
There is a need for efficient and scientifically sound clinical trials that examine the clinical utility of individualized treatment rules (ITRs) that individualize 2 or more medical treatments. These are trials assessing the effectiveness of the medical decision process taking place in the individualization of the treatments, not the effectiveness of the particular treatments involved. We developed a new experimental design for confirming the superiority of an ITR for 2 treatments over decision processes that ignore patient covariates. The design is built on a new published idea, the ITR’s individualization potential (IP), which compares the average effect on clinical response (AER) of the ITR with that of treatment without individualization (TWI). The AER of a decision approach compares the expected potential response of a patient treated with the treatment recommended by the approach versus the expected response should the patient be treated with the alternative therapy instead. The IP concept implies a novel way of constructing the control group of our design and testing the ITR’s utility. The experimental design provides estimates of the AERs of both the ITR and TWI, which we use to test the individualization potential’s statistical significance. In personalized medicine, one of the 2 treatments is often recommended to a subpopulation of lower-frequency patients more difficult to recruit. Thus, as another innovation, our formulation implies convenient sample size calculations that minimize the number of these patients in the total sample size, contrary to traditional methods of powering clinical trials that usually oversample them. Our design usually needs smaller sample sizes than popular designs and has a more appropriate control group. We illustrate calculating sample sizes for a hypothetical ophthalmology clinical trial.
Objectives: Flexible survival models can accommodate time-varying treatment effects in network meta-analyses (NMAs) but may yield biased estimates from naïve extrapolations when trial data have varying maturity. To enable more reliable comparison of long-term survival outcomes when there are differences in follow-up duration between studies, we proposed employing parametric spline NMA models leveraging historical trial data in a Bayesian multiparameter evidence synthesis (B-MPES) framework. Methods: B-MPES NMA models were applied to reconstructed interim overall survival data from 4 randomized phase 3 trials of first-line immunotherapy plus tyrosine kinase inhibitor (IO+TKI) versus sunitinib in metastatic renal cell carcinoma, for which median follow-up duration ranged between 18.1 and 30.6 mo. Trial observations were supplemented with conditional survival estimates from a historical trial of dual IO therapy versus sunitinib with a median follow-up of 55 mo. An alternative B-MPES model was informed by older historical trial data for sunitinib and a hazard ratio condition. Treatment effects were quantified by 5-y restricted mean survival time. B-MPES model performance was compared with a naïve proportional hazards (PH) spline model. Results: Leveraging historical trial data via B-MPES stabilized non-PH models that captured the diminishing treatment effects, whereas naive PH models were biased in favor of the IO+TKI therapy for which data were least mature, and naïve non-PH models yielded spurious and highly uncertain predictions. Relative effect estimates and treatment rankings were only moderately sensitive to the choice of historical trial data, while absolute survival probabilities exhibited high sensitivity. Conclusions: B-MPES survival models informed by suitable external data provide an approach to estimate long-term relative effectiveness in a network of treatments when there are nonproportional hazards and data from some studies remain immature.
Purpose: Patient-reported measurement of shared decision making (SDM) is widely used in research and quality improvement (QI). The reliability of measurement instruments is not well established, making it difficult to interpret scores, evaluate interventions and design policies to strengthen SDM. This study examined the test–retest reliability of the CollaboRATE and SHARED instruments and explored factors influencing the stability of patient-reported experiences of SDM. Design: This mixed-methods prospective study was nested within a QI programme at a single tertiary National Health Service hospital in the United Kingdom. Adults (>18 y) booked for elective surgery completed CollaboRATE (3 items assessing the perceptions of clinicians’ SDM skills) and SHARED (10 items assessing patient experiences of making treatment decisions) questionnaires at baseline (surgical booking consultation) and retest (5–10 d later) time points. Intraclass correlation (ICC) and kappa coefficients at item, total and top-score levels assessed test–retest reliability, with Bland–Altman plots describing agreement. Semi-structured interviews exploring the stability of patients’ experiences of SDM were conducted with a purposive subsample and inductive coding identified themes. Results: A total of 86 patients completed retest measurement (median interval = 8 d). Test–retest reliability was weak for CollaboRATE (ICC = 0.34, P < 0.001) and moderate for SHARED (ICC = 0.52, P < 0.001) total scores and moderate for CollaboRATE top scores (κ = 0.47, P < 0.001). Integration of qualitative results indicated 2 drivers of instability: 1) reflection on the SDM process and 2) need for greater support in SDM. Conclusion/Implications: CollaboRATE and SHARED patient-reported SDM instruments show fair-to-moderate test–retest reliability in a real-world surgical setting. Instability may reflect not only the measurement properties of instruments also but an evolution in patients’ perceptions of SDM. Clinicians, researchers and policy makers should consider the dynamic nature of SDM when planning the timing of SDM measurement and interpreting instrument scores.
