Feasibility of the SmartPrompt for Improving Everyday Function in Dementia (SmartPrompt)

Difficulty completing everyday tasks is a primary reason for the high cost of care, loss of caregiver paid hours, and general caregiver burden associated with dementia. Electronic reminder applications hold promise as a low-cost solution to improve daily functioning, promote aging in place, and reduce caregiver burden and cost of care, particularly as older adults become more computer literate. There are many electronic reminders available for healthy individuals, but few have been developed to target the specific cognitive difficulties that impede completion of everyday tasks in people with dementia (i.e., premature decay of task goals, decreased motivation to perform tasks, distractibility, semantic knowledge degradation, etc.). Furthermore, there is a dearth of feasibility research on the fundamental efficacy and usability of reminder applications for people with dementia. This R21 proposal addresses these gaps with a feasibility study of the SmartPrompt, an enhanced electronic reminder aid designed for people with dementia that is used with an inexpensive smartphone. A diverse sample of older adults with mild dementia (N = 40) and their caregivers (N =40) will be trained to use the SmartPrompt and then asked to use the application to perform a target task (hydration, meals, or medication) twice per day in their homes for two weeks. Aim 1 will test the hypothesis that the SmartPrompt is effective at promoting everyday task completion (i.e., efficacy) relative to a one- week control period without the SmartPrompt. Using a single-group crossover design, efficacy outcomes will be obtained during the SmartPrompt and Control Conditions and will include participant and caregiver reports of task completion, caregiver report of burden, and participant report of frustration Aim 2 will investigate whether the SmartPrompt will be perceived favorably by participants and caregivers and the extent to which technical support is needed for its use (i.e., usability). Usability measures will be obtained from caregivers (report of technical problems, questionnaire), participants (questionnaire), the study team (training time, technical support required), and the smartphone (i.e., measures of smartphone use, response times to prompts). A third exploratory aim is to examine participant and caregiver features that are associated with efficacy and usability outcomes, including participant cognitive profile, participant/caregiver demographics, computer proficiency and self-efficacy, desire to change, etc. Results will be used to inform 1) a working model of barriers and facilitators for the use and efficacy of prompting applications that may be tested in future studies and 2) SmartPrompt design modifications for a future Phase II clinical trial.

The exorbitant health care costs associated with dementia relative to other illnesses are a significant concern in the US and other countries with aging populations. Difficulty completing everyday tasks is a primary reason for the high cost of care, loss of caregiver paid hours, and general caregiver burden. As older adults become more computer literate, assistive technologies hold promise as an inexpensive solution to improve daily functioning and ultimately promote aging in place, reduce caregiver burden, and reduce healthcare expenditures due to secondary health issues that arise when everyday tasks are neglected. Electronic reminder applications are a type of assistive technology that deliver cues to perform important daily activities (e.g., take medication). Electronic reminders are widely available, but, few of these applications have been designed specifically to 1) target the numerous cognitive difficulties that impede the recall and completion of everyday tasks in people with dementia (i.e., premature decay of task goals, decreased motivation to perform tasks, etc.); 2) circumvent the sensory/motor difficulties that older adults face when using technology (i.e., simple, user-friendly interface); and 3) be accessible at a low cost (i.e., easily integrated with consumer devices like smartphones). Additionally, there is a dearth of fundamental knowledge and quantitative data on the feasibility of electronic reminder applications for older adults with dementia. This revised R21 application (PAR-15-351/PAR-18-179) proposes a Stage I feasibility study that will address gaps in fundamental knowledge on the efficacy and usability of an electronic reminder aid for commercially available smartphones specifically designed for people with dementia. Aim 1 (efficacy): To investigate the hypothesis that a reminder application designed to target the specific cognitive impairments that preclude task completion in dementia and to circumvent sensory/motor challenges faced by many older adults (SmartPrompt) is efficacious for promoting everyday task completion. The SmartPrompt includes six features to address cognitive impairments that cause task failure in dementia: 1) reminder alert (prospective memory impairment); 2) choice to respond to reminder immediately or at a later time (distractibility); 3) instructions (semantic knowledge degradation); 4) persistent reminder until task is completed (poor monitoring); 5) activity logging via photos (episodic memory impairment); 6) gaming rewards for task completion (apathy). Diverse participants with mild dementia (N = 40) and their caregivers will be asked to track performance of a target task (medication, meals, or hydration) twice per day for two weeks. Using an AB/BA crossover design, participants will use the SmartPrompt for one of the two weeks (SmartPrompt Condition) and will perform the task as usual during the other week (Control condition; order counterbalanced across participants). Efficacy outcomes, including caregiver/participant daily logs of task completion and reports of caregiver burden and participant frustration, will be examined using repeated-measures analyses comparing the Control vs. SmartPrompt Conditions. Aim 2 (usability)- To evaluate the usability of the SmartPrompt using measures obtained from caregivers, participants with dementia, the study team, and the study smartphone. The investigators expect that the design features of the SmartPrompt will be perceived favorably by participants and will require minimal technical support from the study team and caregivers during the one-week SmartPrompt testing period. The investigators also expect that the structure and support embedded in the SmartPrompt application design will lead to favorable caregiver and participant usability ratings relative to published standards. All technical problems reported by the participant and/or caregiver will be logged and analyzed by the study team for future design modifications. Metrics obtained directly from the smartphone (e.g., response times to reminder alarms) also will be evaluated as objective measures of usability. Exploratory Aim: To examine participant and caregiver features that are associated with SmartPrompt efficacy/usability. Features of primary interest include participant sex, level of education, level and type of cognitive/functional impairment, caregiver and participant computer literacy/self-efficacy, and desire to change. The investigators will recruit a diverse older adult sample and administer measures of cognition, performance-based testing of everyday function, and questionnaires on mood/anxiety, computer literacy/proficiency/self-efficacy, and desire for behavior change. Bivariate correlations and multiple regression analyses will be performed with participant/caregiver features as predictors and efficacy/usability outcomes as the dependent variables. Results of this aim will be used to 1) develop a working framework of barriers and facilitators for the efficacy and usability of electronic reminder applications that will be tested in future studies and 2) inform modifications of the SmartPrompt training protocol and/or design for a future Phase II clinical trial. Considering the limited success of pharmaceutical treatments for dementia, there has been surprisingly little rigorous research on the potential to leverage relatively simple assistive technologies to improve daily functioning for people with dementia. Even relatively modest improvements in daily function could have a host of positive downstream effects, including increased autonomy and quality of life and improved health for people with dementia and reduced cost of care and reduced burden for caregivers.

