A method and device for generating psychological support prompts for improving depressive symptoms in frail elderly patients

Abstract

The application discloses a kind of psychological support prompt generation method, device for the depression symptom improvement of senile frailty patient, for the elderly population of age≥65 years, FRAIL≥1 and PHQ-9≥5 or GDS≥6.Method includes: collection frailty assessment data, objective activity data (sleep, step count etc.) and depression state scale data;Frailty grade F and depression risk grade D are calculated;Matched prompt information is retrieved from psychological support prompt content library;According to frailty grade, set push control parameter, including single interaction duration≤2 minutes, daily push number≤3 times, push time window 06:00-22:00;Feedback data is collected, when consecutive refusal number≥2 times, switch low-load mode and update prompt type.The application forms "data acquisition-state evaluation-prompt generation-push control-feedback update" technical closed loop, solves the technical problems of individualization deficiency, poor compliance, lack of objective data support, improves the accessibility of psychological support and user compliance.

Description

Technical Field

[0001] This invention relates to the field of health management and smart terminal application technology, specifically to a method and device for generating psychological support prompts to improve depressive symptoms in elderly and frail patients. Background Technology

[0002] Frailty in the elderly refers to a clinical syndrome characterized by a decline in the physiological reserves of multiple systems and a reduction in stress resistance in people aged 65 and above. It is often assessed using the FRAIL scale or indicators such as walking speed and grip strength. People with frailty in the elderly are characterized by fatigue, limited activity, and difficulty walking, and their daily living abilities and tolerance to external stimuli are significantly reduced.

[0003] Studies have shown a high degree of coexistence between frailty and depressive symptoms in the elderly. The incidence of depressive symptoms in frail elderly individuals is 2 to 3 times higher than in healthy elderly individuals, creating a vicious cycle: frailty leads to decreased activity levels and reduced social interaction, which in turn exacerbates depressive moods; depressive moods, in turn, lead to decreased self-care abilities and insufficient nutritional intake, further accelerating the frailty process. Therefore, monitoring depressive symptoms and providing psychological support to frail elderly individuals is of great significance.

[0004] Currently, the management of depression in the elderly mainly relies on community follow-ups, regular assessments in nursing homes, or outpatient follow-ups. However, existing technologies have the following problems: First, the content and frequency of follow-up visits lack individualization. Traditional follow-up visits often use fixed frequencies and uniform content, failing to differentiate management based on the degree of frailty of the elderly. Elderly people with higher degrees of frailty experience significant fatigue, short attention spans, and poor tolerance for high-frequency or long-duration interactions. Fixed, high-frequency push notifications or lengthy interactive content often lead to decreased compliance and may even trigger negative emotions.

[0005] Second, there is a lack of objective data support, resulting in insufficient timeliness of assessments. Existing management methods mainly rely on subjective scale assessments (such as PHQ-9 and GDS), lacking real-time collection of objective indicators such as sleep quality and activity levels. This makes it difficult to detect early changes in depressive states in a timely manner, and the assessment results are limited by the follow-up period, resulting in a lag.

[0006] Third, the timing and frequency of push notifications are poorly controlled, which can easily cause disruption. Existing follow-up systems mostly use fixed time slots for push notifications, failing to consider the sleep patterns of the elderly. Nighttime pushes or excessively frequent pushes can disturb their rest and reduce user acceptance. Furthermore, the lack of constraints on push frequency limits can easily lead to information overload.

[0007] Fourth, there is a lack of feedback loop mechanism. Most existing systems only perform one-way push, without collecting or making full use of user interaction feedback data (such as completion rate and number of rejections), and cannot dynamically adjust the push strategy according to the actual user response, resulting in wasted resources and poor intervention effect.

