Large Language Models (LLMs) are increasingly used to support cancer patients and clinicians in decision-making. This systematic review investigates how LLMs are integrated into oncology and evaluated by researchers. We conducted a comprehensive search across PubMed, Web of Science, Scopus, and the ACM Digital Library through May 2024, identifying 56 studies covering 15 cancer types. The meta-analysis results suggested that LLMs were commonly used to summarize, translate, and communicate clinical information, but performance varied: the average overall accuracy was 76.2%, with average diagnostic accuracy lower at 67.4%, revealing gaps in the clinical readiness of this technology. Most evaluations relied heavily on quantitative datasets and automated methods without human graders, emphasizing "accuracy" and "appropriateness" while rarely addressing "safety", "harm", or "clarity". Current limitations for LLMs in cancer decision-making, such as limited domain knowledge and dependence on human oversight, demonstrate the need for open datasets and standardized evaluations to improve reliability.

Abstract Social discounting describes the tendency to show less generosity to more socially distant recipients. With respect to age differences, prior studies (Pornpattananangkul et al., 2019; Lin & Zhang, 2020) found that older adults showed more generosity to distant others (i.e., lower social discounting rates) than younger adults. However, prior research relied on small samples, compared extreme age groups, and focused on Asian countries. To examine generalizability, this pre-registered study examined social discounting and charitable giving in a U.S. adult life-span sample (N = 426, age 18-96, Mage = 50.93, SDage = 19.51, 51% female, 61% non-Hispanic White). The social discounting measure was adapted from Jones and Rachlin (2006) and involved five social distances (1, 5, 10, 50, 100) on a 100-point scale (0 = closest, 100 = most distant). For each social distance, participants were asked how much of $100 they would allocate to themselves versus a person at that distance. They made the same decision for their favorite charity. Contrary to previous findings and our predictions, results revealed a significant positive association between age and social discounting rates (p< .01), such that older adults’ generosity declined more with increasing social distance. Age differences remained robust after controlling for demographics. There were no significant age differences in average giving across social distances and in charitable giving, but, regardless of age, charitable giving was higher among those with lower social discounting rates (p< .001). Findings suggest that age differences in social discounting may be sensitive to cultural context.

Self-continuity (i.e., people's sense of connectedness with their past and future selves) declines with increasing distance from the present moment. Questions remain about the specific patterns of such temporal declines and about potential associations between self-continuity and temporal discounting (i.e., the tendency to devalue delayed outcomes). In response, the present research assessed both past and future self-continuity as well as temporal discounting across multiple time intervals in two U.S. adult lifespan samples (n = 86/461). Analyses examined the associations between self-continuity and temporal discounting and fitted mathematical models from the temporal discounting literature (exponential/hyperbolic/q-exponential) to self-continuity. The qexponential model provided the best fit for both self-continuity and temporal discounting data. However, associations between temporal discounting and past or future self-continuity were only found at single intervals but not for average levels or slope parameters. Theoretical implications, methodological limitations, and future directions are discussed.

To examine why older adults avoid choices more, two pre-registered, U.S.-based studies (N1 = 164, Mage_1 = 50.71, N2 = 485, Mage_2 = 51.08) evaluated the roles of cognitive load and affect. In Study 2, half were randomized to a writing condition theorized to decrease avoidance. Across studies, older age was positively associated with avoiding choices and improved affect after doing so. Avoidance was linked to perceiving higher cognitive loads but neither load nor other covariates accounted for older adults' avoidance tendencies. The writing condition resulted in lower cognitive loads and more positive decision-related affect but did not decrease avoidance.

