Study population and recruitment.

The study population comprises patients diagnosed with dementia and MCI who visit each participating institution and meet the inclusion criteria. Each participating institution plans to promote the study by posting recruitment posters on hospital websites or offline bulletin boards. Upon patient visits, screening evaluations will be conducted to assess whether they meet the inclusion and exclusion criteria. Patients who agree to participate will be provided with detailed information about the study and all safety-related matters before they are asked to provide informed consent. Furthermore, because this study is a government-funded project of the National Development Institute of Korean Medicine (NIKOM), data may be anonymized and linked to secondary data sources, such as national health insurance claims data, for future use. Accordingly, separate consent will be obtained for granting access to third parties and use in secondary research. The electronic Case Report Form (e-CRF) is established by the NIKOM and the information is managed by the government. Such registries are being conducted simultaneously for other diseases, and the government wants to continue research and manage data. We will conduct an observational study for 7 years with funding and then continue to receive funding to build the registry and conduct related research.

Inclusion criteria.

1. Adults aged 55–85 years.
2. Individuals diagnosed with a major neurocognitive disorder caused by Alzheimer disease or vascular disease, or a mild neurocognitive disorder based on the criteria described in the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5).
3. Individuals who voluntarily agree to participate or whose legal representatives have provided signed informed consent.

Exclusion criteria.

1. Patients with dementia caused by conditions other than Alzheimer disease or vascular dementia (e.g., degenerative brain diseases such as Parkinson disease, Huntington disease, frontotemporal disorder, and Creutzfeldt–Jakob disease).
2. Individuals with systemic conditions that may cause dementia (e.g., hypothyroidism, vitamin B12 or folic acid deficiency, niacin deficiency, hyperkalemia, neurosyphilis, and human immunodeficiency virus disease).
3. Individuals with a history of being diagnosed with a psychotic disorder or substance-related disorder based on DSM-5 criteria (e.g., schizophrenia, delusional disorder, bipolar disorder, and alcohol or substance abuse disorder).
4. Individuals with a history of neurological disorders such as epilepsy, focal brain injury, and head trauma.
5. Individuals with gastrointestinal, endocrine, or cardiovascular diseases that cannot be controlled through dietary or drug therapy.
6. Individuals in a serious unstable medical condition (as determined by a physician based on laboratory tests, electrocardiogram results, chest X-rays, and vital signs).
7. Others whom the investigator deems ineligible to participate in the study.

Sample size.

The sample size was calculated as follows: Through discussions among the research team, we pre-selected 15 covariates that are expected to affect changes in MOCA-K, an important continuous variable for evaluating treatment progress. The 15 covariates are as follows: gender, age, smoking, alcohol consumption, weight, height, cognitive impairment severity, region of residence, duration of illness, income, education level, occupation, marital status, medication use, and family history. We aimed to conduct a multiple linear regression analysis to identify which covariates influence changes in MOCA-K for patients receiving Korean medicine treatment. It is generally known that multiple linear regression analysis requires approximately 20 samples for each independent variable, so the total sample size was estimated to be around 300. As this study is intended as an epidemiological investigation inherent to the nature of the registry, no predefined statistical hypotheses have been formulated. As new research questions emerge over time, additional statistical analysis plans will be devised. Consequently, each participating institution has estimated the number of eligible patients who could reasonably be registered to calculate the final sample size. The average number of patients with dementia and MCI across the three institutions is 121 per year. With an estimated registration rate of approximately 50%, the objective is to enroll 60 new participants annually, with the goal of obtaining a total of 300 new registrations in total over five years.

Data analysis.

The analysis plan for the research questions that arise over time is as follows. Descriptive analyses will be employed to demonstrate the baseline data and disease-related variables. Nominal variables will be presented as frequencies and ratios, whereas continuous variables will be presented as means and standard deviations. Baseline data will be subjected to statistical analysis to identify associations between the data and several variables, including treatment characteristics, treatment effects, factors influencing KM treatment, health outcome variables, and prognosis. Changes in continuous variables in all patients or specific subgroups will be tested using either paired t-tests or Wilcoxon signed-rank tests, depending on the normality test results. Changes in the nominal variables will be tested using McNemar’s test. Correlations between continuous variables will be analyzed using Pearson (or Spearman) correlation analysis. Differences in continuous variables between subgroups will be analyzed using independent t-tests (or Wilcoxon rank-sum tests, depending on the normality test results). Differences in nominal variables will be analyzed using chi-squared tests (or Fisher’s exact test, depending on the normality test results). Multiple regression analysis will be used to test the effects of the covariates on a particular continuous dependent variable, and multinomial logistic regression analysis will be used to test the effects of the covariates on the occurrence of binominal indicators. Survival analysis will be performed to test the effects of the covariates on the occurrence of a particular binomial variable over time. The proportion of missing data that can be included in the analysis is limited to 10%. The missing data will be imputed using one of several strategies, including unconditional or conditional mean imputation and expectation maximization.

