Importance Anticholinergic and sedative medications contribute to poor clinical outcomes in older inpatients. Evidence-based deprescribing strategies are needed. Objective To evaluate whether an intervention bundle increases deprescribing of anticholinergic and sedative medications in older inpatients. Design, Setting, and Participants This stepped-wedge cluster randomized clinical trial was conducted in 6 Australian public hospitals from January 2022 to July 2023. Participants were patients 65 years and older with Drug Burden Index (DBI; a measure of anticholinergic and sedative medication exposure) score greater than 0 at admission and hospitalized for 48 hours or more. Data were analyzed from October 2023 to May 2026. Interventions The intervention bundle had 2 components. The informational component (automated DBI score calculation displayed in the electronic medical record, deprescribing guides, patient leaflets, and education module) was present in all trial phases, although clinicians were not informed of it during the control phase. The stewardship component (study stewardship pharmacists who promoted the informational component and made patient-level deprescribing recommendations to attending teams) was delivered in the intervention phase. The informational component was promoted 1 to 2 times monthly during the sustainability phase. Main Outcomes and Measures The primary outcome was the proportion of patients with 1 or more DBI-contributing medications deprescribed (ceased or dose reduced) at discharge compared with admission. Secondary outcomes included deprescribing recommendations in discharge summaries and in-hospital adverse events. Results Among 4827 included patients, 2798 (58.0%) were female, and the median (IQR) age was 83.0 (76.0-88.0) years. The proportion of patients with 1 or more DBI-contributing medications deprescribed was 38.1% (788 of 2069) in the control phase, 41.4% (326 of 788) in the intervention phase, and 37.0% (729 of 1970) in the sustainability phase. The adjusted absolute difference (aAD) between intervention and control was 5.3 percentage points (95% CI, 0.1-10.6), smaller than hypothesized. Deprescribing was predominantly by cessation; in the intervention phase, 256 of 788 patients (32.5%) had 1 or more DBI-contributing medications ceased, 91 (11.5%) had 1 or more doses reduced, and 21 (2.7%) had both. The aAD between the sustainability and intervention phase estimates was -5.3 percentage points (95% CI, -10.9 to 0.3). Deprescribing recommendations in discharge summaries were more prevalent in the intervention (176 of 751 admissions [23.4%]) than the control phase (287 of 1968 admissions [14.6%]; aAD, 10.5 percentage points; 95% CI, 6.0-15.1) and were lower in the sustainability phase (362 of 1910 [19.0%]; aAD, -4.0 percentage points; 95% CI, -8.7 to 0.8). In-hospital adverse events were similar across phases. Conclusions and Relevance In this stepped-wedge cluster randomized clinical trial, the intervention bundle increased deprescribing less than hypothesized, and the increase was not sustained once the stewardship component ended. Trial Registration Australian New Zealand Clinical Trials Registry: ACTRN12622000374763.
Importance Older adults with advanced chronic kidney disease (CKD) face consequential decisions about whether to initiate dialysis, yet advance care planning (ACP), the communication that prepares patients for such decisions, is documented infrequently in ambulatory settings. Objective To determine whether a video decision aid about treatment options for advanced CKD, combined with patient-created video declarations, increases ACP documentation among older adults with advanced CKD. Design, Setting, and Participants This multicenter, parallel-group randomized clinical trial was conducted at 12 ambulatory clinics from June 2021 to March 2025, with 12-month follow-up. Participants were adults 65 years or older with advanced CKD. Outcome assessors were masked to group assignment. Data were analyzed in April 2026. Interventions Intervention participants viewed a certified video decision aid depicting 3 treatment options for advanced CKD (hemodialysis, peritoneal dialysis, and conservative kidney management) and then created a video declaration, a recorded statement of goals, values, and treatment preferences in video, audio, or written format, that was sent to their treating clinician. Control participants received usual medical care. Main Outcomes and Measures The primary outcome was ACP documentation in the electronic health record within 12 months of enrollment or death, whichever came sooner. Documentation was ascertained by validated natural language processing-assisted human adjudication. Prespecified secondary outcomes included documented discussion of kidney failure treatment options and stated treatment preferences. Results A total of 453 patients were included (228 intervention [50.3%] and 225 control [49.7%]; mean [SD] age, 78.8 [6.4] years; 190 women [41.9%]; 4 American Indian or Alaska Native [1.1%], 8 Asian [1.8%], 38 Black [8.4%], 35 Hispanic or Latino [7.7%], 3 Native Hawaiian or other Pacific Islander [0.7%], and 375 White [82.8%] individuals). At 12 months, ACP documentation was higher in the intervention group than the control group (56.6% vs 40.0%; rate difference, 16.6%; 95% CI, 7.5%-25.7%; P < .001), as was documented discussion of kidney failure treatment options (46.1% vs 31.6%; rate difference, 14.5%; 95% CI, 5.6%-23.4%; P = .001). After viewing the video, fewer intervention participants than controls at enrollment were undecided about treatment (30 of 227 [13.2%] vs 92 of 223 [41.3%]), and more preferred conservative kidney management (115 of 227 [50.7%] vs 71 of 223 [31.8%]; P < .001 for the overall distribution). Conclusions and Relevance This randomized clinical trial found that, among older adults with advanced CKD, a video decision aid increased ACP documentation and reduced the proportion of patients who were undecided about treatment. Trial Registration ClinicalTrials.gov Identifier: NCT04347629
