Neurology · Medical Literature
NegotiAge-An AI-Based Caregiver Negotiation Training: Results From the Multiphase Optimization Strategy (MOST) Trial.
Abstract
Family caregivers experience conflicts in caring for people with Alzheimer's dementia (PWD; e.g., siblings disagreeing, advocating with providers, PWD refusing assistance). Subsequently, family caregivers experience frustration, burden, and stress. Furthermore, these caregivers rarely receive formal conflict resolution training to ameliorate these concerns. We developed NegotiAge-an artificial intelligence (AI)-based negotiation training intervention that teaches caregivers how to resolve conflicts and provides the opportunity to practice negotiating with online avatars in real-world conflicts. We sought to optimize the potency and test the early effect of NegotiAge on family caregivers following the multiphase optimization strategy (MOST). National randomized factorial trial of NegotiAge with family caregivers of PWD (> 65 years). Caregivers were randomized to one of eight conditions and variables/outcomes assessed at baseline and 1 month following NegotiAge. The mixed methods project used dependent sample t-tests (quantitative) and constant comparative analysis (qualitative). Across 23 states, 125 family caregivers (mean-age 54 years, 77% female) completed the study. Participants experienced significantly less caregiver burden (-3.24, p < 0.002) and negative affect (-2.1, p < 0.0001) between baseline and 1 month following. Those receiving two or more exercises experienced further significant burden reduction (-5.28, p < 0.04). Qualitative analyses supported these findings: "There was a time when my mother was crying. Because of the training I was able to calm her."; "When we have a conflict, I focus on the interest of both of us-helped tremendously." In the month following, 54% (n = 67) of participants experienced conflicts and 74% (n = 93) applied skills learned. Among utilizers, there was a significant decrease in forcing behaviors (-1.24, p < 0.0001). NegotiAge, an AI-based negotiation training intervention, significantly reduced burden and improved negative affect among family caregivers. Engaging in more negotiation exercises further reduced burden. NegotiAge has the potential to improve the lives of caregivers and PWD. Further testing is needed to determine the direct impact on older adults and long-term effects. ClinicalTrials.gov identifier: NCT04837937.
Clinical relevance
Among family caregivers of persons with Alzheimer's dementia, can the AI-based NegotiAge negotiation training intervention — optimized via a national randomized factorial trial using the Multiphase Optimization Strategy (MOST) — improve caregiver outcomes (including positive affect, well-being, conflict-handling, and negotiation knowledge) at one month compared with alternative component combinations? __SNAPSHOT__ {"studyDesign":"Randomized factorial trial (2³ design) embedded within the Multiphase Optimization Strategy (MOST) framework; mixed-methods design combining quantitative analysis (dependent-sample t-tests) and qualitative analysis (constant comparative analysis)","setting":"National (United States); community-based, fully remote/online; participants enrolled across 23 states","sampleSize":"n=125 family caregivers who completed the trial (exact allocation across the 8 factorial conditions not reported in the available abstract)","population":"Family caregivers of persons with Alzheimer's dementia (PWD) aged >65 years; mean caregiver age 54 years; 77% female; 23 U.S. states represented; racial/ethnic composition not fully reported in the available abstract excerpt","intervention":"NegotiAge — an online, AI-based negotiation training program comprising instructional videos (interests-rights-power framework), print planning materials, and avatar-based negotiation practice via the Interactive Arbitration Guide Online (IAGO) platform. Caregivers were randomized to one of eight treatment-component combinations (factorial design) including variations in avatar scenario type (care recipient, family caregiver, physician) and training component dosage","comparator":"Other factorial treatment combinations within the 2³ design; no single inert placebo or no-treatment control arm — consistent with MOST optimization-phase methodology","primaryOutcome":"Caregiver positive affect and well-being (per the published MOST protocol); negotiation knowledge and conflict-handling style also assessed; exact primary endpoint labeling not explicitly confirmed in the available abstract excerpt","followUp":"1 month post-intervention (assessments at baseline and 1 month)","fundingSource":"National Institutes of Health / National Institute on Aging (NIH/NIA; grant numbers R01AG068421, R01AG058777, R01AG083034, and P30AG059988 referenced in related prior publications); no industry funding reported"}
Key findings
The available abstract is truncated and does not report quantitative outcome data (effect sizes, p-values, confidence intervals, or mean differences). Accordingly, no specific primary outcome results, NNT/NNH, or between-group comparisons can be summarized. From the separately published qualitative substudy of this same trial cohort (n=119 completers), 73% of participants reported managing at least one caregiving conflict in the month following NegotiAge, and 86% reported handling conflicts differently as a result of the training — with behavioral themes including shifting away from power tactics toward shared-interest negotiation, identifying root causes of conflict, and enhancing communication. These qualitative findings are self-reported and not verified against objective outcome measures. No adverse events attributable to the intervention are reported in any available source. Specific quantitative MOST optimization findings (which component combinations were superior, and by what magnitude) are not available from the truncated abstract and remain unreported in accessible sources at this time.