Abstract Arthritis, which affects tens of millions of middle-aged and older adults, is one of the most common causes of activity limitation, disability, and chronic pain. Despite evidence that total joint replacement (TJR) can improve quality of life and physical functioning in arthritis patients, many eligible TJR candidates hesitate to pursue surgical intervention, and this reluctance is higher among older patients. Research shows that willingness to undergo TJR is influenced by a variety of psychosocial factors. Social relationships in particular can be an important source of support and information for patients when considering surgery. It is well documented that people’s social networks tend to be smaller and more close-knit as they age, and recent research suggests that decision support networks for hypothetical healthcare decisions show the same trend. The present study examined whether this pattern extends to realistic decisions about TJR. Using a name-elicitation approach, we collected information on the size of patients’ decision support networks (Msize = 3.30; rangesize = 0-8) in a sample of potential TJR candidates (N = 98; Mage = 67.88; rangeage = 40-93 years). As predicted, linear regression revealed that, compared to younger potential TJR candidates, older potential TJR candidates identified fewer members in their decision support networks (B= -0.033; p =.048). Additional analyses examined the role of potential covariates including demographics (e.g., gender, marital status), health (e.g., pain, functional status), and psychosocial variables (e.g., personality, health literacy). Theoretical implications and ramifications for clinical practice are discussed.

Objectives Many choices are made in the context of one's social network, but age differences in consulting others during decision making remain underexplored. This preregistered study examined age differences in the composition of decision-related social networks and associated social preferences and motivations. Based on prior findings regarding age differences in information-seeking and socioemotional preferences, we predicted that with advanced age people would consult fewer and more emotionally close social partners, consult them less frequently, rate consultations as more helpful, and prioritize social goals during decision making.Methods An adult lifespan sample (N = 485; Mage = 51.08; SDage = 19.63, rangeage = 18-96 years) considered hypothetical choices and reported how many and what types of people they would consult and how often they would consult them. They also identified their most important decision consultants and evaluated interactions with those individuals.Results As predicted, age was associated with consulting fewer people less frequently and this was not explained by having a smaller range of potential consultants in one's network. Contrary to predictions, the proportion of close social partners in decision networks and the perceived helpfulness of consultants did not vary significantly by age. Moreover, older adults were less likely than younger adults to prioritize relationship quality over decision quality. Controlling for potential covariates did not diminish the observed age effects.Discussion These findings align with the prior literature on age-related decrements in pre-decisional information seeking but suggest that age-related preferences for close social partners and socioemotional goals do not extend to decision-related consultations.

Abstract: This review examines how age-related differences in the experience of regrets relate to decision making with particular emphasis on memory-based and anticipatory processes. First, we summarize how reports of different types of regrets vary across the lifespan. For long-term/life regrets, older (vs. younger) adults tend to recall more intense and omission-based regrets, but not a higher number of regrets. For short-term/everyday regrets, older (vs. younger) adults tend to recall fewer and less intense regrets as well as similar or higher levels of decision satisfaction. For anticipated or future regrets, older adults are typically as accurate as younger adults in forecasting feelings, but they expect to experience lower levels of negative affect after missing out on desirable decision outcomes. Second, we link age-related differences in the experience of regrets to potential neural, cognitive, emotional, and motivational mechanisms. With respect to neural and cognitive processes, age-related cognitive changes may alter older adults’ ability to process regret-related information and to engage in counterfactual thinking. With respect to affect, older adults tend to attend more strongly to positive decision-relevant information and remember past decisions and experiences as more positively than they initially experienced them. Finally, with respect to motivation, older age is typically associated with a shift toward maintenance and loss prevention goals. In addition, older adults focus more on controlling their response to regrets rather than the regrettable decision itself. We conclude our review by identifying limitations of the extant literature and making suggestions for future research directions on the topic.

This study examined whether the average levels and the temporal extension of self-continuity varied over the course of the COVID-19 pandemic. Three demographically matched adult life-span samples (n = 223 each) were gathered in the U.S. in fall 2016, summer 2020, and winter 2022/23. Participants rated their self-continuity 1/5/10 years into the past and future. Multi-level analyses examined the effects of temporal distance, past/future direction, and assessment time while controlling for demographics. Average self-continuity did not vary across assessments, but the tendency to report lower self-continuity for more distant times was weaker during the pandemic, and the tendency to report lower self-continuity for past versus future was weaker during and after the pandemic. Discussion focuses on the role of slowed time perception during the pandemic and the possibility that mid- and post-pandemic reports of past self-continuity were elevated by nostalgic self-reflection, coupled with increased uncertainty about the future.

