Automated IADL Sensing to Refine Measurement of Older Adult Daily Activity

University of Pittsburgh20 enrolled

Overview

The instrumental activities of daily living (IADL) refer to complex daily activities required for adult independence, such as preparing a meal or taking medications. This study will assess the efficacy of sensing technologies (smartwatch, computer vision, eye tracking) for recognizing IADL activities in naturalistic settings and score performance relative to ratings from occupational therapists. If successful in assessing the efficiency of IADL, the sensing technologies will be a valuable addition to geriatric assessment.

With declines in motor and cognitive function, even older adults living independently may be less efficient in performing daily activities, such as cooking and light housekeeping, which may signal an impending need for caregiver support and healthcare services. Clinicians currently lack automated tools for detecting early declines in daily activity. This research will assess the utility of motion sensors and computer vision assessment in detecting early deficits in the instrumental activities of daily living (IADL), in this case structured cooking and cleaning tasks performed in a standardized kitchen. Older adults with normal cognitive status and those with mild cognitive impairment will be recruited from a research registry to assess differences in performance.

Study Type

OBSERVATIONAL

Enrollment

Conditions

Cognitive Ability General Functional Abilities

Eligibility

Sex: FEMALEMin age: 75 YearsHealthy volunteers:

Medical Language ↔ Plain English

Inclusion Criteria: 1. Participant in University of Pittsburgh Pepper Center research registry 2. Age 75+ 3. Female 4. Residing in community 5. Meets criteria for normal cognition or mild cognitive impairment on telephone screening (Memory Impairment Screen \>= 5) 6. Daily, independent performance of cooking and light cleaning tasks Exclusion Criteria: 1. Inability to provide informed consent 2. Meets criteria for possible dementia (Memory Impairment Screen \<= 4) 3. Reports difficulty with activities of daily living (dressing, feeding oneself, using toilet, bathing) 4. Uses mobility assistance device for indoor ambulation 5. Medical conditions that may interfere with participation (e.g., severe untreated psychiatric disorders, unstable cardiovascular conditions, Parkinson's disease) 6. Current participation in other interventional studies or clinical trials.

Outcomes

Primary Outcomes

Efficiency of IADL Performance

Machine learning composite based on candidate sensing metrics, such as time to complete each element of kitchen task, pacing of activity, corrections, repetition of movement, adjustments of posture, and need to review directions.

Time frame: 15-minute telephone screening, 90-minute in-person assessment

Secondary Outcomes

Concordance with Occupational Therapist Rating

Agreement between machine-learning sensor categorization and occupational therapist assessment of standardized kitchen tasks.

Time frame: One 90-min in-person assessment