The aim of this trial is to evaluate the effect of SENATOR software on incident, adverse drug reactions (ADRs) in older, multimorbid, hospitalized patients. The SENATOR software produces a report designed to optimize older patients’ current prescriptions by applying the published STOPP and START criteria, highlighting drug-drug and drug-disease interactions and providing non-pharmacological recommendations aimed at reducing the risk of incident delirium. We will conduct a multinational, pragmatic, parallel arm Prospective Randomized Open-label, Blinded Endpoint (PROBE) controlled trial. Patients with acute illnesses are screened for recruitment within 48 h of arrival to hospital and enrolled if they meet the relevant entry criteria. Participants’ medical history, current prescriptions, select laboratory tests, electrocardiogram, cognitive status and functional status are collected and entered into a dedicated trial database. Patients are individually randomized with equal allocation ratio. Randomization is stratified by site and medical versus surgical admission, and uses random block sizes. Patients randomized to either arm receive standard routine pharmaceutical clinical care as it exists in each site. Additionally, in the intervention arm an individualized SENATOR-generated medication advice report based on the participant’s clinical and medication data is placed in their medical record and a senior medical staff member is requested to review it and adopt any of its recommendations that they judge appropriate. The trial’s primary outcome is the proportion of patients experiencing at least one adjudicated probable or certain, non-trivial ADR, during the index hospitalization, assessed at 14 days post-randomization or at index hospital discharge if it occurs earlier. Potential ADRs are identified retrospectively by the site researchers who complete a Potential Endpoint Form (one per type of event) that is adjudicated by a blinded, expert committee. All occurrences of 12 pre-specified events, which represent the majority of ADRs, are reported to the committee along with other suspected ADRs. Participants are followed up 12 (+/− 4) weeks post-index hospital discharge to assess medication quality and healthcare utilization. This is the first clinical trial to examine the effectiveness of a software intervention on incident ADRs and associated healthcare costs during hospitalization in older people with multi-morbidity and polypharmacy. Clinicaltrials.gov NCT02097654 , 27 March 2014.
Early identification of frailty can prevent functional decline. Although multiple frailty screens exist for use in Emergency Departments (EDs), few are validated against diagnostic standards such as comprehensive geriatric assessment. To examine the diagnostic accuracy of ED screens for frailty, scientific databases were searched for prospective diagnostic accuracy test studies from January 2000 to September 2022. Studies were assessed for risk of bias using QUADAS-C. Psychometric properties were extracted and analysed using R. Six studies involving 1,663 participants describing seven frailty screening instruments (PRISMA-7, CFS, VIP, FRESH, BPQ, TRST, and ISAR), representing 13 unique data points, were included. The mean age of participants ranged from 76 to 86 years. The proportion that was female ranged from 45 to 60%. The pooled prevalence rate of frailty was high at 59%. The pooled estimate for sensitivity was 0.85 (95% CI: 0.76–0.91) versus 0.77 (95% CI: 0.62–0.88) for specificity. Pooled accuracy based on area under the ROC curve was 0.89 (95% CI: 0.86–0.90). Although few studies were found, limiting the ability to conduct a meta-analysis of individual instruments, available frailty screens can accurately diagnose frailty in older adults attending the ED. As specificity was comparatively low, additional assessment may be required to identify those requiring inpatient management or onward community referral. Further study is therefore required.
Sleep disturbance is common among people with cognitive impairment and, when present, is an important target for intervention because it potentially leads to negative outcomes and cognitive decline. Given this association, sleep represents a potential public health target, although evidence for efficacy is lacking. For this study, a systematic review and meta-analysis was undertaken of controlled clinical trials of pharmacological and non-pharmacological interventions to improve sleep in mild cognitive impairment and dementia. A total of 144 trials involving 13,471 participants (median 50 per trial) were included. To measure sleep, 68 trials used subjective measures exclusively, and 41 used only objective measures, while 35 used a combination. In all, 287 discrete sleep outcome measures were reported. Bright light therapy was the most frequently examined non-pharmacological intervention, but results were equivocal. Other non-pharmacological interventions (such as physical activity, cognitive behavioural therapy for insomnia, music, and continuous positive airway pressure) showed promise but require further evidence. Results for melatonin, the most frequently examined pharmacological intervention, were inconclusive, but lower doses may be more effective. Other pharmacological interventions (such as trazadone and orexin-receptor antagonists) demonstrated effectiveness in a small number of trials but require further evidence. Overall, there is insufficient evidence upon which to base clinical decisions regarding the treatment of sleep disturbance in this population. Existing research is marked by wide heterogeneity, which limits opportunities for data synthesis. A core outcome set is urgently required to ensure that future research provides more coherent and reliable evidence to improve outcomes for people with cognitive impairment and sleep disturbance.