Abstract Self-continuity, the perceived connectedness with past and future selves, varies across the life span. Previous research found that both past and future self-continuity are positively associated with age and that such effects are more pronounced for more distant time intervals (Rutt & Löckenhoff, 2016). Questions remain about the appropriate mathematical modeling of such patterns and potential associations between age differences in self-continuity and temporal discounting, the tendency to devalue future outcomes. In response, the present study assessed both past and future self-continuity as well as temporal discounting across four temporal distances (0.5, 1, 5, 10 years) in a U.S. adult life-span sample (N=461, aged 18–96, M = 51.38, SD =19.54, 52% female, 65% Non-Hispanic White). When compared to linear and exponential models, hyperbolic models provided the best fit for both self-continuity and temporal discounting. Subsequent analyses therefore focused on hyperbolic slope parameters (k) derived for each participant. Chronological age was negatively associated with k for both self-continuity and temporal discounting, indicating that, with age, the perceived connectedness with past/future selves and the perceived value of delayed rewards show fewer decrements with increasing temporal distance. However, although k parameters for past/future self-continuity were significantly associated with each other, they were not significantly associated with temporal discounting. Moreover, age differences in temporal discounting remained significant after controlling for age differences in past/future self-continuity. The present study extends our understanding of previously reported age differences in self-continuity and temporal discounting but suggests that they cannot be traced to the same underlying mechanism.

This review examines how age-related differences in the experience of regrets relate to decision making with particular emphasis on memory-based and anticipatory processes. First, we summarize how reports of different types of regrets vary across the lifespan. For long-term/life regrets, older (vs. younger) adults tend to recall more intense and omission-based regrets, but not a higher number of regrets. For short-term/everyday regrets, older (vs. younger) adults tend to recall fewer and less intense regrets as well as similar or higher levels of decision satisfaction. For anticipated or future regrets, older adults are typically as accurate as younger adults in forecasting feelings, but they expect to experience lower levels of negative affect after missing out on desirable decision outcomes. Second, we link age-related differences in the experience of regrets to potential neural, cognitive, emotional, and motivational mechanisms. With respect to neural and cognitive processes, age-related cognitive changes may alter older adults' ability to process regret-related information and to engage in counterfactual thinking. With respect to affect, older adults tend to attend more strongly to positive decision-relevant information and remember past decisions and experiences as more positively than they initially experienced them. Finally, with respect to motivation, older age is typically associated with a shift toward maintenance and loss prevention goals. In addition, older adults focus more on controlling their response to regrets rather than the regrettable decision itself. We conclude our review by identifying limitations of the extant literature and making suggestions for future research directions on the topic.

Abstract Decision avoidance is more frequent with age, but associated mechanisms remain underexplored. To address this gap, a mixed-methods study captured decision makers’ subjective evaluation of choice options prior to making or avoiding choices. An adult lifespan sample (N = 227, 18-88 years, M = 50.71, SD = 19.22, 55% women) evaluated four health and consumer decision scenarios. Specifically, participants were asked to write down what they liked and disliked about the available choice options before making or avoiding a choice. Decision avoidance was positively associated with age. Open-ended responses were on average 33.08 words long (SD = 22.38) and contained a sum of 21.18 statements (SD = 16.42). The number of words or statements was unrelated to age. Responses were coded for remarks about positive (κ =.79-.88), negative (κ =.60-.74), and fixable/acceptable (κ =.32-.85) choice attributes, as well as choice preferences (κ =.34-.67). No significant findings emerged with respect to the latter two categories; among the former, the number of negative remarks dominated the positive ones. Decision makers who focused relatively more on negative attributes were older and reported poorer mental well-being. Among middle-aged and older adults, but not younger adults, a stronger focus on negative compared to positive attributes was associated with more pronounced decision avoidance tendencies. All findings remained significant after accounting for the number of words or statements. In combination, this may suggest that older age groups hold stricter decision standards or are less willing to trade-off between favorable and unfavorable choice attributes.

