In an investigation evaluating how well outside observers understand the daily experiences of older persons living in long-term care facilities, staff proxy assessments aligned more closely with resident self-ratings of thriving than evaluations completed by relatives. The cross-sectional investigation, conducted across public care homes in a medium-sized Finnish city, analyzed responses using the 15-item short-form Thriving of Older People Assessment Scale to evaluate subjective, place-related well-being. Across pairwise comparisons, quadratic weighted kappa coefficients for resident versus staff ratings ranged from 0.22 to 0.64, spanning fair to substantial agreement. Conversely, agreement between residents and their relatives registered substantially lower, yielding coefficients from 0.00 to 0.50, which encompassed poor to moderate agreement. The findings demonstrate that while proxy reporting offers essential insights when cognitive impairment limits resident self-expression, different proxy groups provide distinctly divergent perspectives on the quality of life and environmental contentment experienced by older residents.
The concept of thriving within aged care research is defined as place-related well-being that emerges through the dynamic interaction between an individual and their physical, social, and care environment. Although quality of life generally captures broad well-being and health status, thriving emphasizes situated contentment, positive relationships, and the experience of settling into long-term residential surroundings. Because cognitive impairments affect a large share of individuals residing in care facilities, direct self-assessment is frequently complicated by difficulties recalling daily care events or articulating personal preferences. Under such circumstances, proxy respondents represent an indispensable, albeit secondary, method for gathering observational data. However, previous literature has shown that proxy evaluations may introduce systematic bias depending on the rater’s relationship with the resident, the frequency and continuity of their contact, and whether raters adopt a proxy-patient perspective or impose their own external standards.
To rigorously quantify these observational discrepancies, the researchers examined a sample drawn from all four public municipal care homes in a Finnish city of approximately 50,000 residents. Across these facilities, which offered 24-hour staffing and single-occupancy rooms with private bathrooms, 237 residents were initially invited to take part. Ultimately, 188 residents participated alongside 79 relatives and 145 care staff members, yielding 47 complete resident-relative-staff triads. The participating residents had a mean age of 85.53 years and an average stay of 29 months, with women representing 77 percent of the cohort. Standardized assessments from the statutory interRAI Long-Term Care Facilities system characterized the high care needs of the sample: 84.9 percent exhibited functional dependency with Activities of Daily Living Hierarchy scores exceeding 1, and 83.6 percent recorded Cognitive Performance Scale scores greater than 2, reflecting widespread cognitive impairment among the resident population.
The data collection, conducted between October and November 2022, administered the short-form TOPAS instrument, which evaluates 15 distinct statements regarding the care environment, daily activities, staff interactions, and personal autonomy on a 6-point Likert-type scale ranging from 1, representing complete disagreement, to 6, indicating complete agreement. Total scale scores can theoretically span from 15 to 90 points. Because the scale was originally developed and validated in Sweden, where staff proxy assessments demonstrated high internal consistency with a Cronbach’s alpha of 0.90, the research team applied a forward-backward translation method to produce a Finnish version suited to the bilingual municipal context. Both staff and relatives were explicitly directed to provide ratings from a proxy-resident perspective by answering each questionnaire item as they believed the residents themselves would respond, ensuring that the comparative analyses examined intended resident viewpoint concordance rather than external rater preferences.
At the item level, residents reported an overall mean score of 4.5 across the 15 items, registering their highest average agreement on item 5 regarding staff kindness at 5.5 and item 7 regarding staff competence at 5.4. Residents recorded their lowest average score of 3.6 on item 11, which addressed having interesting things to do during the day. Staff gave comparable responses, generating an overall item-level mean of 4.3, with their highest mean rating of 5.4 assigned to item 4 regarding adjusted care assistance and their lowest mean rating of 3.6 assigned to item 13 concerning opportunities to visit important external places. Discrepancies between resident and staff item means remained narrow, ranging between 0 and 0.4 points. In contrast, relatives yielded an overall item-level mean of 4.2, systematically rating resident experiences lower than residents themselves on 13 of the 15 scale items. The lowest ratings among relatives appeared for item 11 and item 13, both averaging 3.2.
Linear mixed-effects models testing equality across raters revealed statistically significant mean differences for three specific domains: having an enjoyable time with fellow residents, possibilities for outdoor access according to resident wishes, and opportunities to visit important places outside the home. The largest raw score disparities between residents and relatives emerged on outdoor access and community outings, where resident scores exceeded relative ratings by 0.8 points, as well as peer socialization, which showed a 0.5-point gap. Similar differences separated relatives from staff, who rated outdoor access 0.8 points higher and daytime activities 0.5 points higher than family members did. These patterns underscored a pervasive tendency among staff to record more favorable conditions than relatives, while relatives consistently assumed residents experienced greater deprivation in recreation, outdoor access, and social contact than either the residents or their direct daily caregivers reported.
When assessing categorical concordance through quadratically weighted kappa statistics, which accounted for ordinal distance and incorporated cluster-robust bootstrap confidence intervals to address nesting within care homes, staff ratings mirrored resident evaluations far more reliably than family evaluations did. The highest pairwise agreement between residents and staff reached a substantial kappa of 0.637 for item 9, which measured having an enjoyable time with fellow residents. Moderate agreement between residents and staff was also observed across several core items, including participating in enjoyable activities at 0.528, deciding how items are arranged in one’s room at 0.524, daytime engagement at 0.508, experiencing housing as a home at 0.423, and viewing the facility as the best possible living place at 0.417. The lowest agreement between residents and staff appeared on items concerning direct caregiver relationships, dropping to fair levels of 0.216 for staff kindness and 0.227 for having desired relationships with staff.