Importance Cancer drugs are frequently approved based on uncertain benefits and risks that are not communicated to patients. Objective To evaluate the effect of communicating uncertainties about the benefits and risks of new cancer drugs on changes in individuals' hypothetical decisions to take the drug. Design, Setting, and Participants Two randomized clinical trials recruited online a demographically diverse sample of US adults (census matched for age, sex, education, income, and race and ethnicity) from May 9 to July 26, 2025. Intervention Plain-language statements describing uncertainties about drug benefits and risks. Trial 1 randomized participants to 1 of 5 uncertainties: (1) single-arm trial design, (2) unrepresentative population in terms of race and ethnicity, (3) short study duration, (4) surrogate end point, or (5) no noticeable improvement. Trial 2 randomized participants to 1, 2, 3, or 4 uncertainties or to a control group. Main Outcomes and Measures The primary outcome in trial 1 was changes in hypothetical decisions to take a new cancer drug (within-group difference), while in trial 2 the primary outcome was the change in between-group differences in hypothetical decisions (control vs intervention groups). Secondary outcomes included perceived treatment effectiveness, worry associated with taking the drug, and trust in the information and the US Food and Drug Administration. Results A total of 6272 adults were included in the primary analysis (3102 [49.5%] women). In trial 1, learning about uncertainties reduced hypothetical decisions to take the drug by a mean of 10.1 percentage points (pp; 95% CI, -11.7 to -8.5 pp; P < .001). The reduction in hypothetical decisions to take the drug was observed for all uncertainties, including among those in the group that learned that it was unknown whether patients would notice an improvement with the drug (-14.1 pp; 95% CI, -17.9 to -10.4 pp; P < .001) and those who learned that the drug was not studied in patients of similar race or ethnicity (-12.8 pp; 95% CI, -16.4 to -9.2; P < .001). In trial 2, hypothetical decisions to take the drug decreased with each additional uncertainty, demonstrating a dose-dependent effect. Learning about uncertainties reduced perceived effectiveness and trust but did not affect worry. Conclusions and Relevance In 2 randomized clinical trials including a demographically diverse sample of US adults, learning about common uncertainties with a new cancer drug's benefits and risks reduced hypothetical decisions to take the drug.
Importance Anticholinergic and sedative medications contribute to poor clinical outcomes in older inpatients. Evidence-based deprescribing strategies are needed. Objective To evaluate whether an intervention bundle increases deprescribing of anticholinergic and sedative medications in older inpatients. Design, Setting, and Participants This stepped-wedge cluster randomized clinical trial was conducted in 6 Australian public hospitals from January 2022 to July 2023. Participants were patients 65 years and older with Drug Burden Index (DBI; a measure of anticholinergic and sedative medication exposure) score greater than 0 at admission and hospitalized for 48 hours or more. Data were analyzed from October 2023 to May 2026. Interventions The intervention bundle had 2 components. The informational component (automated DBI score calculation displayed in the electronic medical record, deprescribing guides, patient leaflets, and education module) was present in all trial phases, although clinicians were not informed of it during the control phase. The stewardship component (study stewardship pharmacists who promoted the informational component and made patient-level deprescribing recommendations to attending teams) was delivered in the intervention phase. The informational component was promoted 1 to 2 times monthly during the sustainability phase. Main Outcomes and Measures The primary outcome was the proportion of patients with 1 or more DBI-contributing medications deprescribed (ceased or dose reduced) at discharge compared with admission. Secondary outcomes included deprescribing recommendations in discharge summaries and in-hospital adverse events. Results Among 4827 included patients, 2798 (58.0%) were female, and the median (IQR) age was 83.0 (76.0-88.0) years. The proportion of patients with 1 or more DBI-contributing medications deprescribed was 38.1% (788 of 2069) in the control phase, 41.4% (326 of 788) in the intervention phase, and 37.0% (729 of 1970) in the sustainability phase. The adjusted absolute difference (aAD) between intervention and control was 5.3 percentage points (95% CI, 0.1-10.6), smaller than hypothesized. Deprescribing was predominantly by cessation; in the intervention phase, 256 of 788 patients (32.5%) had 1 or more DBI-contributing medications ceased, 91 (11.5%) had 1 or more doses reduced, and 21 (2.7%) had both. The aAD between the sustainability and intervention phase estimates was −5.3 percentage points (95% CI, −10.9 to 0.3). Deprescribing recommendations in discharge summaries were more prevalent in the intervention (176 of 751 admissions [23.4%]) than the control phase (287 of 1968 admissions [14.6%]; aAD, 10.5 percentage points; 95% CI, 6.0-15.1) and were lower in the sustainability phase (362 of 1910 [19.0%]; aAD, −4.0 percentage points; 95% CI, −8.7 to 0.8). In-hospital adverse events were similar across phases. Conclusions and Relevance In this stepped-wedge cluster randomized clinical trial, the intervention bundle increased deprescribing less than hypothesized, and the increase was not sustained once the stewardship component ended. Trial Registration Australian New Zealand Clinical Trials Registry: ACTRN12622000374763