Decision makers are more likely to passively accept than to actively reject preselected default options. Age differences in such "default effects" have not been systematically examined, but prior research reports age-related variation in several known determinants of default effects suggesting that they may be more common in older as compared to younger adults. To address this question, a representative life span sample (N = 500; Mage = 49.90, SDage = 19.34; 51% female, 49% male; 67% non-Hispanic White) responded to a preregistered online study. Participants completed a default effect task comprising two conditions, one requiring opt-out and one requiring opt-in decisions (i.e., 15 vs. 0 preselected features each). Susceptibility to defaults was assessed as the discrepancy between the number of features selected within each condition. In addition, we collected data on known determinants of default effects (i.e., perceived endowment, endorsement, ease, experience making similar choices, importance of the choice, and affective responses to the choice). Finally, we screened demographic background, personality, socioemotional and health status, and cognitive ability. Susceptibility to default effects was evident both at the individual and the group level. Unlike hypothesized, older age did not predict greater susceptibility to defaults, and older adults were less likely to endorse determinants of default effect compliance. Of the covariates assessed, only identifying as non-Hispanic White, greater perceived endorsement, greater perceived ease, and lower perceived importance of making the right choice predicted decision makers' susceptibility to default effects. Thus, our findings suggest that susceptibility to decision defaults does not vary by age. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Prior research suggests that older adults seek less information in consumer choices than younger adults do. However, it remains unclear if intentional information avoidance plays a role in such effects. To test this possibility, we examined age differences in deliberate information avoidance in consumer decisions and explored a range of potential motives. Adult lifespan samples completed two pre-registered online studies, which assessed information avoidance using a slider scale (Study 1, N =195) and a forced-choice task (Study 2, N = 500). In Study 1, age differences in information avoidance were not significant, but methodological limitations could have obscured age effects. In Study 2, age was associated with higher information avoidance. Avoidance was higher among participants who reported that the information would not impact decision preferences, would elicit more negative affect, and would be useless. Although age was associated with lower perceived impact on decision preferences and lower concerns about affective responses, age differences in information avoidance remained significant when these variables were statistically controlled. In conclusion, in the context of consumer choices, deliberate information avoidance is higher among older consumers. Thus, interventions to promote the acquisition of relevant information would benefit from being tailored to the target age group.

Objectives Previous research suggests that self-continuity is higher in older ages, especially for more distant intervals. This study extends prior work by examining age and temporal patterns of self-continuity in 2 adult life-span samples from Germany and the United States. Methods German data (n = 1,656, aged 18-93) were drawn from the German Socioeconomic Panel. U.S. data (n = 230, aged 18-87) were collected through a survey research firm. Preregistered multilevel analyses examined the roles of age, temporal direction (past/future), and temporal distance (1/5/10 years), and explored the role of demographic covariates. Results In both data sets, self-continuity was higher in older ages and decreased with distance from the present, especially for the past. Interaction effects among age, temporal distance, and temporal directions were complex and varied across samples. Self-continuity was higher among married and more educated German participants and more affluent U.S. participants, but age differences remained robust when including demographic covariates. Discussion Findings replicate prior evidence for age-related increments in self-continuity but suggest that patterns vary by temporal distance and direction and may be sensitive to contextual factors.