Agreement between residents and their relatives proved much weaker across almost all functional and environmental measures. The only domains achieving moderate agreement between resident self-reports and relative proxy reports were item 8, addressing relationships with staff at a kappa of 0.497, and item 3, capturing efforts to see positive aspects of living in the facility at 0.442. For five questionnaire items, resident-relative agreement fell to slight levels between 0.003 and 0.204, specifically encompassing adjusted assistance, staff effort, peer interaction, daytime activities, and community visits. On item 11, which evaluated daytime activities, agreement collapsed entirely to a kappa of 0.003, accompanied by a 95 percent confidence interval spanning from negative 0.107 to positive 0.113. This striking divergence occurred despite nearly identical mean scores on adjusted care, illustrating that mean comparisons alone conceal pronounced individual dispersion and fundamental rater discordance.
Pairwise concordance between relatives and staff generated the lowest overall coefficients, where no item reached moderate agreement. The highest agreement between relatives and staff occurred on room autonomy at a kappa of 0.399, activity participation at 0.369, staff kindness at 0.315, and maintaining outside contact at 0.308, all within the fair agreement category. The remaining eight items yielded slight agreement ranging between 0.040 and 0.186, including a kappa of 0.040 on staff relationships and 0.075 on staff effort. Furthermore, several confidence intervals for comparisons involving relatives crossed zero, indicating instances where agreement did not exceed chance. These wide confidence intervals reflected both the subjective complexity of the items and the smaller relative sample size, which increased statistical uncertainty and highlighted the differing observational prisms through which family members and professional caregivers interpret care quality.
The authors contextualized these divergent ratings by considering the relational perspectives, psychological adjustments, and operational realities governing each rater group. Staff members interact continuously with residents across daily shifts, allowing them to witness spontaneous peer engagement, institutional activities, and personal adaptations that visiting family members rarely observe. This routine familiarity likely explains why staff evaluations of social life and activity participation aligned substantially with resident self-reports. However, the relatively low agreement between staff and residents regarding caregiver kindness and competence may reflect subtle rating inflation or social desirability bias among staff assessing their own workplace, alongside lingering resident hesitancy to express dissatisfaction regarding individuals responsible for their daily assistance.
Conversely, relatives often evaluate a resident’s circumstances by comparing their present institutional existence to their previous independent life prior to the onset of severe physical and cognitive decline. This frame of reference frequently magnifies feelings of grief, loss, and caregiver burden, leading family members to perceive living conditions as poorer and less fulfilling than the residents themselves report. The research team also noted significant external constraints during the study window, including pandemic-related visiting limitations that restricted family access to residential units and may have reduced relatives’ insight into everyday care routines. Coupled with a 33.3 percent relative response rate that introduces potential non-response bias, these factors emphasize that relatives’ assessments must be understood as subjective approximations shaped by emotional distance and limited direct observation.
The study authors emphasized methodological boundaries that necessitate cautious interpretation, noting that the cross-sectional design within four municipal care homes restricts broad generalizability, while full psychometric validation of the Finnish version of the short-form TOPAS remains ongoing. Nevertheless, the findings offer vital implications for aged care management and clinical practice. Rather than treating proxy data as interchangeable or assuming that family members inherently reflect resident perspectives better than institutional caregivers, long-term care organizations must recognize resident, staff, and relative assessments as complementary channels of information. Establishing structured communication forums where staff and relatives systematically discuss resident thriving can bridge observational disparities, ensuring that care planning honors resident autonomy, mitigates social isolation, and fosters meaningful everyday well-being for vulnerable older adults.
Subject of Research: Interrater agreement on older persons' thriving between care home residents, relatives, and staff
Article Title: Interrater Agreement Between Resident‐Self Assessments and Proxy Rated Assessments of Older Persons' Experience of Thriving in Care Homes
Article References: Silvennoinen, A.-S., Johansson, E., Baxter, R., Finne‐Soveri, H., Bergland, Å., Edvardsson, D., & Fagerström, L. (2026). Interrater Agreement Between Resident‐Self Assessments and Proxy Rated Assessments of Older Persons' Experience of Thriving in Care Homes. Nursing Open, 13(11), Article e70929. https://doi.org/10.1002/nop2.70929
Image Credits: AI Generated
DOI: 10.1002/nop2.70929
Keywords: thriving in care homes, older persons, long-term care facilities, TOPAS assessment scale, interrater agreement, proxy assessments, nursing care, Finnish care homes, cognitive impairment, interRAI scales, Cohen's weighted kappa, resident self-assessment, relative proxy ratings, staff proxy ratings, quality of life in aged care, social engagement
Cite Scienmag News
Ophelia Keating. (November 20, 2026). Care Home Staff Align More Closely With Residents Than Relatives on Thriving. Scienmag. https://scienmag.com/care-home-staff-align-more-closely-with-residents-than-relatives-on-thriving/
Ophelia Keating. "Care Home Staff Align More Closely With Residents Than Relatives on Thriving." Scienmag, 20 November 2026, https://scienmag.com/care-home-staff-align-more-closely-with-residents-than-relatives-on-thriving/. Accessed 20 November 2026.
Ophelia Keating. "Care Home Staff Align More Closely With Residents Than Relatives on Thriving." Scienmag. November 20, 2026. https://scienmag.com/care-home-staff-align-more-closely-with-residents-than-relatives-on-thriving/