Objectives: To test whether the assumption of constant health-related inequality aversion holds for Atkinson and Kolm-Pollak social welfare functions (SWFs) and to assess whether there is public support in Sweden for prioritizing the worse off when allocating health care resources.Methods: In an online survey experiment, respondents (n = 1,654) completed two task sets in which they chose between interventions benefiting better-off and worse-off groups to different extents. To test for constant inequality aversion, we checked whether the inequality-aversion parameter from the second task set was one of the parameters closest to that elicited from the first set.Results: A majority displayed nonconstant inequality aversion (71.4% for Atkinson, 72.3% for Kolm-Pollak). However, the design may not fully isolate inequality aversion from age preferences and cannot rule out learning effects. Among the valid responses (n = 1,029), 50.4% valued health gains to worse off higher, but only up to a point; 19.4% were unwilling to help the better off before nothing more could be done for the worse off; 14.4% preferred a more egalitarian system; 10.8% were unwilling to sacrifice total health gains to benefit one group more than the other; and 4.8% were willing to sacrifice total health to benefit the better off more than the worse off.Conclusions: If our findings were not confounded by age preferences or learning effects, the Atkinson and Kolm-Pollak SWFs would not seem to reflect the sample's preferences for prioritizing the worse off. Most respondents traded efficiency for equity in ways largely consistent with Swedish health care priority setting.
Purpose: To evaluate the association between patient-physician trust, treatment choice, patient-reported outcomes, and patient-reported experiences among patients with localized prostate cancer.Materials and Methods: We conducted a secondary analysis of data from a multicenter randomized controlled trial among patients with localized prostate cancer. Trust was measured at baseline, prior to treatment, using the Patient Trust-Wake Forest Physician Trust Scale (PTWF). Outcomes were assessed at baseline and at 3, 6, 12, and 24 mo. Treatments were robotic-assisted laparoscopic radical prostatectomy, radiation therapy, and active surveillance. The primary treatment outcome was receipt of active treatment (surgery or radiation therapy) versus active surveillance. Multivariable logistic regression and linear mixed-effects models were adjusted for age, race/ethnicity, marital status, education, employment, insurance, prostate cancer risk group, intervention assignment, and study site. Sensitivity analyses modeling trust as a continuous variable yielded findings that were materially consistent with the primary analyses.Results: The high-trust group had higher odds of receiving active treatment than the low-trust group did (adjusted odds ratio [OR] = 1.80, 95% confidence interval [CI] = 1.14-2.87; P = 0.013). Higher baseline trust was also consistently associated with better physical and mental health-related quality of life and higher satisfaction with care and treatment decision making across follow-up assessments through 24 mo. Sensitivity analyses using trust as a continuous measure produced similar findings.Conclusions: Higher patient-physician trust was independently associated with treatment choice, patient-reported outcomes, and patient-reported experience. These findings underscore the importance of patient-physician trust as a component of patient-centered cancer care and shared decision making. Future studies should evaluate interventions designed to strengthen trust and improve patient-centered outcomes across diverse clinical settings.