Data collection.

This study is a pure research project with no feedback to participating hospitals and aims to collect basic patient sociodemographic characteristics, survey information, and clinical practice data. All documents pertaining to registry participants will be recorded using identification codes rather than names and stored on computers as encrypted files with access restricted to approved researchers only. Trained trial staff and research coordinators will be responsible for collecting baseline and follow-up data. All data will be stored in an e-CRF using an electronic data capture system called MyTrial, which is managed by the Korea Institute of Oriental Medicine. The online database permits information to be modified and updated as required and real-time exportation for the purposes of quality management and statistical analysis purposes. Once permission has been granted for the data to be used for public purposes, it will be accessible via the online database.

Data quality control and queue maintenance.

Data completeness and reliability will be evaluated by independent monitoring personnel separate from the researchers and sponsors. The monitoring personnel will be responsible for ensuring compliance with the observational study protocol, endorsing modifications of the registry protocol, appropriate and accurate data collection, and confirmation of participants’ informed consent. If data completeness and reliability are compromised, face-to-face or telephone interviews will be conducted with the participants to facilitate the necessary corrections. Furthermore, a random sample of less than 10% from each institution will be selected up to five times during the study period for comparison between the data entered in the e-CRF and the paper case report forms maintained at each institution. If the input errors exceed a predetermined acceptable level, the institution will compare all case records and input data. To minimize dropout rates, materials related to health and cognitive disorders, including lifestyle information, will be distributed at least once per year. Additionally, the study coordinator will contact the participants to monitor their health management status.

Sociodemographic characteristics.

Information on the sociodemographic characteristics and medical history of the participants and their caregivers will be collected.

Laboratory tests and medical device inspections.

Data will be collected from participants regarding hematology tests, blood chemistry tests, urinalysis, Vital signs, and medical device inspections (electrocardiogram, chest X-ray, heart rate variability, quantitative electroencephalograph, and functional near-infrared spectroscopy)

Long-term care.

Information will be collected on participants’ long-term care insurance grade, nursing home admission, and long-term care services currently in use.

Questionnaires.

Pattern identification tool for cognitive disorder: The PIC-C [20,27] comprises 22 items across 10 categories and classifies cognitive impairment into four patterns: Qi Deficiency, Yin Deficiency, Dampness, and Heat Syndrome. In addition to determining the primary pattern, it quantitatively assesses the tendencies of other patterns. Consequently, it can be utilized for evaluating the selection of primary prescriptions for herbal medicine and acupuncture treatment, adjusting herbal dosages, selecting adjunctive acupuncture treatments, and assessing treatment progress.

Barthel index: Various versions of the Barthel Index are available. Originally, the total score ranged from 0 to 100 points; however, this study used the version revised by Collin et al. [28], which was translated into Korean as the Korean version of Barthel Activities of Daily Living Index (K-BADL) [29]]. The K-BADL assesses the ability to perform basic activities of daily living. The scale is based on a total score of 20 points, with 11–15 points indicating moderate disability and ≤ 10 points indicating severe disability.

Seoul-instrumental activities of daily living: The S-IADL [30] was developed and standardized using items suitable for Korea’s cultural characteristics. It comprises 15 items rated on a scale of 0–3 points, with a total score ranging from 0 to 45 points. Higher scores indicate greater disability in daily life. This tool evaluates instrumental activities of daily living by dividing them into current and potential abilities. It has been demonstrated that a cut-off score of 7.5 can be used to differentiate between normal individuals and those with dementia. An S-IADL score of 8 points or above indicates the likely presence of dementia.

Brief questionnaire form of neuropsychiatric inventory: The NPI-Q [31] is a streamlined version of the original NPI [32] designed for straightforward use in clinical settings. The instrument evaluates the severity of 12 common neuropsychiatric symptoms observed in patients with dementia and their caregivers. It assesses both the severity of neuropsychiatric symptoms (rated 1–3) and caregiver distress caused by these symptoms (rated 0–5). Higher scores indicate greater severity. The Korean version of the NIP-Q (KNPI-Q) [33] will be used in this study.

