Machine learning analyses identify multi-modal frailty factors that selectively discriminate four cohorts in the Alzheimer's disease spectrum: a COMPASS-ND study.

BACKGROUND: Frailty indicators can operate in dynamic amalgamations of disease conditions, clinical symptoms, biomarkers, medical signals, cognitive characteristics, and even health beliefs and practices. This study is the first to evaluate which, among these multiple frailty-related indicators, are important and differential predictors of clinical cohorts that represent progression along an Alzheimer's disease (AD) spectrum. We applied machine-learning technology to such indicators in order to identify the leading predictors of three AD spectrum cohorts; viz., subjective cognitive impairment (SCI), mild cognitive impairment (MCI), and AD. The common benchmark was a cohort of cognitively unimpaired (CU) older adults. METHODS: The four cohorts were from the cross-sectional Comprehensive Assessment of Neurodegeneration and Dementia dataset. We used random forest analysis (Python 3.7) to simultaneously test the relative importance of 83 multi-modal frailty indicators in discriminating the cohorts. We performed an explainable artificial intelligence method (Tree Shapley Additive exPlanation values) for deep interpretation of prediction effects. RESULTS: We observed strong concurrent prediction results, with clusters varying across cohorts. The SCI model demonstrated excellent prediction accuracy (AUC = 0.89). Three leading predictors were poorer quality of life ([QoL]; memory), abnormal lymphocyte count, and abnormal neutrophil count. The MCI model demonstrated a similarly high AUC (0.88). Five leading predictors were poorer QoL (memory, leisure), male sex, abnormal lymphocyte count, and poorer self-rated eyesight. The AD model demonstrated outstanding prediction accuracy (AUC = 0.98). Ten leading predictors were poorer QoL (memory), reduced olfaction, male sex, increased dependence in activities of daily living (n = 6), and poorer visual contrast. CONCLUSIONS: Both convergent and cohort-specific frailty factors discriminated the AD spectrum cohorts. Convergence was observed as all cohorts were marked by lower quality of life (memory), supporting recent research and clinical attention to subjective experiences of memory aging and their potentially broad ramifications. Diversity was displayed in that, of the 14 leading predictors extracted across models, 11 were selectively sensitive to one cohort. A morbidity intensity trend was indicated by an increasing number and diversity of predictors corresponding to clinical severity, especially in AD. Knowledge of differential deficit predictors across AD clinical cohorts may promote precision interventions.

Frailty inclusive care in acute and community-based settings: a systematic review protocol.

BACKGROUND: Frailty is a known risk factor for an array of adverse outcomes including more frequent and prolonged health services use and high health care costs. Aging of the population has implications for care provision across the care continuum, particularly for people living with frailty. Despite known risks associated with frailty, there has been limited research on care pathways that address the needs of persons living with frailty. Our study aims to review and examine, in a rigorous way, the quality of evidence for multi-component interventions and care pathways focused on frailty. METHODS: A comprehensive electronic search strategy will be used to identify studies that evaluate multi-component interventions or care pathways for persons living with frailty. The search strategy will include terms for frailty, multi-component interventions, effectiveness, and cost effectiveness applied to the following databases: MEDLINE (OVID), EMBASE (OVID), CINAHL (EBSCO), Cochrane Central Register of Controlled Trials (CENTRAL), and Cochrane Database of Systematic Reviews. An adapted search for Google Scholar and gray literature databases will also be used. References of included studies will be hand-searched for additional citations of frailty-inclusive care. Known experts and corresponding authors of identified articles will be contacted by email to identify further eligible studies. Risk of bias will be assessed using the Effective Public Health Practice Project Quality Assessment tool. Data will be extracted from eligible studies and it is anticipated that narrative analysis will be used. If studies with sufficient homogeneity are found, then pooled effects will be reported using meta-analysis. DISCUSSION: This review will appraise the evidence currently available on multi-component frailty interventions. Results will inform on clinical pathway development for people living with frailty across the care continuum and will guide future research to address gaps in the literature and areas in need of further development. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42020166733.