Background: Prompt and efficient identification and stratification of patients who are frail is important, as this cohort are at high risk of adverse healthcare outcomes. Numerous frailty screening tools have been developed to support their identification across different settings, yet relatively few have emerged for use in emergency departments (EDs). This protocol provides details for a systematic review aiming to synthesize the accumulated evidence regarding the diagnostic accuracy and clinimetric properties of frailty screening instruments to identify frail older adults in EDs. Methods: Six electronic databases will be searched from January 2000 to March 2021. Eligible studies will include adults aged ≥60 years screened in EDs with any available screening instrument to identify frailty (even if not originally designed for this purpose). Studies, including case-control, longitudinal, and cohort studies, will be included, where instruments are compared to a reference standard to explore diagnostic accuracy. Predictive accuracy for a selection of outcomes, including mortality, institutionalization, and readmission, will be assessed. Clinical and methodological heterogeneity will be examined, and a random effects meta-analysis performed if appropriate. Conclusion: Understanding whether frailty screening on presentation to EDs is accurate in identifying frailty, and predicting these outcomes is important for decision-making and targeting appropriate management.
Dementia is a leading and growing cause of disability worldwide. Insomnia is highly prevalent in mild cognitive impairment (MCI) and dementia and is associated with impaired cognition, functional decline, and reduced quality of life (QoL). Although sleep disturbance represents a potentially modifiable risk factor within the trajectory of cognitive impairment, there have not been many studies conducted to examine the feasibility or preliminary efficacy of digital cognitive behavioural therapy for insomnia (dCBT-I) in this population. The aim of this pilot randomised controlled feasibility study is to evaluate the acceptability, adherence, and potential clinical effects of a multi-component dCBT-I programme (Sleepio) in adults with MCI or mild dementia and comorbid insomnia. Thirty community-dwelling adults aged ≥50 years with established MCI or mild dementia (Mini-Mental State Examination ≥18; Clinical Dementia Rating 0.5–1.0) and insomnia (Sleep Condition Indicator ≤16) will be randomised (1:1) to Sleepio or a wait-list control. Feasibility outcomes include recruitment and retention rates, intervention adherence (completion of ≥4/6 sessions), and acceptability measured using validated usability scales. Secondary outcomes include changes in sleep, mood, QoL, cognition, and function over 10 weeks. Adverse events will be monitored to assess safety. Findings will inform the design of a future definitive trial evaluating digital sleep interventions in cognitively impaired populations. Ethical approval has been granted. The trial is registered at ClinicalTrials.gov (NCT07363928).
Insomnia and dementia are major public health issues. Although cognitive behavioural therapy for insomnia (CBT-I) is the recommended first line treatment for insomnia, few studies have examined its usability or effectiveness in people with cognitive impairment. Digital CBT-I (dCBT-I) is an effective and easily accessible alternative to CBT-I. We undertook a mixed-methods analysis of the usability and acceptability of dCBT-I (SleepioTM) in patients attending a memory clinic. Participants were community dwellers aged ≥50 years with insomnia (Sleep Condition Indicator score ≤ 16) and subjective cognitive decline (SCD), mild cognitive impairment (MCI) or dementia. The Mobile Device Proficiency Questionnaire (MDPQ) assessed digital literacy. The System Usability Score (SUS) and Usability Metric for User Experience–Lite (UMUX-Lite) assessed usability. Acceptability was assessed by analysing adherence to the intervention and subjective reports of perceived benefit. Both usability and acceptability were further explored in semi-structured interviews. Among 24 participants (median age 69; 17 females), 2 had SCD, 14 MCI, and 8 dementia. Overall response to SleepioTM was positive. Participants favoured relaxation-based aspects of the programme. Behavioural elements, such as sleep restriction, were generally less favoured. Lack of digital literacy was the main barrier to engagement, though this could be surmounted by assistance from caregivers or family members in certain instances. Regardless of cognitive status, those with above-average digital literacy (MDPQ > 20) undertook a median 6/6 sessions, while those with below-average digital literacy (MDPQ < 20) undertook a median 3/6 sessions. Among those with above-average digital literacy, median SUS and UMUX-Lite were 72.5 and 75, respectively, representing acceptable usability, while overall scores were 63.75 and 56.25, respectively, representing marginally acceptable usability. Our findings may suggest that SleepioTM is usable and acceptable for people with SCD, MCI and selected individuals with mild dementia, particularly where digital literacy is preserved. Additional caregiver support may facilitate usability among people with MCI and lower digital literacy. Usability appears limited in more advanced dementia. However, as some subgroups were small, more research is required to confirm these findings.