Mini-mental state examination 2nd edition: Standard version: The MMSE [34] is the most widely used test for screening cognitive impairment in older adult populations. To provide greater sensitivity and adaptability to each country’s culture, the MMSE-2 was developed, which consists of three versions: Brief Version (BV), Standard Version (SV), and Expanded Version (EV). As with the original MMSE, the K-MMSE-2:SV has a total score of 30 points. It assesses orientation to time (5 points), orientation to place (5 points), registration (3 points), attention and calculation (5 points), recall (3 points), language (8 points), and drawing (1 point). A total score of 23 points is typically considered the cutoff point for cognitive impairment.

Montreal cognitive assessment: The MoCA [35], which has been translated into Korean [36], will be used. The MoCA-K is a questionnaire developed to screen for MCI in individuals with normal findings on the MMSE test. Total scores equal 30 points, and a score of 23 or higher is considered normal.

Global deterioration scale: The Korean version of the GDS will be used in this study [37,38]. The GDS is generally used to quantify the extent of dementia progression. GDS 1 indicates a clinically normal individual without cognitive impairment, GDS 2 indicates subjective memory impairment, and GDS 3 indicates MCI and some with mild dementia. From GDS 4 onwards, a definitive diagnosis of dementia can be made, with GDS 4 representing mild dementia, GDS 5 representing moderate dementia, and GDS 6 and 7 representing severe dementia.

Short form geriatric depression scale: This study will use the Korean version of the S-GDpS-K [39], which has been developed by shortening the original 30-item GDpS [40] to 15 items. This self-report questionnaire allows for simple “yes” or “no” responses. A score of 8 or higher indicates a depressive state.

Core seven-emotions inventory short form: The CSEI-s [41] consists of 28 items across 7 categories, with each item measured on a 5-point Likert scale. This scale uses t-scores (mean 50, standard deviation, 10) as reference points. For emotions other than happiness, including anger, thoughtfulness, sadness, fear, anxiety, and surprise, higher scores indicate a higher risk level. Regarding the specific cutoff points: t-scores ranging from 55 to 60 are considered cautionary, scores from 61 to 65 are considered risky, and scores above 66 are classified as high risk. In contrast, for happiness, lower scores indicate a higher likelihood, with t-scores ranging from 40 to 45 considered slightly likely, 35–39 considered likely, and scores below 34 classified as highly likely.

Korean medical pathologic aging scale: The Korean Medical Pathologic Aging Scale [20] comprises 9 items related to the sensory and physical symptoms of aging. Each item is measured on a 3-point Likert scale, with 0 points indicating “none,” 1 point indicating “sometimes present,” and 2 points indicating “frequently present.” Lower scores signify a better health status in relation to aging.

EuroQol 5 dimensions 5 levels and EuroQol visual analog scale: The EQ-5D [42,43] assesses health-related quality of life and comprises five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension is evaluated on a five-point scale, with 1 indicating the best state and 5 indicating the worst state. The EQ-VAS is a visual analog scale comprising a vertical line of 100 mm. The EQ-VAS represents the respondent’s self-rated health status on a scale of 0–100, on which 100 represents the best imaginable health state and 0 represents the worst imaginable health state. In general, rating scales are recommended as supplementary tools to other instruments. The EQ-VAS was also included as a supplementary component to the standard EQ-5D questionnaire.

Geriatric quality of life-dementia: The GQOL scale [44] is used to assess the comprehensive quality of life of older adults and patients with dementia. Subsequently, the Geriatric Quality of Life-Dementia scale (GQOL-D) [45] was developed to reflect the quality-of-life domains of particular importance to patients with dementia. The GQOL-D comprises 13 items that assess various quality-of-life domains, including physical health, psychological health, independence level, social relations, environment, and religion. It also includes one item for overall health and another item for overall life satisfaction. Each item is measured on a 4-point Likert scale.

Caregiver burden inventory: The CBI [46] assesses the burden experienced by caregivers of patients with dementia. The scale has 24 items in total, divided into five domains: time-dependent burden, developmental burden, physical burden, social burden, and emotional burden. One advantage of the CBI is that it allows for the simultaneous measurement of both objective and subjective burden.

Adverse events.

This registry study is designed without any interventions included and has a low expectation of risks or adverse events (AEs). All medical histories collected during the follow-up visits after the initial visit will be documented as medical histories, and only recorded as serious AEs if the participant dies. Additionally, cumulative mortality cases from all participating institutions will be reported collectively during periodic reviews. Clinical research indemnity insurance has been obtained to mitigate potential harm from trial participation.