Research area
As the most economically developed super metropolis in China, Shanghai has one of the largest populations of aged adults. The number of elder people aged 65 and older has shown a significant upward trend for recent years[15, 16], Figure 1. Importantly, Shanghai is also a pioneer and in the vanguard of NH cultural change pilots. Thus, it was selected for analysis by this study. The flow chart of the study and its details were shown in Figure 2.
Participants
Through stratified random sampling, we selected 310 elders form 15 district-level NHs, which are all on a journey of cultural transformation. The main inclusion criteria for elders participation in the study were; (1) aged 65 years or older, (2) conscious consent to research participation, (3) awareness of place and time, (4) and the ability to communicate and respond to the questionnaire. Participants who had been under extreme stress in the past month were excluded[17]. Finally, there were 10 residents could not complete the survey due to sudden physical limitations. Therefore, of the 310 potential elders, 96.8% (n = 300) participated.
Of the 300 participants, 145 (48.3%) were male and 155(51.7%) were female. Most elders (n= 95, 31.7%) were illiterate, approximately 30% (n = 90) had a primary school education, 21.7% (n = 65) had a high school degree, and 16.6% had attended university (n = 50). Almost all residents (99.3%, n = 298) had a son or daughter. Two did not. Age ranged from 65 to 99 with a mean of 80.9 (standard deviation [SD] = 9.87). Length of stay within the nursing home ranged from 6 months to 10 years (Mean = 39.9 months, SD = 26.43 months). Table 1 shows the general characteristics of all participants.
Table 1. General characteristics of participants (n = 300)
Variable
|
Proportion/Number
|
Sex
|
|
Male %
|
48.3
|
Female %
|
51.7
|
Age
|
|
Mean, year
|
80.9
|
Education
|
|
Illiterate %
|
31.7
|
Primary school %
|
30.0
|
High school %
|
21.7
|
University %
|
16.6
|
Son/daughter
|
|
Yes %
|
99.3
|
No %
|
0.7
|
Length of in NH
|
|
Mean, days
|
1497
|
Measures
QOL Scales
QOL was measured with the Quality of Life Scales for Nursing Home Elders by Kane et al[13, 14]. This instrument examined the elderly perception of QOL in 11 areas with 4-point Likert Scales. Higher scores usually indicated greater QOL. The following are sample items for each QOL scale with alphas derived from this study. Physical comfort: “Do you feel pain in your body when you stay in one place for a long time?” (Cronbach a = 0. 628). Security: “Can you obtain a doctor or nurse quickly when you need help?” (a = 0.773). Meaningful activity: “Apart from religious activities, do you enjoy other recreational activities organized by the nursing home” (a = 0.76). Relationships: “Do you consider one or more members of the staff to be your friend?” (a = 0.794). Functional competence: “Can you get to a bathroom quickly anywhere in the nursing home”? (a = 0.778). Enjoyment: Do you enjoy mealtimes at (nursing home name)?” (a = 0.744). Privacy: “Do staff knock & wait before entering your room?” (a = 0.849). Dignity: “Do staff take time to listen to you when you are depressed?” (a = 0.805). Autonomy: “Can you get up or go to bed at the time you want?” (a = 0.704). Individuality: “Are the staff here willing to learn about your interesting life experience?” (a = 0.305). Spiritual well-being: “How often do you participate in religious activities?” (a = 0.721). Because of poor scale reliability (alpha = 0.305, which is lower than 0.6) for individuality, we excluded this domain.
The remaining 10 QOL domains were evaluated. Each QOL scale was comprised of three to six items with all but three providing a 4-point Likert-type response, usually with four choices, “never, rarely, sometimes, and usual”, and scored from 1 = “never” to 4 = “usual”. It’s worth noting that three items that had different response sets were components of the meaningful activities and relationships scales. Two questions were posed in terms of whether some activities (e.g., being outdoors, participating in recreational activities) happened from “less than once a month” = 1 to “every day” or “as much as you want” = 4. The other question regarding relationships was in a dichotomous format with 1 = “no” and 4 = “yes”. Reverse scoring was used where appropriate, so high scores usually reflected a better QOL. A binary coding system (“mostly no” = 1.5, “mostly yes” = 3.8) was provided for those respondents who were unable to use the 4-point scale. In this study, 10 respondents used the binary system. After completion of almost every question, mean scores (from 1 to 4) for each QOL domain scale were derived. Any missing item scores were replaced by the item mean.