The Impact of Preoperative Frailty on the Clinical and Cost Outcomes of Adult Cardiac Surgery in Alberta, Canada: A Cohort Study.

BACKGROUND: There is limited information about the impact of frailty on public payer costs in cardiac surgery. This study aimed to determine quality-adjusted life-years (QALYs) and costs associated with preoperative frailty in patients referred for cardiac surgery. METHODS: We retrospectively compared costs of frailty in a cohort of 529 patients aged ≥ 50 years who were referred for nonemergent cardiac surgery in Alberta. Patients were screened preoperatively for frailty, defined as a score of 5 or greater on the Clinical Frailty Scale. The primary outcome measure was public payer costs attributable to frailty, calculated in a difference-in-difference (DID) model. RESULTS: The prevalence of frailty was 10% (n = 51; 95% confidence interval [CI], 7%-12%). Median (interquartile range) costs for frail patients were higher in the first year postsurgery ($200,709 [$146,177-$486,852] vs $147,730 [$100,674-$177,025]; P < 0.001) compared to nonfrail; the difference-in-difference attributable cost of frailty was $57,836 (95% CI, $-28,608-$144,280). At 1 year, frail patients had fewer QALYs realized compared to nonfrail patients (0.71 [0.57-0.77] vs 0.82 [0.75-0.86], P < 0.001), whereas QALYs gained were similar (0.02 [-0.02-0.05] vs 0.02 [0.00-0.04], P = 0.58, median difference 0.003 [95% CI, -0.01-0.02]) in frail and nonfrail patients. CONCLUSIONS: Frailty screening identified a population with greater impairment in quality-of-life and greater healthcare costs. Costs attributable to frailty represent opportunity costs that should be considered in future cardiac surgical services planning in the context of our aging population and the growing prevalence of frailty.

CONTEXTE: Il existe peu de renseignements concernant les répercussions de la fragilité sur les coûts pour les payeurs publics en chirurgie cardiaque. Cette étude visait à déterminer les années de vie pondérées par la qualité (QALY, pour Quality-Adjusted Life-Years) et les coûts associés à la fragilité préopératoire chez les patients dirigés vers un service de chirurgie cardiaque. MÉTHODOLOGIE: Nous avons comparé de façon rétrospective les coûts de la fragilité dans une cohorte de 529 patients âgés de 50 ans ou plus qui ont été dirigés vers un service de chirurgie cardiaque pour une intervention non urgente en Alberta. Un dépistage de la fragilité, définie comme un score de 5 ou plus à l'échelle CFS (Clinical Frailty Scale), a été effectué avant l'intervention. Le principal critère d'évaluation était le coût attribuable à la fragilité pour les payeurs publics, calculé selon un modèle d'écart des différences. RÉSULTATS: La prévalence de la fragilité a été de 10 % (n = 51; intervalle de confiance [IC] à 95 % : 7 à 12 %). Les coûts médians (écart interquartile) dans la première année suivant l'intervention chirurgicale ont été plus élevés chez les patients fragiles que chez les patients non fragiles (200 709 $ [146 177 $ à 486 852 $] contre 147 730 $ [100 674 $ à 177 025 $]; p < 0,001); le coût attribuable de la fragilité selon le modèle d'écart des différences a été de 57 836 $ (IC à 95 % : −28 608 $ à 144 280 $). À 1 an, les patients fragiles avaient moins de QALY réalisées que les patients non fragiles (0,71 [0,57 à 0,77] contre 0,82 [0,75 à 0,86]; p < 0,001), alors que le nombre de QALY gagnées était similaire (0,02 [−0,02 à 0,05] contre 0,02 [0,00 à 0,04]; p = 0,58; différence médiane : 0,003 [IC à 95 % : −0,01 à 0,02]) chez les patients fragiles et non fragiles. CONCLUSIONS: Le dépistage de la fragilité a permis de repérer une population associée à une perte plus importante de qualité de vie et à des coûts plus élevés en soins de santé. Les coûts attribuables à la fragilité représentent des coûts de renonciation qui doivent être considérés dans la planification future des services de chirurgie cardiaque, dans le contexte du vieillissement de notre population et de la prévalence croissante de fragilité.