[0008] The aforementioned technical issues limit the effectiveness of existing follow-up management methods in elderly and frail populations, making it difficult to balance the accessibility of psychological support with user compliance. There is an urgent need for a technical solution that can combine frailty assessment, objective activity data collection, individualized push control, and feedback closed-loop updates. Summary of the Invention

[0009] To address the aforementioned shortcomings of existing technologies, the present invention aims to provide a method and apparatus for generating psychological support prompts to improve depressive symptoms in elderly frail patients. This method involves collecting frailty assessment data, objective activity data, and depressive state scale data via a terminal, calculating frailty levels and depression risk levels, retrieving matching individualized prompt information from a pre-set psychological support prompt content library, dynamically setting control parameters such as push frequency, single interaction duration, and push time window based on the frailty level, and automatically adjusting the push strategy based on user interaction feedback data. This forms a closed-loop technology of "data collection - status assessment - prompt generation - push control - feedback update," solving the technical problems of insufficient individualization, poor compliance, lack of objective data support, and missing feedback mechanisms in existing technologies.

[0010] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for generating psychological support prompts to improve depressive symptoms in elderly frail patients, applicable to mobile terminals or nursing station terminals, targeting individuals aged ≥65 years with a FRAIL score ≥1 and a PHQ-9 score ≥5 or a GDS score ≥6, includes the following steps: S1 Data Collection: Collect frailty assessment data (FRAIL scale five dimensions score), objective activity data (sleep duration, number of nighttime awakenings, daily steps, activity duration, or at least one of these) and depression status scale data (PHQ-9 or GDS score) through the terminal. S2 Status Assessment: Calculate the frailty level F (pre-frailty / frailty) and the depression risk level D (low / medium / high risk). S3 Prompt Generation: Based on F and D, retrieve matching prompt information from the psychological support prompt content library and generate a push plan. The content library includes templates such as emotional soothing, companionship and encouragement, low-load behavior activation, family communication, and nursing follow-up reminders. S4 Push Control: Set push control parameters according to the attenuation level F, including push frequency, single interaction duration ≤ 2 minutes, number of pushes per day ≤ 3, push time window 06:00-22:00, and interaction termination condition (threshold for consecutive rejections). S5 Feedback Update: Collects feedback data such as task completion rate, number of consecutive rejections, and emotion self-rating. When there are ≥2 consecutive rejections, it switches to low load mode (the push frequency is reduced to 50% of the original frequency and the number of words in the prompt is reduced), and updates the prompt type for the next cycle.

[0011] The present invention also provides a corresponding device, which includes a data acquisition module, a status evaluation module, a prompt generation module, a push control module, and a feedback update module.

[0012] The beneficial effects of this invention include: (1) By combining frailty assessment with objective activity data, the assessment of depressive state is improved in multiple dimensions and in real time, making up for the lag and one-sidedness of purely subjective scale assessment. (2) Adjust the push frequency, interaction duration and push time window dynamically according to the attenuation level, reduce the interaction load for high attenuation objects, set a night push ban window and a push interval lower limit, significantly reduce disturbance and improve compliance; (3) Establish an automatic adjustment mechanism for push strategies based on feedback data to form a closed-loop optimization, avoid the waste of resources in one-way push, and improve the level of individualized management of psychological support; (4) Provide differentiated prompts for patients, family members and nurses to achieve multi-role collaborative follow-up assistance and enhance the accessibility and systematic nature of psychological support; (5) By controlling the technical parameters clearly (duration threshold, frequency limit, triggering conditions), an executable technical solution can be formed to avoid the legal risks of intellectual activity rules. Attached Figure Description

[0013] Figure 1 This is a block diagram of the system structure of the present invention; Figure 2 This is a flowchart of the method of the present invention; Figure 3 This is a schematic diagram of the push control strategy of the present invention; Figure 4 This is a schematic diagram of the feedback update process of the present invention. Detailed Implementation

[0014] The present invention will be further described in detail below with reference to specific embodiments. The following embodiments are used to illustrate the present invention, but do not limit the scope of the present invention.