Dementia, Dementia, Vascular, Dementia Alzheimers, Dementia, Mixed, Dementia, Mild, Dementia of Alzheimer Type, Mild Cognitive Impairment, Mild Dementia

The SmartPrompt is a smartphone application designed specifically for older adults with difficulties remembering to do daily tasks due to cognitive difficulties.

Caregiver Reported Efficacy Variable - caregiver reports participant's ability to complete each of the two target tasks using the following scale (0 - failed to complete task; 1- completed the task with assistance; 2- completed the task with some difficulty; 3- no difficulty). The total possible caregiver task completion score for each task during each week will range from 0 (unable to complete the task over 7 days) to 21 no difficulty completing the task over one week). A total caregiver task completion score for both tasks (range = 0 -42) will be calculated separately for the SmartPrompt Condition.

Caregiver Reported Efficacy Variable - caregiver reports participant's ability to complete each of the two target tasks using the following scale (0 - failed to complete task; 1- completed the task with assistance; 2- completed the task with some difficulty; 3- no difficulty). The total possible caregiver task completion score for each task during each week will range from 0 (unable to complete the task over 7 days) to 21 no difficulty completing the task over one week). A total caregiver task completion score for both tasks (range = 0 -42) will be calculated separately for the Control Condition.

Caregiver Reported Efficacy Variable - using a score from 0 (no burden) to 4 (severe burden), the caregiver reports the level of burden he/she experienced in helping the participant perform the two target tasks. The total possible caregiver task burden score will range from 0 - 28. A total caregiver task burden score will be calculated separately for the SmartPrompt Condition.

Caregiver Reported Efficacy Variable - using a score from 0 (no burden) to 4 (severe burden), the caregiver reports the level of burden he/she experienced in helping the participant perform the two target tasks. The total possible caregiver task burden score will range from 0 - 28. A total caregiver task burden score will be calculated separately for the Control Condition.

Participant Reported Efficacy Variable -Participant Reported Efficacy Variable - during the SmartPrompt condition participants will log a photo of their completed task (e.g., 0= task not performed or photo not logged; 1= task performed/photo logged). Failure to perform the task or log the photo will be scored as a 0. During Control condition, participants report on task completion using the same scale and a paper activity logbook. The total activity log score will range from 0 (unable to complete or log any task over one week) to 14 (completed/logged both tasks every day of the week). A total activity log score will be calculated separately for the SmartPrompt Condition.