Implementation of population-level screening for frailty among patients admitted to adult intensive care in Alberta, Canada.

PURPOSE: A substantial proportion of patients admitted to intensive care units (ICUs) are frail; however, the epidemiology of frailty has not been explored at a population-level. Following implementation of a validated frailty measure into a provincial ICU clinical information system, we describe the population-based prevalence and outcomes of frailty in patients admitted to ICUs. METHODS: Retrospective cohort study of adult admissions to 17 ICUs. Data were captured using eCritical Alberta. A Clinical Frailty Scale (CFS) score assigned at ICU admission was used to define the exposure (CFS score ≥ 5). Primary outcome was hospital mortality. Secondary outcomes were ICU and hospital stay, and receipt of organ support. RESULTS: Fifteen thousand two hundred and thirty-eight patients (81%) were assigned a CFS score at ICU admission. Of these, 28% (95% confidence interval [CI], 27 to 28) were frail. Prevalence of frailty was 9-43% across ICUs. Frail patients were older [mean (standard deviation) 63 (15) vs 56 (17) yr; P < 0.001], more likely to be male (54% vs 46% female; P < 0.001), and had higher APACHE II scores [22 (8) vs 17 (8); P < 0.001] compared with non-frail patients. Frail patients received less mechanical ventilation (62% vs 68%; P < 0.001) and vasoactive therapy (24% vs 57%; P < 0.001), but more non-invasive ventilation (22% vs 9%; P < 0.001). Frail patients had higher hospital mortality (23% vs 9%; adjusted odds ratio, 1.80; 95% CI, 1.64 to 2.05, along with longer ICU stay (median [interquartile range] 4 [2-8] vs 3 [2-6] days; P < 0.001), and longer hospital stay (16 [8-36] vs 10 [5-20] days; P < 0.001) compared with non-frail patients. CONCLUSION: A validated measure of frailty can be implemented at the population level in ICU. Frailty is common in ICU patients and has implications for health service use and clinical outcomes.

Frailty and the Association Between Long-Term Recovery After Intensive Care Unit Admission.

Frailty is common, although infrequently screened for among patients admitted to intensive care. Frailty has been the focus of research in geriatric medicine; however, its epidemiology and interaction with critical illness have only recently been studied. Instruments to screen for and measure frailty require refinement in intensive care settings. Frail critically ill patients are at higher risk of poor outcomes. Frail survivors of critical illness are high users of health resources. Further research is needed to understand how frailty assessment can inform decision-making before and during an episode of critical illness and during an intensive care course for frail patients.

Review of risk assessment tools to predict morbidity and mortality in elderly surgical patients.

BACKGROUND: Informed surgical consent requires accurate estimation of risks and benefits. Multiple risk assessment tools are available; however, most are not widely used or are specific to certain interventions. Assessing surgical risk is especially challenging in elderly patients because of their range of comorbidities, level of frailty, or severity of illness and a number of available surgical interventions. DATA SOURCES: We searched MEDLINE from January 2014 to July 2017 for studies that used risk assessment tools in studies on elderly surgical patients. We then sought the original articles describing each assessment tool and subsequent validation studies. CONCLUSIONS: We identified risk assessment tools that can improve surgical risk assessment in elderly surgical patients. The majority of the identified tools are not commonly used for pre-operative risk assessment. NSQIP-PMP, mFI and SURPAS are promising tools. Age is commonly used to predict risk, but frailty may be a more appropriate measure.

A prospective multicenter cohort study of frailty in younger critically ill patients.