[0015] Example 1: System Structure Example This embodiment provides a psychological support prompt generation device for improving depressive symptoms in elderly and frail patients, such as... Figure 1As shown, this device can be deployed at nursing station terminals in elderly care facilities or on mobile terminals used by the elderly, and can be used in conjunction with wearable devices such as smart bracelets and smartwatches. The device includes the following modules: 1. Data Acquisition Module The data acquisition module is responsible for collecting basic information, frailty assessment data, objective activity data, and depressive state scale data of the target subjects. The inclusion criteria for target subjects are: age ≥ 65 years, FRAIL scale score ≥ 1 (i.e., already in the pre-frailty or frailty state), PHQ-9 scale score ≥ 5 or GDS scale score ≥ 6 (i.e., at risk of depressive symptoms).

[0016] The frailty assessment data collection uses five dimensions of the FRAIL scale: Fatigue, Resistance, Ambulation, Illnesses (≥5 diseases), and Loss of weight (>5%). Each dimension is scored 1 point for "yes" and 0 points for "no", with a total score of 0-5.

[0017] Objective activity data is automatically collected through wearable devices (smart bracelets or smartwatches), including: - Sleep duration: The total duration (in hours) of sleep each night; - Number of times you wake up during the night: The number of times you wake up during your sleep at night; - Daily Steps: The total number of steps taken each day; - Activity duration: The cumulative duration (in minutes) of moderate to high intensity activities each day.

[0018] At least one of the above should be collected, with sleep duration and daily steps being preferred.

[0019] Depression scale data were assessed by the target population or caregivers via a terminal interface, including: - PHQ-9 scale: 9 items, each item scored from 0 to 3 points, total score from 0 to 27 points; - GDS scale: 15-item summary or 30-item full version. The summary version has a total score of 0-15 points, and the full version has a total score of 0-30 points.

[0020] Data collection frequency: The FRAIL scale and PHQ-9 / GDS scale are collected every 7 days; objective activity data are automatically synchronized daily.

[0021] 2. Status Assessment Module The status assessment module is responsible for calculating the frailty level F and the depression risk level D based on the collected data.

[0022] The calculation rules for the attenuation level F are as follows: - FRAIL score 1-2: Early stage of decline; - FRAIL score 3-5: Weak.

[0023] The calculation rules for depression risk level D: When using the PHQ-9 scale: - Score 5-9: Low risk; - Score 10-14: Medium risk; - Score ≥ 15: High risk.

[0024] When using the GDS scale (15-item short form): - Score 6-10: Low risk; - Score 11-15: Medium risk; - Score ≥ 16 (or full version ≥ 20): High risk.

[0025] The status assessment module also marks anomalies in objective activity data: - Sleep disorders: Sleep duration <6 hours or nighttime awakenings ≥3 times; - Insufficient activity: Daily steps <2000 steps or activity duration <30 minutes.

[0026] 3. Prompt Generation Module The prompt generation module retrieves matching prompt information from a preset psychological support prompt content library based on the frailty level F, the depression risk level D, and the abnormality markers of objective activity data, and generates a push plan.

[0027] The psychological support tips content library includes five types of templates: (1) Emotional reassurance prompt template: Include short positive encouragement phrases, such as "The weather is nice today, you can sit by the window for a while and feel the warmth of the sun", "Every day is a new beginning, just take it slow". Avoid empty preaching and use lifelike and specific expressions.

[0028] (2) Companionship and encouragement prompts template: Include suggestions to reduce loneliness, such as "If it is convenient, you can call your family and hear their voices", "You can sit in the activity room and chat with other elderly people".

[0029] (3) Low-load behavior activation prompt template: Includes single-step, low-energy-consumption behavior suggestions, such as "drink a glass of warm water to replenish your body's energy", "take a few slow steps in the room to move your body", and "open the window to ventilate and breathe fresh air". For sleep disorders, provide sleep improvement suggestions such as "put down your phone an hour before bed and listen to soothing music".

[0030] (4) Family communication prompt template: pushed to family members, including an overview of the patient's current status and suggestions for companionship, such as "Your family member has been less active recently. We suggest you call or video call to greet them when it is convenient. A brief expression of concern will bring great support."

