Abstract / Summary
Mild cognitive impairment represents a critical window for early intervention, yet healthcare-seeking for cognitive symptoms often involves substantial delays. Healthcare-seeking decision-making for mild cognitive impairment is frequently a joint process shaped by both patient and caregiver perceptions and negotiations. However, existing instruments focus primarily on decisional capacity or treatment engagement and are not designed to assess the pre-diagnostic healthcare-seeking decision-making process from both patients’ and caregivers’ perspectives. This study aimed to develop parallel instruments for assessing healthcare-seeking decision-making among older adults screened as having MCI and caregiver dyads. This multi-phase scale development and validation study developed the instruments based on concept analysis of healthcare-seeking decision-making, literature review, qualitative interviews, Delphi expert consultation, and psychometric testing. Initial item generation was informed by systematic literature review and qualitative interviews with 19 older adult–caregiver dyads from December 2024 to January 2025. Two rounds of Delphi consultation with 17 multidisciplinary experts refined the item pool. Psychometric testing involved a survey of 447 older adults screened as having MCI and their caregivers recruited from two tertiary hospitals, two secondary hospitals, and six communities in China from May to July 2025. The sample was randomly split for exploratory factor analysis and confirmatory factor analysis to examine the factor structure. Reliability and validity evidence were evaluated. Psychometric testing in 447 older adults screened as having MCI and their caregive produced two 36-item scales: the Healthcare-Seeking Decision-Making Scale–Patient version and Caregiver version, each comprising six dimensions (Symptom Perception/Recognition, Decision-making Dilemma, Decision-making Help, Decision-making Conflict, Decision-making Balance, and Decision-making Results). EFA identified a six-factor solution explaining 73.769% (HSDM-P) and 72.246% (HSDM-C) of the variance. CFA supported the six-factor model of HSDM-P (χ²/df = 1.833, GFI = 0.832, AGFI = 0.806, CFI = 0.924, IFI = 0.925, TLI = 0.917, RMSEA = 0.053, SRMR = 0.051) and HSDM-C (χ²/df = 1.820, GFI = 0.832, AGFI = 0.806, CFI = 0.941, IFI = 0.941, TLI = 0.935, RMSEA = 0.053, SRMR = 0.049). Internal consistency was acceptable (Cronbach’s α = 0.843–0.914, McDonald’s ω = 0.895–0.937, split-half = 0.787–0.912 for HSDM-P and Cronbach’s α = 0.884–0.915, McDonald’s ω = 0.910–0.936, split-half = 0.829–0.890 for HSDM-C). Content validity was high (I-CVI = 0.882–1.000, S-CVI/Ave = 0.986 for HSDM-P; I-CVI = 0.941–1.000, S-CVI/Ave = 0.994 for HSDM-C), and convergent/discriminant validity were supported (AVE = 0.538–0.748, CR = 0.860–0.918 for HSDM-P and AVE = 0.534–0.839, CR = 0.889–0.954 for HSDM-C; the square root of AVE greater than inter-factor correlations). Domain-level correlations with the external healthcare-seeking measures were in the expected directions (HSDM-P: r = − 0.133 to − 0.322; HSDM-C: r = 0.148 to 0.230), providing preliminary evidence of construct validity based on relationships with conceptually related measures. The HSDM-P and HSDM-C showed preliminary psychometric support for assessing healthcare-seeking decision-making among older adults screened as having mild cognitive impairment and their family caregivers in China. Interpretation should focus primarily on the six domain scores, which may help identify specific barriers to timely healthcare-seeking and inform family-centered assessment and intervention. Further validation in independent and more diverse samples is warranted. Not applicable.