BACKGROUND: Frailty is a multidimensional syndrome characterized by loss of physiologic and cognitive reserve that heightens vulnerability. Frailty has been well described among elderly patients (i.e., 65 years of age or older), but few studies have evaluated frailty in nonelderly patients with critical illness. We aimed to describe the prevalence, correlates, and outcomes associated with frailty among younger critically ill patients. METHODS: We conducted a prospective cohort study of 197 consecutive critically ill patients aged 50-64.9 years admitted to intensive care units (ICUs) at six hospitals across Alberta, Canada. Frailty was defined as a score ≥5 on the Clinical Frailty Scale before hospitalization. Multivariable analyses were used to evaluate factors independently associated with frailty before ICU admission and the independent association between frailty and outcome. RESULTS: In the 197 patients in the study, mean (SD) age was 58.5 (4.1) years, 37 % were female, 73 % had three or more comorbid illnesses, and 28 % (n = 55; 95 % CI 22-35) were frail. Factors independently associated with frailty included not being completely independent (adjusted OR [aOR] 4.4, 95 % CI 1.8-11.1), connective tissue disease (aOR 6.0, 95 % CI 2.1-17.0), and hospitalization within the preceding year (aOR 3.3, 95 % CI 1.3-8.1). There were no significant differences between frail and nonfrail patients in reason for admission, Acute Physiology and Chronic Health Evaluation II score, preference for life support, or treatment intensity. Younger frail patients did not have significantly longer (median [interquartile range]) hospital stay (26 [9-68] days vs. 19 [10-43] days; p = 0.4), but they had greater 1-year rehospitalization rates (61 % vs. 40 %; p = 0.02) and higher 1-year mortality (33 % vs. 20 %; adjusted HR 1.8, 95 % CI 1.0-3.3; p = 0.039). CONCLUSIONS: Prehospital frailty is common among younger critically ill patients, and in this study it was associated with higher rates of mortality at 1 year and with rehospitalization. Our data suggest that frailty should be considered in younger adults admitted to the ICU, not just in the elderly. Additional research is needed to further characterize frailty in younger critically ill patients, along with the ideal instruments for identification.

Long-term association between frailty and health-related quality of life among survivors of critical illness: a prospective multicenter cohort study.

OBJECTIVE: Frailty is a multidimensional syndrome characterized by loss of physiologic reserve that gives rise to vulnerability to poor outcomes. We aimed to examine the association between frailty and long-term health-related quality of life among survivors of critical illness. DESIGN: Prospective multicenter observational cohort study. SETTING: ICUs in six hospitals from across Alberta, Canada. PATIENTS: Four hundred twenty-one critically ill patients who were 50 years or older. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Frailty was operationalized by a score of more than 4 on the Clinical Frailty Scale. Health-related quality of life was measured by the EuroQol Health Questionnaire and Short-Form 12 Physical and Mental Component Scores at 6 and 12 months. Multiple logistic and linear regression with generalized estimating equations was used to explore the association between frailty and health-related quality of life. In total, frailty was diagnosed in 33% (95% CI, 28-38). Frail patients were older, had more comorbidities, and higher illness severity. EuroQol-visual analogue scale scores were lower for frail compared with not frail patients at 6 months (52.2 ± 22.5 vs 64.6 ± 19.4; p < 0.001) and 12 months (54.4 ± 23.1 vs 68.0 ± 17.8; p < 0.001). Frail patients reported greater problems with mobility (71% vs 45%; odds ratio, 3.1 [1.6-6.1]; p = 0.001), self-care (49% vs 15%; odds ratio, 5.8 [2.9-11.7]; p < 0.001), usual activities (80% vs 52%; odds ratio, 3.9 [1.8-8.2]; p < 0.001), pain/discomfort (68% vs 47%; odds ratio, 2.0 [1.1-3.8]; p = 0.03), and anxiety/depression (51% vs 27%; odds ratio, 2.8 [1.5-5.3]; p = 0.001) compared with not frail patients. Frail patients described lower health-related quality of life on both physical component score (34.7 ± 7.8 vs 37.8 ± 6.7; p = 0.012) and mental component score (33.8 ± 7.0 vs 38.6 ± 7.7; p < 0.001) at 12 months. CONCLUSIONS: Frail survivors of critical illness experienced greater impairment in health-related quality of life, functional dependence, and disability compared with those not frail. The systematic assessment of frailty may assist in better informing patients and families on the complexities of survivorship and recovery.

Association between frailty and short- and long-term outcomes among critically ill patients: a multicentre prospective cohort study.

