2.2.1 Sample size calculation
The required sample size was estimated according to the cross-sectional study sample size formula
where n is the estimated sample content; p is the expected rate, q = 1-p, and deff is the design efficiency value. In this study, the test level α = 0.05, uα = 1.96, and the permissible error δ is 0.1p. Based on the prevalence of senile dementia (≥ 65 years) in China at 8%, setting p = 8% and q = 1-p = 92%, the required sample size was calculated as n = 1.962×8%×92%/(0.1×8%)2 = 4417. The required sample size is 4417. Considering the response rate and other factors, a 20% increase in the sample size would require a total of at least 5300.
2.2.2 Investigation procedures and tools
The investigation was divided into two phases.The first stage is the screening stage, using self-designed general-information surveys of the population, including gender, education, chronic disease status, and self-rated health status. The first stage screening tools can be divided into two types: one is the Community Screening Scale for Dementia (CSI-D) and the other is the 8-item AscertainDementia (AD8) questionnaire. The former is a self-assessment questionnaire suitable for elderly people who can answer by themselves.The latter is a peer assessment questionnaire that can be answered by caregivers who live with the elderly. CSI-D scores range from 0 to 9 and assesses memory, cognitive ability, spatial ability, etc., with one point for correct answers. A score of 7–9 indicates normal, a score of 5–6 indicates suspicious dementia, a score of 4 or below highly indicates dementia, and a score of 6 or below is included in the list of the second stage for diagnosis and assessment by specialists [7]. AD8 screening questionnaire includes 8 items: 1) difficulty with judgment, 2) decreased interest in hobbies, activities, 3) repeating the same things over and over again, 4) learning how to use tools, appliances or gadgets, 5) forgetting the correct month and year, 6) Difficulty dealing with complex financial issues, 7) difficulty remembering appointments, 8) daily thinking/memory problems. The answer of "changed" to any question is worth 1 point, and the total score of all questions is the total score of AD8. If the total score of AD8 is greater than or equal to 2, it is highly suspected that the elderly may have early dementia, and all are included in the stage 2 list for diagnosis and assessment by a specialist. [8]
The second stage is diagnostic. The specialist uses the following two tools to further assess the diagnosis of the suspect population identified in the first stage of screening.
There are also two types of assessment tools used in the second stage: The first is mini-mental state examination (MMSE); the second is montreal cognitive assessment (MOCA). MMSE is suitable for people with junior middle school education level and below. MOCA is applicable to people with education level above middle school. MMSE is mainly used to evaluate orientation, memory, attention, calculation, recall, naming, retelling, reading, three-step command execution, writing and structure ability. The total score is 30, and less than 27 is abnormal [9]. MOCA can perform tests on visual-space and executive ability, naming, attention, backward reading, computing power, language ability, abstract thinking, delayed recall, and orientation. The total score is 30, and less than 26 is abnormal [10].
After completing the evaluation of the above tools, the specialist can make a diagnosis. For patients diagnosed with senile dementia, a care burden assessment was completed by a specialist. The care burden assessment is divided into three aspects. 1.Type of care needed: 1)Only ordinary daily care is required, 2)Home care for dementia is needed, 3)Need temporary nursing home care(several nights per week, 4)Day care is required not more than 3 days per week, 5)Day care is required for more than3 days per week, 6)Along-term stay in a nursing home is required 2. Care time of family members or nannies: 1) none, 2) less than 3 hours per week, 3) 3-l0 hours per week, 4) l0-20 hours per week, 5) 20–30 hours per week, 6) more than 30 hours per week. 3. Degree of care distress of family members or nannies: 1) no distress, 2) mild distress, 3) mild distress, 4) moderate distress, 5) severe distress, 6) severe distress.
In this survey, there are 5 979 people in 3 600 households who are older than or older than 65 years old. In fact, a total of 5 616 people have completed the survey, accounting for 93.9%(Fig. 1).