Design And Participants
Prior to conducting this research, we obtained approval from the institutional review board of the main participating hospital. In total, we contacted 189 dyads of PWCIs and their family caregivers (FCGs) through two memory disorder clinics at two teaching hospitals and from local community long-term care service programs in northern Taiwan. Nine dyads did not complete the test battery; therefore, we used the data of only 180 dyads for statistical analysis (Fig. 1).
(Insert Fig. 1 about here)
Once consent from the dyads was received, the PWCIs underwent a standard comprehensive evaluation at their respective clinics. To determine eligibility, a team of neurologists, clinical psychologists, and nurses assessed the type and severity of dementia and cognitive status. The main neurologists at the clinical sites diagnosed dementia through consensus agreement following the criteria in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) (American Psychiatric Association, 1994), guidelines of the National Institute of Neurological and Communicative Disorders and Stroke, and the Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) (McKhann et al., 1984). PWCIs were eligible for the study if they: (1) spoke Mandarin Chinese, Taiwanese, or Hakka dialect (all the recruited participants understood Mandarin which was the official language of Taiwan); and (2) had primary or secondary FCGs. PWCIs were excluded if: (1) their primary residence was a nursing home; (2) they had an acute illness, impaired sensory symptoms (hearing loss and/or severe visual problems), chronic alcohol abuse, or drug usage affecting the central nervous system; or (3) they had a Clinical Dementia Rating (CDR) greater than 3 (Morris et al., 2001).
To address the recall bias of PWCIs, FCGs were included if, for the past three months, they: (1) had provided most of the caregiving assistance to the family member with cognitive impairment; or (2) had provided secondary care by supervising hired care assistants or foreign helpers who were primary caregivers of the PWCI. Co-residency was not required for FCGs because, in Taiwan, it is common for married adult children who live nearby to regularly visit their parents’ houses to assist with chores and/or medical needs. Family caregivers were excluded if they had: (1) been diagnosed with a cognitive or mental disorder, such as severe memory problems or major affective disorders; (2) untreated hearing or visual impairments; (3) been prescribed drugs known to impair or enhance attention, such as antidepressants, barbiturates, other depressants, or amphetamines; or (4) an insufficient command of Chinese, Taiwanese, or Hakka languages that would preclude them from testing.
Translation Of The Chinese Wayfinding Effectiveness Scale (Cwes)
The WES, developed by Algase et al. (2007), has a four-factor solution based on factor analysis results. The factors consist of: (1) complex wayfinding goals (CWG); (2) analytic strategies (AS); (3) global strategies (GS); and (4) simple wayfinding goals (SWG). The internal consistency reliability for the 30-item WES is high (.94 to .95). Cronbach’s alpha values for the four subscales are stable (greater than .70) for these three versions: (1) FCGs reporting PWCIs’ current wayfinding effectiveness; (2) FCGs reporting PWCIs’ wayfinding effectiveness before disease onset; and (3) FCGs’ current wayfinding effectiveness as a controlled condition. Test-retest reliability of the WES is acceptable for persons with dementia. Details of the scale’s development are described in Algase et al. (2007).
Following the forward-and-back method (Brislin, 1970), the first author translated the WES into Chinese. Then, the second author, an English-fluent Taiwanese faculty member with a specialization in gerontology, back-translated the CWES to English. After comparing the 30 items on the back-translated version with the original English version, Dr. Donna Algase who was the original developer for the instrument, rated all 30 items as equivalent in meaning.
A six-member panel of experts in gerontological nursing, neurology, and clinical psychology evaluated the face and content validity of the CWES. The panel’s comments guided revisions in the Chinese wording to clarify the starting point and endpoint of a travel path. We pilot-tested the updated translation with four FCGs to determine acceptability by asking them to rate whether any part of the translated scale was difficult to understand or confusing. Their feedback was positive, and no further modifications were made.
Measurements
We collected demographic data, such as age, gender, and education level, of the dyads. We also collected the PWCIs’ disease data (such as type and severity of dementia and cognitive status), FCGs’ relationship status with PWCIs, and if other helpers existed. The test battery for patients is described as follows.