Practice implications
This MOST optimization-phase trial represents a methodologically novel approach to refining a multi-component behavioral intervention for family caregivers of persons with Alzheimer's dementia — a population with documented high burden, stress, and conflict exposure who currently receive no formal negotiation training. The MOST framework is explicitly not an efficacy trial; its purpose is to identify the most potent component combinations to inform a subsequent full evaluation-phase RCT. Therefore, this trial was designed to optimize, not to establish definitive efficacy. Whether NegotiAge reduces caregiver burden, stress, or improves patient care outcomes cannot be concluded from this study. Qualitative data from this cohort suggest plausible behavioral change in conflict management strategies at one month, but these are subject to self-report and social desirability bias and lack a control comparator. The population is predominantly female, middle-aged caregivers recruited nationally across 23 states, but racial/ethnic diversity and caregiver relationship type (spouse vs. adult child) are not fully characterized in available sources. Persons with limited English proficiency, low digital literacy, or without internet access are likely underrepresented. Because only the truncated abstract is available, a complete assessment of analytic rigor, attrition, and component-level results is not possible. Clinicians should regard this as a promising optimization study awaiting a confirmatory efficacy trial before any practice implications can be drawn. __LIMITATIONS__ Only the truncated abstract is publicly available; primary quantitative outcome data, effect sizes, p-values, confidence intervals, and component-level factorial results are not reportable from the available text — a critical limitation of this summary.,MOST optimization-phase design: the trial was not powered or intended to establish superiority over a control condition; absence of an inert control arm precludes efficacy conclusions.,Dependent-sample t-tests assess within-group pre-post change rather than between-group differences; this cannot determine whether observed changes exceed natural variation or placebo effects.,One-month follow-up is brief; durability of negotiation skill acquisition and any effects on caregiver burden or patient outcomes are unknown.,Qualitative outcomes rely on self-report and are subject to recall and social desirability bias.,Sample size of 125 distributed across 8 factorial conditions yields limited power for detecting interaction effects within each cell (approximately 15-16 per cell if evenly distributed — calculated).,Population skews female (77%) and middle-aged (mean 54 years); generalizability to male caregivers, spousal caregivers of advanced age, or non-English-speaking caregivers is limited.,Digital exclusion: the fully online format excludes caregivers without reliable internet access or digital literacy.,No clinical outcomes for the persons with dementia (PWD) are assessed; downstream impact on patient care is entirely speculative at this stage.,This trial represents only the preparation and optimization phases of MOST; a separate evaluation-phase RCT is required before efficacy can be established. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ NegotiAge is an AI-based, avatar-driven negotiation training program for family caregivers of persons with Alzheimer's dementia, evaluated here in a MOST optimization-phase factorial trial to identify the most potent component combinations — not to establish clinical efficacy. Quantitative primary outcome results are not available from the truncated abstract, and the trial design does not permit efficacy conclusions versus a control; a subsequent evaluation-phase RCT is needed. Clinicians caring for dementia caregivers should be aware of this innovative approach to conflict resolution training but should not change practice based on this optimization-phase study alone.