[0031] (5) Nursing follow-up reminder template: pushed to the nursing end, including patient compliance data and follow-up suggestions, such as "Patient Zhang's task completion rate in the past 3 days is 40%, and he has not responded to the push for 2 consecutive days. It is recommended to arrange a face-to-face follow-up to understand the reason".

[0032] Search matching rules: - Sleep abnormality marker → Prioritize searching for activation prompts of low-load behaviors related to sleep improvement; - Insufficient activity flag → Prioritize searching for low-load behavior activation prompts related to activity encouragement; - If the score for "Difficulty getting along with others" in PHQ-9 is ≥2 or the score for "Prefer to stay at home" in GDS is positive, prioritize searching for companionship encouragement tips or family communication tips; - Other situations → Search for emotional soothing tips.

[0033] Multi-platform content generation: - Patient-side prompts: 50-100 words, concise and gentle language, accompanied by voice broadcast; - Family member's suggestion: 100-150 words, including a brief description of the patient's condition and suggestions for accompanying the patient; - Nursing staff tip: 150-200 words, including data trends and professional follow-up suggestions.

[0034] 4. Push Control Module The push control module sets the push control parameters according to the attenuation level F and performs push within the specified time window.

[0035] Push control parameters include: (1) Push frequency: Based on a joint decision of frailty level F and depression risk level D, such as Figure 3 As shown.

[0036] - Early stage of weakness + high risk: 3 times daily; - Early stage of weakness + medium risk: twice daily; - Early stage of weakness + low risk: once every 2 days; - Weakness + High Risk: Twice daily; - Weakness + Medium or Low Risk: Once daily.

[0037] (2) Duration of a single interaction: - Early stage of weakness: ≤2 minutes (120 seconds); - Weakness: ≤90 seconds.

[0038] Interaction duration refers to the maximum time from when a push notification is triggered until the user completes their response or the notification is automatically closed.

[0039] (3) Push interval time: - Early stage of decline: The interval between two pushes is ≥3 hours; - Weakness: The interval between two pushes is ≥4 hours.

[0040] (4) Push time window: All objects are uniformly set to 06:00 to 22:00. Pushes are prohibited from 22:00 at night to 06:00 the next day to avoid disturbing rest.

[0041] (5) Interaction termination condition: When the number of consecutive failures or rejections is ≥2, the low load prompt mode switch is triggered (see feedback update module for details).

[0042] Push method: - Patient-side: Mobile terminal pop-up notification + voice broadcast, or smart bracelet vibration reminder; - Family side: Push notifications via mobile app; - Nursing terminal: Nursing station terminal message list reminders.

[0043] 5. Feedback Update Module The feedback update module collects user interaction feedback data on push notifications and automatically adjusts the push strategy based on the feedback, such as... Figure 4 As shown.

[0044] Interactive feedback data includes: - Task completion rate: Number of completions / Number of push notifications × 100%. Completion is defined as: the user clicks to view the prompt message and stays for ≥30 seconds, or completes a simple action in the prompt to confirm (such as clicking the "I have completed" button).

[0045] - Number of consecutive rejections: The number of times there is no response (no click within 10 minutes) or the page is quickly closed (staying for less than 10 seconds).

[0046] - Emotional self-rating: The push notification includes an option to rate your emotions from 1 to 5 (1 point is very depressed, 5 points is good mood), and users can choose voluntarily.

[0047] Feedback update rules: (1) Low-load alert mode trigger: When the number of consecutive incomplete or rejected transactions is ≥2, the system automatically switches to low-load alert mode. Low-load mode includes: - Reduce the push frequency to 50% of the original frequency (e.g., reduce from twice a day to once a day). - The number of characters in the notification message has been reduced to less than 50; - Remove interactions that require text input, leaving only the emoticon radio button or the "Yes / No" button; - Voice broadcast speed reduced to 0.8x.