Control Variables
Functional Impairment
Elderly residents’ abilities to conduct activities of daily living (ADL) were identified using four ADL measures; bed mobility, toilet use, transferring, and eating[18]. These results were tabulated as a single ADL score. The final score for each elder ranged from 1 = “independent” to 4 = “completely dependent”. A higher score indicated lower functional capacity.
Cognitive Impairment
Referring to the seven-category Minimum Data Set Cognitive Performance Scale (CPS) [19]proposed by Morris et al, we designed five questions in five dimensions (state of consciousness, the level of memory, cognitive skills for making decisions, ability to be understood by others, and ability to enjoy meals independently) to evaluate the cognitive ability of the subjects. The final CPS score ranged from 1 = “intact” to 4 = “severe impairment”. A lower score indicated higher cognitive capacity.
Outcome Variables
Satisfaction with nursing services
In this module, the Satisfaction with Nursing Care Scale (SNCS) [20] was used to assess participants’ satisfaction with the quality of care. The SNCS is a part of the Newcastle Satisfaction with Nursing Scale[21]developed by Thomas et al that consists of 19 items using 5-point Likert scales. Scores for satisfaction ranged from 1 = “very dissatisfied” to 5 = “completely satisfied”. Final SNCS scores ranging from 1 to 5 were replaced by means determined for a total of 19 items.
Satisfaction with epidemic prevention and control (EPC)
Since the end of 2019, COVID-19 has swept the world with more than 100 million people infected[22]. As the elderly have generally weakened immune systems and other predisposing conditions, NHs are hardest hit by epidemics. Based on existing researches[23, 24], we designed five items to evaluate resident satisfaction with EPC based on three dimensions; body temperature measurement, availability of new coronavirus knowledge, and procedures for EPC. The scores for each item ranged from 1 = “extremely dissatisfied” to 5 = “extremely satisfied”. Satisfaction with EPC scores ranged from 1 to 5 with replacement of means derived from a total of five items.
Overall satisfaction with NHs
For comparison to QOL indicators, we used two items to measure elder satisfaction with NHs. One item required residents to rate their satisfaction with NHs using a 5-point Likert Scale format from 1 (“extremely dissatisfied”) to 5 (“extremely satisfied”). The other item examined whether the elderly were willing to recommend the NH to their relatives or friends with a 5-point scale, which ranged from 1 = “completely not” to 5 = “certainly”. The final NH satisfaction score ranged from 1 to 5 with replacement of means for two questions.
Procedure and Data Analysis
The research was approved by the Shanghai Institution of Technology Human Research Ethics Board. Before conducting the survey, we informed NHs and participants of the purposes and procedures of the study. Also, written informed consent was obtained from all participants.
With the help of research assistants, the elderly completed the survey through face-to-face interviews in a private and quiet atmosphere, usually in the elder’s room. Interviews were also conducted in other NH places such as the library and chess room. Each interview required approximately 40 minutes to complete. If necessary, the duration of the interview was extended. When participants felt unwell, they could interrupt the interview and complete the interview later.
Internal consistency reliability of the QOL and the three satisfaction scales were examined by Cronbach a. Second-order confirmatory factor analysis (CFA) was used to examine domain loading between overall QOL and related domains. Also, we examined potential bivariate correlations among demographic details and the 10 QOL domains. In regression models, only those variables that had significant relationships with one or more outcome variables were considered. Finally, hierarchical regression models were used to analyze possible predictors of different satisfaction categories. Two control variables were included in the first step regression models and the QOL domains in the second step models. Further, no multicollinearity was detected. SPSSAU and Excel 2010 software were used for data analysis.