BACKGROUND: Frailty is a multidimensional syndrome characterized by loss of physiologic and cognitive reserves that confers vulnerability to adverse outcomes. We determined the prevalence, correlates and outcomes associated with frailty among adults admitted to intensive care. METHODS: We prospectively enrolled 421 critically ill adults aged 50 or more at 6 hospitals across the province of Alberta. The primary exposure was frailty, defined by a score greater than 4 on the Clinical Frailty Scale. The primary outcome measure was in-hospital mortality. Secondary outcome measures included adverse events, 1-year mortality and quality of life. RESULTS: The prevalence of frailty was 32.8% (95% confidence interval [CI] 28.3%-37.5%). Frail patients were older, were more likely to be female, and had more comorbidities and greater functional dependence than those who were not frail. In-hospital mortality was higher among frail patients than among non-frail patients (32% v. 16%; adjusted odds ratio [OR] 1.81, 95% CI 1.09-3.01) and remained higher at 1 year (48% v. 25%; adjusted hazard ratio 1.82, 95% CI 1.28-2.60). Major adverse events were more common among frail patients (39% v. 29%; OR 1.54, 95% CI 1.01-2.37). Compared with nonfrail survivors, frail survivors were more likely to become functionally dependent (71% v. 52%; OR 2.25, 95% CI 1.03-4.89), had significantly lower quality of life and were more often readmitted to hospital (56% v. 39%; OR 1.98, 95% CI 1.22-3.23) in the 12 months following enrolment. INTERPRETATION: Frailty was common among critically ill adults aged 50 and older and identified a population at increased risk of adverse events, morbidity and mortality. Diagnosis of frailty could improve prognostication and identify a vulnerable population that might benefit from follow-up and intervention.

Frailty and outcome in elderly patients with acute coronary syndrome.

BACKGROUND: Frailty is superior to chronological age as a predictor of outcome. The Edmonton Frail Scale (EFS) is a simple valid measure of frailty, covering multiple important domains, with scores ranging from 0 (not frail) to 17 (very frail). The purpose of this pilot study was to assess the EFS in a group of elderly patients with acute coronary syndrome (ACS). METHODS: The EFS was administered to 183 consecutive patients with ACS aged ≥ 65 years admitted to a single centre in Edmonton, Alberta, Canada. RESULTS: Scores ranged from 0-13. Patients with higher EFS scores were older, with more comorbidities, longer lengths of stay (EFS 0-3: mean, 7.0 days; EFS 4-6: mean, 9.7 days; and EFS ≥ 7: mean, 12.7 days; P = 0.03), and decreased procedure use. Crude mortality rates at 1 year were 1.6% for EFS 0-3, 7.7% for EFS 4-6, and 12.7% for EFS ≥ 7 (P = 0.05). After adjusting for baseline risk differences using a "burden of illness" score, the hazard ratio for mortality for EFS ≥ 7 compared with EFS 0-3 was 3.49 (95% confidence interval [CI], 1.08-7.61; P = 0.002). CONCLUSIONS: The EFS is associated with increased comorbidity, longer lengths of stay, and decreased procedure use. After adjustment for burden of illness, the highest frailty category is independently associated with mortality in elderly patients with ACS. Further work is needed to determine whether the use of a validated frailty instrument would better delineate medical decision making in this important, often disadvantaged population.

An assessment of neurocognitive speed in relation to frailty.

OBJECTIVES: to evaluate the relationship between neurocognitive speed (NCS) and frailty; to consider how this relationship is affected by how frailty is operationalised. DESIGN: secondary analysis of the baseline cohort of the Oxford Project To Investigate Memory and Aging (OPTIMA), a longitudinal observational cohort. SUBJECTS: of 388 participants who underwent a comprehensive intake assessment followed by an annual follow-up for at least 3 years, data on all measures were available on 164 people. MEASUREMENTS: NCS was defined as a combined score of <18 on the pattern comparison test (<11 is abnormal) and letter comparison test (<7 is abnormal). Frailty was defined from a modified Phenotype model, the Edmonton Frailty Scales (EFS) and a frailty index (FI); the latter two were adapted here to exclude cognitive measures. RESULTS: in multivariate logistic (NCS as < or ≥18) and linear regression (NCS as continuous variable), only the FI (OR = 0.87) was significant (P < 0.05). When all frailty measures were included in the multivariate analysis only, FI (OR = 0.88) was significant (P < 0.05). Mini-mental Status Examination remained significantly related to NCS throughout all analysis. CONCLUSION: NCS slows with increasing frailty as shown with the FI.