[0048] (2) Periodic Assessment and Strategy Adjustment: Each assessment period lasts 7 days, during which PHQ-9 or GDS scores and objective activity data are re-collected. Data from before and after each period are compared. - If the PHQ-9 score decreases by ≥2 points, or the task completion rate increases by ≥20%, or the average emotional self-rating score increases by ≥0.5 points, the current push strategy will be maintained; - If the above indicators do not improve or worsen, adjust the combination of prompt types, increase the proportion of low-load behavior activation prompts and companionship encouragement prompts, and reduce the proportion of emotional soothing prompts.

[0049] (3) Exit low load mode: After entering low load mode, if the task completion rate is ≥60% for 3 consecutive days, the original push frequency will be automatically restored.

[0050] Example 2: Method Flow Example This embodiment provides a method execution flow based on the above-described apparatus, such as... Figure 2 As shown, the following example illustrates a home-based follow-up scenario in a certain community.

[0051] Application Scenario: A community health service center provides follow-up management services for depressive states to elderly residents aged 65 and above within its jurisdiction. The service recipients are elderly individuals who have completed initial screening for frailty and depression risk, have a FRAIL score ≥1, and a PHQ-9 score ≥5. The service tools are the elderly individuals' smartphones (with a dedicated health management app installed) and smart bracelets.

[0052] Method execution flow: Step S1: Data Acquisition On the first day, community nurses visited the target individuals to complete the FRAIL and PHQ-9 assessments and entered the data into the health management app. For example: - Subject A: Female, 72 years old, FRAIL=2 (fatigue+, limited activity+), PHQ-9=11; - Subject B: Male, 78 years old, FRAIL=4 (fatigue+, limited mobility+, difficulty walking+, number of diseases+), PHQ-9=16.

[0053] At the same time, a smart bracelet is provided to the recipient and linked to the APP to begin collecting objective activity data. The bracelet automatically syncs sleep duration, number of nighttime awakenings, and daily step count data to the APP daily.

[0054] Step S2: Status Assessment The system calculates based on the collected data: - Subject A: Weakness level F = Pre-weakness (FRAIL=2), Depression risk level D = Medium risk (PHQ-9=11); - Subject B: Frailty level F = Frail (FRAIL=4), Depression risk level D = High risk (PHQ-9=16).

[0055] Objective data from day 3 shows: - Subject A: Sleep duration 5.2 hours (<6 hours, marked as abnormal sleep), daily steps 1800 steps (<2000, marked as insufficient activity); - Subject B: Sleep duration 6.8 hours (normal), daily steps 800 (marked as insufficient activity), 3 points in the PHQ-9 item "difficulty in getting along with others" (≥2 points, marked as social need).

[0056] Step S3: Prompt generation The system generates a push plan based on the evaluation results: Subject A (Pre-deterior frailty + intermediate risk + sleep disturbances + insufficient activity): - Push frequency: Twice a day; - Prompt content: The first push (8:00 AM) will be a low-load behavior activation prompt for sleep improvement, such as "You didn't get enough rest last night. Take a 20-minute nap this afternoon, which will help restore your energy"; the second push (3:00 PM) will be a low-load behavior activation prompt for activity encouragement, such as "The weather is nice. Take a 5-minute walk in the community garden to get some exercise".

[0057] - The patient's input should be around 80 characters, accompanied by voice prompts.

[0058] Subject B (frail + high risk + lack of activity + social needs): - Push frequency: Twice a day; - Prompt content: The first push (9:00 AM) will be a reminder to spend time with your children, such as "You can call your children, chat for a few minutes, and listen to their voices"; the second push (4:00 PM) will be a reminder to activate low-intensity behaviors, such as "Go to the window at home, look at the scenery outside, and breathe some fresh air".

[0059] - The number of characters for patient prompts should be around 60 (to reduce the workload), accompanied by voice broadcast.

[0060] - At the same time, a notification is generated for family members: "Your father has been less active recently and may be in a bad mood. We suggest you call him when it is convenient for you. A brief expression of concern will be of great support." (120 characters) Step S4: Push Control The system executes push notifications at the set push time and controls the push parameters: Object A: - Single interaction duration ≤ 2 minutes; - The interval between two push notifications is ≥3 hours (7 hours between 8:00 AM and 3:00 PM). - Push notification time window: 06:00-22:00; - Push method: Mobile APP pop-up + voice broadcast.

