**Background:** The COVID-19 pandemic led to strict visitation bans in residential care facilities, increasing social isolation, loneliness, and mental health deterioration among institutionalized older adults. Prior evidence indicates that reminiscence therapy can improve cognitive function, mood, and social engagement in older populations. The authors developed the 'Geriatric Proximity' intervention, grounded in person-centered care and reminiscence principles, to address these pandemic-induced deficits.
**Methods:** This pilot, quantitative, quasi-experimental study enrolled 34 institutionalized older adults from two Portuguese residential homes. Institution A (n=17) served as the experimental group (EG), and institution B (n=17) as the control group (CG), assigned by paper draw. The EG was predominantly male (64.7%), age range 66–94 years; the CG was predominantly female (64.7%), age range 68–97 years. Mean ages did not differ significantly (p=0.463). Widowed status predominated (EG 64.7%, CG 58.8%). The intervention comprised 10 group sessions (max 45 min each) led by the same researcher-nurse and an occupational therapist. Themes (e.g., marriage, children, occupational life, school, trips, dreams, birthdays, childhood games) were chosen by participants. Sessions used a tablet for recording and displaying images. The EG was divided into three subgroups (two of 5, one of 7 participants). Data were collected at three time points: baseline (May 2021), midpoint (23 June 2021), and post-intervention (16 July 2021). Instruments included: Social Support Satisfaction Scale (ESSS, range 15–75, higher=better), Elderly Nursing Core Set (ENCS, 25 items, higher=worse functioning), Positive and Negative Affect Schedule—Reduced Version (PANAS-VRP, 10 items), and UCLA Loneliness Scale (18 items, range 18–72, higher=worse). Statistical analysis used t-tests, Wilcoxon Mann–Whitney, mixed ANOVA with Bonferroni correction, and robust ANOVA where assumptions were violated (α=0.05).
**Key Results:** At baseline, significant differences existed between groups on all scales (all p<0.005), with the CG showing better functional profile, lower loneliness, lower negative affect, higher social support satisfaction, and higher positive affect. By the second measurement, no significant differences remained between groups for social support satisfaction or loneliness (both p>0.05). At the third measurement, only general functional profile and negative affect differed significantly (both p<0.05). In the EG, from first to third measurement, there was a significant reduction in general functional profile (improvement), loneliness scale scores, and negative affect, and a significant increase in satisfaction with general support and positive affect (all p<0.05). Within the ENCS domains, only 'learning of mental functions' showed significant improvement over time in the EG (p=0.008); communication and self-care domains showed no significant change. The CG showed no significant differences across any measurement time points.
**Clinical Implications:** The 'Geriatric Proximity' intervention demonstrated positive effects on loneliness, affectivity, social support satisfaction, and cognitive function in institutionalized older adults during pandemic restrictions. The improvement in learning/mental functions aligns with prior reminiscence therapy research showing cognitive benefits. Enhanced social support satisfaction likely stemmed from group cohesion and shared reminiscence. The reduction in negative affect and loneliness suggests the intervention addressed key mental health risks of isolation. The CG's stability indicates that observed changes were attributable to the intervention rather than time or usual care. Limitations include the small sample (n=34), significant baseline differences between groups, non-randomized allocation, pandemic-related disruptions (including an outbreak that postponed the study), and inability to involve family or multidisciplinary team members. The authors note that self-care and communication domains did not improve, suggesting additional resources and structural changes are needed. Future research should examine durability of effects over time.