wedge clinical trial (NCT03110692) to test the effectiveness of introducing a default order option in the electronic health record (EHR) versus usual practice to reduce unnecessary daily IGRT for palliative RT. The trial was conducted in a large academic health system at 5 radiation facilities with a common EHR where 35 radiation oncologists practice. We defined 2 study groups: the main academic center versus 4 network community sites. Both groups had a 12-month control period (pre-intervention). In a stepped fashion, the main center and community groups were then randomly assigned to receive the intervention in 4-month periods. Subsequently, both groups had a 4-month post-intervention period. The intervention consisted of a templated default imaging order for palliative RT consisting of clinical set up with weekly portal imaging; physicians could also opt out and enter their own order. The primary outcome measure was the change in the proportion of palliative courses with daily IGRT. Generalized estimating equations were used to fit the outcome measure clustering at the physician level with period and group fixed effects and adjusting for secular trends. In secondary analyses, we adjusted for other covariates possibly associated with use of IGRT, including age, ECOG, insurance type, fraction number, dose per fraction, prior RT, and target (bone, soft tissue, brain, other). The trial was approved by the institutional IRB. In the pre-intervention/control arms, the mean rate of daily IGRT was 68% (465/681 courses). In the intervention arms, the mean rate of daily IGRT was 31% (92/297 courses). In the primary analysis, the default prescription intervention yielded a 53% reduction in the odds of daily IGRT (odds ratio (OR) 0.47, p=0.046, 95% CI: 0.22–0.98). This finding remained significant on secondary analyses adjusted for possible confounding variables (OR 0.44, p=0.04; CI 0.19–0.96). Use of weekly portal imaging, rather than daily IGRT, reduced mean set-up time by 2.0 minutes or 13%. Patients receiving palliative RT to the brain were least likely to be treated with daily IGRT (p=0.001); patients who had prior RT were more likely to be treated with daily IGRT (p=0.005). In a large academic health system comprised of academic and community sites, introducing a default order option significantly reduced unnecessary daily IGRT during palliative RT. This trial demonstrates that behavioral economic approaches can be effective in improving the quality of cancer care.
Read full abstract