Abstract Older adults are more likely to avoid decisions, which puts them at risk of experiencing negative consumer or health outcomes. Because the underlying reasons remain unclear, this pre-registered online study assessed the roles of (1) affect (especially regret), (2) subjective difficulty, and (3) cognitive ability in explaining age-related differences in decision avoidance. In addition, half of all participants were randomized to a writing intervention aimed at reducing avoidance tendencies. An adult lifespan sample (N = 432, Mage = 50.94, SDage = 19.53 years, 50% women, 62% Non-Hispanic White) chose between making or avoiding four consumer and health-based decisions. Affect, including regret, was assessed before (pre-decisional), during (peri-decisional), and after making decisions (post-decisional). Subjective difficulty was measured during the decision process. Finally, participants completed assessments of demographic background, socio-emotional experience, health, personality, and cognitive ability. We found that older adults were more likely to avoid decisions (p<.001), but age was unrelated to peri-decisional affect and subjective difficulty. Peri-decisional and post-decisional affect were not consistently related to avoidance, but participants who reported more difficulty were more likely to avoid decisions (p=.002). Covariates and cognitive measures could not explain age-related differences, but lower growth-goal-orientation and stronger maximization tendencies predicted decision avoidance after accounting for age (ps<.05). The intervention was successful at improving peri-decisional cognitive load and peri-decisional affect (ps<.05) but did not reduce decision avoidance. In sum, none of the assessed mechanisms could explain age-related increases in decision avoidance. Across age groups, perceived difficulty but not actual cognitive ability predicted avoidance tendencies.

OBJECTIVES:Clinical observations and studies of retrospective observer ratings point to changes in personality in persons with cognitive impairment or dementia. The timing and magnitude of such changes, however, are unclear. This study used prospective self-reported data to examine the trajectories of personality traits before and during cognitive impairment. DESIGN:Longitudinal observational cohort study. SETTING AND PARTICIPANTS:Older adults from the United States in the Health and Retirement Study were assessed for cognitive impairment and completed a measure of the 5 major personality traits every 4 years from 2006 to 2020 (N = 22,611; n = 5507 with cognitive impairment; 50,786 personality and cognitive assessments). METHODS:Multilevel modeling examined changes before and during cognitive impairment, accounting for demographic differences and normative age-related trajectories. RESULTS:Before cognitive impairment was detected, extraversion (b = -0.10, SE = 0.02), agreeableness (b = -0.11, SE = 0.02), and conscientiousness (b = -0.12, SE = 0.02) decreased slightly; there was no significant change in neuroticism (b = 0.04, SE = 0.02) or openness (b = -0.06, SE = 0.02). During cognitive impairment, faster rates of change were found for all 5 personality traits: neuroticism (b = 0.10, SE = 0.03) increased, and extraversion (b = -0.14, SE = 0.03), openness (b = -0.15, SE = 0.03), agreeableness (b = -0.35, SE = 0.03), and conscientiousness (b = -0.34, SE = 0.03) declined. CONCLUSIONS AND IMPLICATIONS:Cognitive impairment is associated with a pattern of detrimental personality changes across the preclinical and clinical stages. Compared with the steeper rate of change during cognitive impairment, the changes were small and inconsistent before impairment, making them unlikely to be useful predictors of incident dementia. The study findings further indicate that individuals can update their personality ratings during the early stages of cognitive impairment, providing valuable information in clinical settings. The results also suggest an acceleration of personality change with the progression to dementia, which may lead to behavioral, emotional, and other psychological symptoms commonly observed in people with cognitive impairment and dementia.

Abstract Many consequential decisions are made in the context of a social network, but it is not clear how lifespan developmental changes may influence the way people integrate others into their decision-making processes. Research shows that, compared to younger adults, older adults perform less exhaustive pre-decisional information searches, have smaller and more close-knit social networks, and place more value on harmonious relationships (e.g., Hess et al., 2015). Drawing on these findings, a pre-registered survey study examined age differences in the size, composition, and utilization of decision-related social networks as well as subjective goals and perceptions regarding joint decisions. An adult lifespan sample (N = 432; M age = 50.9375; SD age = 19.53267, range age = 18-96 years) considered hypothetical decisions involving specific (apartment, car, physician, and insurance plan) and generic (important vs. minor) scenarios. As predicted, linear regressions revealed that compared to younger respondents, older respondents reported that they would consult fewer people fewer times, and they were less likely to recruit distant social partners or professionals (ps