Cognitive function
The Mini-Mental Status Exam (MMSE), a popular instrument developed by Folstein et al. (1975), was used to screen and monitor global cognitive impairment. The MMSE measures orientation, registration, recall, language, and spatial capacity for a total possible score of 30. A higher score indicates a higher level of cognitive functioning. A psychologist assessed cognitive function in PWCIs with the Chinese version of the MMSE (CMMSE), which has good validity and reliability (Guo et al., 1988; Shyu & Yip, 2001). The MMSE scores were copied from the patients’ charts with their permission. Therefore, we did not calculate its Cronbach’s alpha for this study.
Severity of dementia
The Clinical Dementia Rating (CDR) scale determines the stages of dementia for PWCIs using six domains: (1) memory; (2) orientation; (3) judgment and problem-solving; (4) community affairs; (5) home and hobbies; and (6) personal care. An overall score indicates the severity of dementia: 0 = none; 0.5 = very mild; 1 = mild; 2 = moderate; and 3 = severe. In this study, a psychologist evaluated dementia severity using the Chinese version of the CDR (CCDR). The global score for the Chinese version of the scale has an interrater reliability of kappa 0.63 (Lin & Liu, 2003). The CDR scores were copied from the patients’ charts with their permission. Therefore, we did not calculate its Cronbach’s alpha for this study.
Neuropsychological inventory
Behavioural and psychiatric symptoms of the PWCIs were assessed using the Chinese version of the Neuropsychiatric Inventory (CNPI) developed by Leung et al. (2001). Leung et al. evaluated the psychometric properties of the CNPI with a population of PWCIs in Hong Kong. Concurrent validity of all eight subscales of the CNPI correlated significantly (p < .001) with the corresponding domains of the Behavioural Psychology of Alzheimer’s Disease scale, which measures behavioural and psychological symptoms of dementia (Reisberg et al., 1997). The CNPI has a high level of internal stability, according to its internal consistency reliability of .84 for overall reliability, and both severity and frequency of .86. Kappa coefficients for interrater reliability are acceptable (range = .7 to 1.00), with intraclass correlation coefficients (ICC) for all subscales greater than .90 (Leung et al., 2001). Its overall Cronbach’s alpha for this study was .88.
Chinese Wayfinding Effectiveness Scale
The 30 scale items were randomly ordered for the final CWES instrument. Items were scored on a 5-point Likert scale ranging from 1 (never or unable) to 5 (always); higher scores indicated better wayfinding effectiveness (Algase et al., 2007). We asked FCGs to complete the CWES and respond to each item as it applied to current behaviours of their care recipients.
Data analysis
To identify the conceptual constructs of the CWES, we conducted an exploratory factor analysis using principal component analysis with varimax rotation (Kim & Mueller, 1978). To determine internal consistency and reliability of the CWES, we calculated Cronbach’s alpha for the total scale and each subscale. To find one-week test-retest and interrater reliability, we calculated intraclass correlations with a subset of FCGs (n = 8). To determine construct validity, we used correlation coefficients for the subscale and total CWES scores, CMMSE, and CNPI. Finally, to determine sensitivity and specificity, we used a Receiver Operating Characteristic (ROC) analysis of CDR scores to distinguish PWCIs with mild cognitive impairment from those in the early stages of dementia.
Ethical Considerations
This project was approved by the Institutional Review Board of the participating institution (IRB No. 97-0483B). Additional protection was provided to the vulnerable participants with cognitive impairment under the Declaration of Helsinki regulation. For example, we made sure that the PWCIs understood the purpose of this study by explaining this study in plain language and verifying their response at least three times. We also invited their family caregivers to help to explain this study to the PWCIs. If the PWCIs refused to participate in this study, we would ask them three times (30 minutes apart) and respected their final decision. Those who agreed to participate provided written informed consents and were assured of anonymity and confidentiality. All the informed consents of the dyads were obtained.