Ancillary studies

Assessed via the Quality of Communication questionnaire. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Feeling Heard and Understood scale. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Death and Dying Distress scale. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Generalized Anxiety Disorder scale. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Patient Health Questionnaire, 9 items. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Functional Assessment of Cancer Therapy - General. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Edmonton Symptom Assessment System. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.

Assessed via the Ease of Use, Acceptability, Usefulness, and Safety questionnaire.

Assessed via Goals of Care. The Likert-scale and composite score responses will be modeled by mixed-effect analysis of variance with relation to treatment group and time point, with interaction. Change from baseline at each time point will be assessed by contrasts, without adjustment for multiple comparisons since this is a pilot study. Baseline variables which show evidence of differences may be utilized as model covariates to control for associated bias. Binary responses (including the primary objective) will be similarly modeled by mixed-effect logistic regression. Survey responses which are neither Likert, composite, nor binary, may have categories collapsed such that they can be analyzed as binary.