At-Home Cancer Directed Therapy Versus in Clinic for the Treatment of Patients With Advanced Cancer

Mean patient-reported rating of Cancer Connected Access and Remote Expertise

This hypothesis test will use patient ratings from a single 0-10 item from the Consumer Assessment of Healthcare Providers and Systems Cancer Care Survey assessing "your overall cancer care experience". Will be compared between arms using a two-sample t-test.

Patient-preferred treatment location

The proportion of patients who preferred care at home or expressed no preference will be computed and compared to 50% using a one-group test of proportions. Additional Likert patient feedback questions and "Was It Worth It" questions at each time point will be described using frequencies and relative frequencies by arm (or overall if applicable) and compared between arms (if applicable) using chi-squared tests. Numeric analog scale questions will be described using means and standard deviations and compared between arms (if applicable) using t-tests.

Patient level of comfort with receiving infusions at home

Patient responses to comfort level with receiving infusions at home will be described using frequencies and relative frequencies. The proportion of patients who express comfort (quite a bit or very much) will also be tabulated and a proportion greater than 70% will signify acceptance. Additional Likert patient feedback questions and "Was It Worth It" questions at each time point will be described using frequencies and relative frequencies by arm (or overall if applicable) and compared between arms (if applicable) using chi-squared tests. Numeric analog scale questions will be described using means and standard deviations and compared between arms (if applicable) using t-tests.

Patient-reported worthwhileness

Measured by the Was it Worth It questionnaire. The Was It Worth It questionnaire asks patients whether they thought that receiving chemo/infusions at home was worthwhile, whether they would do it again, and whether they would recommend it to others.

Patient-reported function

Measured by the European Organization for Research and Treatment of Cancer. Mean and standard deviation of each scale will be computed by arm. Means and standard deviations of each scale will also be computed at all other assessment time points. Mean changes from baseline in each scale at 8 weeks (and other assessment time points) will be compared between arms using a linear combination of parameters from a general linear mixed model. Each model will include all available data from all time points. Fixed effects will include arm, time point, and arm-by-time point interaction. Repeated observations by patient will be modeled using compound symmetric correlation structure over time. Such values as the mean change from baseline at 8 weeks by arm, and difference in mean change from baseline at 8 weeks between arms will be estimated with confidence intervals based on the mixed model. Comparisons at other time points will also be carried out and graphically displayed using mean plots.