The assessment of frailty in older people in acute care.

AIM: Develop a measure of frailty for older acute inpatients to be performed by non-geriatricians. METHOD: The Reported Edmonton Frail Scale (REFS) was adapted from the Edmonton Frail Scale for use with Australian acute inpatients. With acute patients aged over 70 years admitted to an Australian teaching hospital, we validated REFS against the Geriatrician's Clinical Impression of Frailty (GCIF), measures of cognition, comorbidity and function, and assessed inter-rater reliability. RESULTS: REFS was moderately correlated with GCIF (n = 105, R = 0.61, P < 0.01), Mini-Mental State Examination impairment (n = 61, R = 0.49, P < 0.001), Charlson Comorbidity Index (n = 59, R = 0.51, P < 0.001) and Katz Daily Living Scale (n = 59, R = 0.51, P < 0.001). Inter-rater reliability of REFS administered by two researchers without medical training was excellent (kappa = 0.84, n = 31). CONCLUSION: In this cohort of older acute inpatients, REFS is a valid, reliable test of frailty, and may be a valuable research tool to assess the impact of frailty on prognosis and response to therapy.

Frailty is associated with postoperative complications in older adults with medical problems.

We sought to test whether frailty may be predictive of operative risk in older adults with medical problems. One hundred and twenty-five patients at least 70 years of age had a previously developed frailty screen, the Edmonton Frail Scale (EFS), administered at a pre-surgical clinic, prior to elective non-cardiac surgery. A blinded chart audit assessed for postoperative medical complications, length of stay and inability to be discharged home. The mean age of patients was 77 (range 70-92) and most (82%) underwent orthopedic procedures. Increasing frailty was associated with postoperative complications (p=0.02), increased length of hospitalization (p=0.004) and inability to be discharged home (p=0.01), independent of age. EFS scores of 3 or less were associated with a lower risk of having a complication (age-adjusted OR 0.27, 95% CI 0.09-0.80, likelihood ratio of 0.33) and a higher chance (80%) of being discharged home (p<0.02). EFS scores exceeding 7 were associated with increased complications (OR 5.02, 95% CI 1.55-16.25, likelihood ratio of 3.9) and a lower chance of being discharged home (40%, p<0.02). This study suggests that a frailty screen can refine risk estimates of postoperative complications in older adults undergoing elective non-cardiac surgery.

Improving recruitment into geriatric medicine in Canada: Findings and recommendations from the geriatric recruitment issues study.

As the number of Canadians aged 65 and older continues to increase, declining recruitment into geriatric medicine (GM) raises concerns about the future viability of this medical subspecialty. To develop effective strategies to attract more GM trainees into the field, it is necessary to understand how medical students, residents, GM trainees, and specialists make career choices. The Geriatric Recruitment Issues Study (GRIST) was designed to assess specific methods that could be used to improve recruitment into geriatrics in Canada. Between November 2002 and January 2003, 530 participants were invited to complete the GRIST survey (117 Canadian geriatricians, 12 GM trainees, 96 internal medicine residents, and 305 senior medical students). Two hundred fifty-three surveys (47.7%) were completed and returned (from 54 participating geriatricians, 9 GM trainees, 50 internal medicine residents, and 140 senior medical students). The survey asked respondents to rate factors influencing their choice of medical career, the attractiveness of GM, and the anticipated effectiveness of potential recruitment strategies. Although feedback varied across the four groups on these issues, consistencies were observed between medical students and residents and between GM trainees and geriatricians. All groups agreed that role modeling was effective and that summer student research programs were an ineffective recruitment strategy. Based on the GRIST findings, this article proposes six recommendations for improving recruitment into Canadian geriatric medicine training programs.