[0061] Object B: - Single interaction duration ≤ 90 seconds; - The interval between two push notifications is ≥4 hours (7 hours between 9:00 AM and 4:00 PM). - Push notification time window: 06:00-22:00; - Push method: Mobile APP pop-up window + voice broadcast (speech speed 0.9x).

[0062] If the recipient does not click to view within 10 minutes of the push notification, the system records it as "No Response"; if the recipient clicks to view but stays for less than 30 seconds, the system records it as "Partially Completed"; if the stay time is ≥30 seconds or the recipient clicks the "I have completed" button, the system records it as "Completed".

[0063] Step S5: Feedback Update On days 3-5, the system collects feedback data: Subject A: A total of 6 push notifications were sent over 3 days. 4 were completed, 1 was partially completed, and 1 was not responded to. The task completion rate is approximately 67% (4 / 6). There were 0 consecutive rejections. The average self-evaluation score for emotion is 3.5. Feedback is positive; the current push notification strategy will be maintained.

[0064] Target B: Pushed 6 times in 3 days, completed 2 times, partially completed 1 time, and did not respond 3 times (including 2 consecutive days of no response on the 4th and 5th days). Task completion rate = 2 / 6 ≈ 33%, consecutive rejections = 2, and the average self-evaluation score for emotion is 2.0. Low load prompt mode is triggered.

[0065] Starting from day 6, the push strategy for object B will be adjusted as follows: - Push notification frequency reduced to once a day (50% of the original twice a day); - Reduce the number of words in the prompt to less than 50, such as "Drink a glass of warm water today, and take it slow"; - Cancel text input, only keep the single selection of emotion emojis (smiley face / neutral face / crying face); - Voice broadcast speed reduced to 0.8x.

[0066] From day 7 to day 9, the task completion rate of object B in low-load mode increased to over 60%, meeting the target for three consecutive days. On day 10, the system automatically resumed the original push frequency (twice a day).

[0067] Step S6: Periodic Assessment On day 7, the system reminded the community nurse to recollect the PHQ-9 score: - Target A: PHQ-9=9 (decrease of 2 points), task completion rate=67%, maintain the current push strategy; - Subject B: PHQ-9=15 (a decrease of 1 point, <2 points). Although there was a decrease, the magnitude was insufficient. The system adjusted the combination of prompt types, increasing the proportion of prompts for companionship and encouragement and prompts for communication with family members.

[0068] Reassessment on day 14: - Target A: PHQ-9=7 (down 2 points), task completion rate=75%, continue to maintain; - Subject B: PHQ-9=12 (down 3 points), task completion rate=58%, emotional self-rating score improved to 3.2, prompt type adjustment effective, continue execution.

[0069] Example 3: Verification Example Application scenario: During June-July 2025, a nursing home used the method of this invention to conduct a 28-day psychological support follow-up management for 20 elderly residents.

[0070] Sample characteristics: - Age: 68-85 years old, average 75.2 years old; - Gender: 9 males and 11 females; - Weakness status: 12th place with FRAIL score of 1-2 (early stage of weakness), 8th place with FRAIL score of 3-5 (weakness); - Depression status: Baseline PHQ-9 score ranged from 5 to 18, with an average of 10.5. Among them, 7 were at low risk (5-9 points), 9 were at medium risk (10-14 points), and 4 were at high risk (≥15 points).

[0071] Implementation method: - Provide each senior citizen with a smart bracelet to collect data on sleep duration, number of nighttime awakenings, and daily steps; - Deploy the device of this invention at the terminal of the nursing station in a nursing home to generate an individualized push plan based on the elderly person's frailty level and depression risk level; - Push methods: Nursing station terminal message notification + delivery by caregiver + wristband vibration reminder; - Re-collect PHQ-9 scores every 7 days and record task completion status daily.