Participant Reported Efficacy Variable -Participant Reported Efficacy Variable - during the SmartPrompt condition participants will log a photo of their completed task (e.g., 0= task not performed or photo not logged; 1= task performed/photo logged). Failure to perform the task or log the photo will be scored as a 0. During Control condition, participants report on task completion using the same scale and a paper activity logbook. The total activity log score will range from 0 (unable to complete or log any task over one week) to 14 (completed/logged both tasks every day of the week). A total activity log score will be calculated separately for the Control Condition.

Participant Reported Efficacy Variable -Using the paper log (Control Condition) or the SmartPrompt logging feature, participants will report on their level of frustration in completing each target task using a scale from 0 (not frustrated) to 2 (very frustrated).The total possible participant frustration score will range from 0 - 28 per week for both tasks. A total participant frustration score will be calculated separately for the SmartPrompt Condition.

Participant Reported Efficacy Variable -Using the paper log (Control Condition) or the SmartPrompt logging feature, participants will report on their level of frustration in completing each target task using a scale from 0 (not frustrated) to 2 (very frustrated).The total possible participant frustration score will range from 0 - 28 per week for both tasks. A total participant frustration score will be calculated separately for the Control Condition.

Caregiver Usability Measure - The C-SUS includes 10 items evaluated on a 1 (strongly disagree) - 5 (strongly agree) point scale. After correction (-1 for odd items and -5 for even, reverse-scored items), the total is multiplied by 2.5 to yield a total score ranging from 0 (poor usability) to 100 (excellent usability). The investigators also added a single seven-point adjective rating question regarding the usability of the SmartPrompt (1 = worst imaginable to 7 = best imaginable) at the end of the questionnaire to facilitate interpretation.

Caregiver Usability Measure - During the SmartPrompt testing period, the caregiver reports yes or no whether there has been any problem with the with the SmartPrompt system that day and may record a memo detailing the problem. Responses will be coded by the study team as 0 (no problem), 1 (problem affecting only 1 task), or 2 (problem affecting both tasks). A total score ranging from 0 to 28 (problem reported everyday with both tasks) as well as problem descriptions will be summed for the week.

Usability Measure from the Study Team- The study team will keep a Technical Support Log of each caregiver/participant's help request. Each problem and the number of communications over the one-week test period will be tallied.

Usability Measure from the Study Team - the total Training Time is a measure of the length of the SmartPrompt training session, number of times the training quiz/checklist needed to be repeated during the initial training session, and whether the caregiver required additional training during the one-day phone follow-up session. The total Training Time (initial training session time + additional training time by phone) will be summed for each participant-caregiver pair. The investigators expect this variable will range from approximately 30 - 120 minutes.

Usability Measure from Smartphone - computed from the accelerator, gyroscope, and other sensor data, as an estimate of the amount of time the smartphone was carried during the day (values will remain static if the phone was not used; higher value = more use).

Usability Measure from Smartphone - Response Time (RT) to all reminder alarms and subsequent prompts (shorter RT = greater engagement).

Participant Usability Measure- participant is asked open-ended questions about the SmartPrompt design and usability.

Caregiver Usability Measure- caregiver is asked open-ended questions about the SmartPrompt design and usability.

The study sample will include older adults with mild dementia (N =40) and their caregivers (N=40). Inclusion/exclusion criteria for participants with dementia are as follows: at least 65 years old; fluency in the English language; diagnosis of mild dementia regardless of suspected etiology consistent with McKhann et al. (2011) criteria a for "all-cause dementia" ; mild level cognitive impairment (i.e., Mini Mental Status Exam > 19/30); availability of a caregiver (see inclusion criteria for caregiver below); no lifetime history of severe psychiatric disorder (e.g., schizophrenia, bipolar disorder), nervous system infections or disorders (e.g., epilepsy, brain tumor, large-vessel stroke, major head trauma); no major intellectual disability; no current metabolic or systemic disorders (e.g., B12 deficiency, renal failure, cancer); no severe sensory deficits that would preclude viewing a smartphone or other everyday objects or the inability to hear the task directions (e.g., blindness, total hearing loss); and no severe motor weakness that would preclude the use of a smartphone or other everyday objects (e.g., severe deformities or paralysis of both upper extremities) The inclusion/exclusion criteria for caregivers are as follows: age 18 years or older; fluency in the English language; resides with or has daily contact with the participant with dementia; and cognitively healthy with no diagnosis of dementia or other major health problems that would preclude reliable reporting and participation in the study.