Feasibility of the Glenrose Grocery Game for Cognitive and Functional Outcomes in Older Inpatients

Occupational therapists at the Glenrose Rehabilitation Hospital (GRH) developed the Glenrose Grocery Game (GGG)-a serious game designed to simulate shopping, meal planning, and budgeting in a virtual environment. However, the GGG had not yet been tested for its impact on cognitive and occupational outcomes.

The purpose of this feasibility study was to explore the implementation and effects of the GGG as part of the rehabilitation program at GRH, assessing its impact on specific mental functions, occupational performance, and satisfaction with performance in older adults. The study also aimed to estimate the parameters needed to design a future randomized controlled trial (RCT).

Research Questions

Was the designed protocol feasible for conducting a future definitive RCT?

Did playing the GGG as part of the rehabilitation program at GRH affect mental functions, occupational performance, and satisfaction in older adults, compared to standard care?

How engaged were older adults while playing the GGG?

Methods Design A pretest-posttest control group design was used. The experimental group received sessions with the GGG, while the control group received standard care (i.e., compensatory strategies). The study followed the CONSORT guidelines for randomized pilot studies (Eldridge et al., 2016).

Participants

Inclusion Criteria:

Adults aged 50 or older

With or without cognitive impairment (MOCA ≥19)

Functional vision and hearing (with or without aids)

Functional upper extremity use

Required no more than one-person assist for mobility

Exclusion Criteria:

Diagnosed with influenza or a virus affecting cognitive performance

Unable to count, speak, or follow simple English instructions

Sample Size Based on power analysis (α = 0.05, power = 0.8, effect size = 1.2), a minimum of 32 participants was required. We aimed to recruit 16 participants per group to account for a 25% dropout rate.

Sampling and Allocation Stratified permuted block randomization was used, with stratification based on MOCA scores. Participants with MOCA scores between 19-24 were categorized as having moderate cognitive impairment; those with scores of 25-30 were considered to have mild/no impairment. Allocation was blinded to research staff administering and evaluating the intervention.

Recruitment Protocol OTs at GRH inpatient units identified eligible participants. Informed consent (or assent via substitute decision-makers) was obtained. Allocation and coding were conducted by the PI (AMRR), and group assignment remained blinded to those delivering and evaluating the intervention.

Independent Variable

GGG Intervention: Participants played the GGG, completing meal planning and grocery shopping tasks in a gamified format. Sessions (8 × 30 minutes over 4 weeks) were delivered by trained RAs using tablets or computers.

Standard Care: Included cognitive assessment, education, and compensatory strategies (e.g., journaling, memory logs). Interventions were tailored to comprehension level and included caregiver involvement when needed.

Dependent Variables

Specific Mental Functions:

Trail Making Test (TMT) Parts A and B

Walking and Remembering Test (WART)

Montreal Cognitive Assessment (MOCA)

Occupational Performance:

Objective: PASS Task H8 (Money Management: Shopping)

Subjective: Canadian Occupational Performance Measure (COPM)

Engagement:

8-point Likert engagement scale developed in a prior study (Ríos Rincón et al., 2017)

All tools have strong metric properties and are suitable for use with older adults.

Data Collection The experimental group received GGG sessions in a dedicated room at GRH. Engagement was assessed post-session. All participants completed pretest and posttest assessments, taking ~80 minutes total. Blinded evaluators (RA 2) conducted assessments; a second rater (RA 3) scored at least 25% of measures to ensure reliability.

Data Analysis Data are currently being analyzed using SPSS under intention-to-treat principles. Descriptive statistics are being used to characterize the groups at pretest. Due to the small sample size, non-parametric tests (Wilcoxon signed-rank and Mann-Whitney U) are being used for within- and between-group comparisons for MOCA, TMT, WART, PASS, and COPM scores. Clinical significance in COPM scores is being evaluated using the two-point threshold (Law et al., 1998). Engagement data are being analyzed descriptively.

Potential Impact Findings from this study will provide clinicians with preliminary evidence regarding the GGG's impact on cognition, occupational performance, and engagement in older adults. If results are favorable, they will support the implementation of the GGG as a rehabilitation tool at GRH, benefiting inpatient clients. Insights will also help guide discharge planning and caregiver support and inform the design of a future, larger-scale RCT.