Abstract Older adults are more likely to underuse and actively avoid decision-relevant information. Although the mechanisms underlying age-related differences in information avoidance are still unknown, Fuzzy-Trace Theory (Reyna, 2012) suggests they may be due to age-related differences in information processing preferences: According to this framework, information is stored using both gist (subjective meaning) and verbatim representations (objective details/numbers). Older adults prefer gist information and gist-based information processing, whereas younger adults favor verbatim information and verbatim processing. To examine whether age-related differences in information avoidance are affected by age-related information and processing preferences, an adult lifespan sample (N = 432) completed a pre-registered online study. Participants responded to scenarios containing potentially aversive information and chose between accepting or avoiding the information. Information formatting was experimentally manipulated, with participants either being faced with receiving gist or verbatim-formatted information. Across age groups, participants were more likely to avoid verbatim numbers than qualitative gist information (p=.026); no associations were found between avoidance and information or processing preferences. At odds with prior research, age was associated with lower information avoidance (p=.010). Follow-up analyses revealed that older adults were less likely to avoid information in health-related contexts but not in contexts unrelated to health (e.g., consumer information; p=.010). Overall, participants were more likely to avoid non-health information than health information (p<.001). In sum, our findings revealed that older adults show a decreased likelihood of avoiding potentially aversive health information, but that age-related differences in information avoidance cannot be linked to age-related information or processing preferences.

Objectives Prior research has documented age differences in risky decisions and indicates that they are susceptible to gain versus loss framing. However, previous studies focused on "decisions from description" that explicitly spell out the probabilities involved. The present study expands this literature by examining the effects of framing on age differences in the Balloon Analogue Risk Task (BART), a widely used and ecologically valid measure of experience-based risky decision making that involves pumping a virtual balloon. Methods In a preregistered study, younger (aged 18-30, n = 129) and older adults (aged 60 and older, n = 125) were randomly assigned to either a gain version of the BART, where pumping the balloon added monetary gains, or a loss version, where pumping the balloon avoided monetary losses. Results We found a significant age by frame interaction on risk-taking: in the loss frame, older adults pumped more frequently and experienced more popped balloons than younger adults, whereas in the gain frame no significant age differences were found. Total performance on the BART did not vary by age or frame. Supplementary analyses indicated that age differences in pumping rates were most pronounced at the beginning of the BART and leveled off in subsequent trials. Controlling for age differences in motivation, personality, and cognition did not account for age differences in risk-taking. Discussion In combination, findings suggest that age differences in risk-taking on the BART are more pronounced when the task context emphasizes avoiding losses rather than achieving gains.

Abstract Self-continuity, the perceived connectedness to one’s past and future selves, predicts well-being and tends to be higher among older adults (Rutt & Löckenhoff, 2016). Self-continuity is also susceptible to major life changes (Sani, 2008). The pandemic brought dramatic changes in multiple life domains, especially for older adults who experienced higher rates of COVID-19 morbidity and mortality and may have been more likely to restrict their behavior as a result. This study examined whether age differences in self-continuity remained robust in response to the pandemic using representative U.S. survey data. Pre-pandemic data (Fall 2016, N = 230, aged 18 – 87, M = 50.85, SD = 16.40, 47.83% female, 67.83% white) and mid-pandemic data (Summer 2020, N = 230, aged 18 – 88, M = 50.20, SD = 19.49, 48.27% female, 64.78% white) were demographically matched using propensity scores. Participants rated their self-continuity one, five, and ten years into the past and future. Multi-level modeling examined the effects of age, temporal distance (in years), temporal direction (past vs. future), and assessment time (pre- vs. mid-pandemic) on self-continuity. Consistent with prior research, self-continuity was lower for more distant intervals, especially for the past. An interaction between temporal direction and assessment time indicated that future self-continuity was lower mid-pandemic than pre-pandemic. Across samples, however, advanced age was associated with higher self-continuity and the size of this effect did not vary by assessment time. Overall, findings suggest that even though future self-continuity decreased during the pandemic, existing age effects remained robust to this disruption.