The Healthcare Improvement Scotland Think Frailty Tool: A Preliminary Psychometric Evaluation
Author(s)
Martin Taylor-Rowan, PhD.
University of Glasgow, Glasgow, United Kingdom.
University of Glasgow, Glasgow, United Kingdom.
OBJECTIVES: Identification of frailty is crucial to appropriately tailor patient care. Most frailty screening tools are difficult to use in unscheduled hospital settings. The Healthcare Improvement Scotland (HIS) ‘think frailty’ tool was developed to easily screen for frailty in frontline healthcare. We evaluated the psychometric properties of the HIS ‘think frailty’ tool.
METHODS: We conducted a prospective observational cohort study. Consecutive adults aged ≥65 who presented as unscheduled admissions to hospital were recruited. Recruitment occurred in 3 phases between 2017-2023. Participants completed the HIS ‘think frailty’ tool and the Rockwood Clinical Frailty Scale (CFS). We assessed HIS 'think frailty' tool's concurrent validity (Pearson’s correlations), overall discrimination (area-under-curve; AUC), intra-rater reliability (adjusted Kappa), inter-rater reliability between specialist and non-specialist assessors (adjusted Kappa), feasibility (percentage of missing data within each HIS ‘think frailty’ domain), and predictive validity for 30-day mortality and duration of hospital stay (Regression). A Maximum likelihood factor analysis was conducted to investigate the underlying factor structure.
RESULTS: A total of 2447 older adults were recruited. HIS ‘think frailty’ scores were significantly correlated with CFS (0.78,p<0.001) and discrimination was excellent (AUC:0.934,95%CI=0.921-0.946). Intra-rater agreement was substantial (Kappa=0.74,95%:0.61-0.86), as was Interrater agreement between a specialist and non-specialist assessor Kappa=(0.83,95%CI:0.68-0.97). All domains of HIS ‘think frailty’ had <1.5% missing data. HIS ‘think frailty’ was significantly associated with an increased risk of mortality up to 30 days (OR:1.49,95%CI=1.32-1.69) and likelihood of being in hospital at 30 days (OR:1.727,95%CI=1.52-1.96). Factor analysis suggested correlations between the items in the HIS ‘think frailty’ were within an acceptable range. The The Kaiser-Meyer-Olkin (KMO) test was acceptable (0.69) and Bartletts test of Sphericity was highly significant (<0.001). The scree plot indicated a single underlying factor accounting for 43.4% of the variance.
CONCLUSIONS: The HIS ‘Think Frailty’ tool has good psychometric properties for ‘font door’ frailty screening and can be used by non-specialists.
METHODS: We conducted a prospective observational cohort study. Consecutive adults aged ≥65 who presented as unscheduled admissions to hospital were recruited. Recruitment occurred in 3 phases between 2017-2023. Participants completed the HIS ‘think frailty’ tool and the Rockwood Clinical Frailty Scale (CFS). We assessed HIS 'think frailty' tool's concurrent validity (Pearson’s correlations), overall discrimination (area-under-curve; AUC), intra-rater reliability (adjusted Kappa), inter-rater reliability between specialist and non-specialist assessors (adjusted Kappa), feasibility (percentage of missing data within each HIS ‘think frailty’ domain), and predictive validity for 30-day mortality and duration of hospital stay (Regression). A Maximum likelihood factor analysis was conducted to investigate the underlying factor structure.
RESULTS: A total of 2447 older adults were recruited. HIS ‘think frailty’ scores were significantly correlated with CFS (0.78,p<0.001) and discrimination was excellent (AUC:0.934,95%CI=0.921-0.946). Intra-rater agreement was substantial (Kappa=0.74,95%:0.61-0.86), as was Interrater agreement between a specialist and non-specialist assessor Kappa=(0.83,95%CI:0.68-0.97). All domains of HIS ‘think frailty’ had <1.5% missing data. HIS ‘think frailty’ was significantly associated with an increased risk of mortality up to 30 days (OR:1.49,95%CI=1.32-1.69) and likelihood of being in hospital at 30 days (OR:1.727,95%CI=1.52-1.96). Factor analysis suggested correlations between the items in the HIS ‘think frailty’ were within an acceptable range. The The Kaiser-Meyer-Olkin (KMO) test was acceptable (0.69) and Bartletts test of Sphericity was highly significant (<0.001). The scree plot indicated a single underlying factor accounting for 43.4% of the variance.
CONCLUSIONS: The HIS ‘Think Frailty’ tool has good psychometric properties for ‘font door’ frailty screening and can be used by non-specialists.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
P32
Topic
Clinical Outcomes, Epidemiology & Public Health, Health Service Delivery & Process of Care
Topic Subcategory
Comparative Effectiveness or Efficacy
Disease
Geriatrics, No Additional Disease & Conditions/Specialized Treatment Areas