[0072] Output metrics: (1) Trend of changes in scores on the Depression Scale - Baseline (Day 1): Average PHQ-9 = 10.5 points; - Day 7: Average PHQ - 9 = 9.1 points (average decrease of 1.4 points); - Day 14: Average PHQ-9 = 8.3 points (average decrease of 2.2 points); - Day 28: Average PHQ-9 = 7.6 points (average decrease of 2.9 points).

[0073] Of the 20 elderly participants, 18 showed a downward trend in their PHQ-9 scores (a decrease of 1-6 points), 2 maintained a relatively stable score, and none showed an increase. Among the 13 elderly participants in the medium-to-high risk group (baseline PHQ-9 score ≥ 10), 11 declined to a low-risk level (< 10 points), and 2 remained at medium risk but their scores decreased by 2-3 points.

[0074] (2) Changes in task completion rate - Week 1 (Days 1-7): Average task completion rate 56.8%; - Week 2 (Days 8-14): Average task completion rate 68.3% (an increase of 11.5%). - Week 3 (Days 15-21): Average task completion rate 74.2% (an increase of 17.4%). - Week 4 (Days 22-28): Average task completion rate 77.5% (an increase of 20.7%).

[0075] The task completion rate for the early-stage weakening group (FRAIL 1-2 points) increased from 58.3% to 82.1%; the task completion rate for the weakening group (FRAIL 3-5 points) increased from 54.2% to 70.6%. The significant improvement in compliance among the weakening groups demonstrates the effectiveness of the push control tailored to the weakening level.

[0076] (3) Effectiveness of push control - Nighttime disturbance incidents: 0 times (strictly enforced during the time window of 06:00-22:00, with no nighttime push notifications); - Low load mode trigger: 6 out of 20 elderly people triggered low load mode in the first week. After switching, the task completion rate increased by an average of 23.5%. 4 people resumed the original push frequency in the second week. - Complaints about exceeding the push frequency limit: 0 times (the daily push limit of ≤3 times has been enforced).

[0077] (4) Changes in self-rating of mood - Baseline: Average self-emotional rating of 2.4 points (1-5 subscale); - Day 28: Average mood self-rating score 3.6 (an improvement of 1.2 points).

[0078] The results show that the method of the present invention, by combining frailty assessment, objective activity data collection, individualized push control, and feedback closed-loop update, has achieved a downward trend in depression assessment scores, an increase in task completion rate, and improved compliance in elderly frail individuals. At the same time, it effectively avoids nighttime disturbances and excessive push issues, verifying the feasibility and effectiveness of the technical solution.

[0079] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of the invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for generating a psychological support cue for improving depressive symptoms in a frail elderly patient, characterized by, Applied to mobile terminals or nursing station terminals, targeting individuals aged ≥65 years with a Frailty Assessment Scale (FRAIL) score ≥1 and a Depression Symptom Risk Assessment Scale (PHQ-9) score ≥5 or a Geriatric Depression Scale (GDS) score ≥6, the following steps are included: S1 Data Collection: Collect frailty assessment data, objective activity data, and depression scale data of the target subjects through the terminal. The frailty assessment data includes scores of five dimensions of the FRAIL scale: fatigue, activity limitation, difficulty walking, number of diseases, and weight loss. The objective activity data includes at least one of the following: sleep duration, number of nighttime awakenings, daily steps, and activity duration. The depression scale data includes PHQ-9 score or GDS score. S2 Status Assessment: The frailty level F is calculated based on frailty assessment data, and the depression risk level D is calculated based on objective activity data and depression status scale data. The frailty level F is divided into two levels: pre-frailty and frailty, and the depression risk level D is divided into three levels: low risk, medium risk, and high risk. S3 Prompt Generation: Based on the frailty level F and the depression risk level D, retrieve matching prompt information from the preset psychological support prompt content library and generate a push plan. The psychological support prompt content library includes emotional comfort prompt templates, companionship and encouragement prompt templates, low-load behavior activation prompt templates, family communication prompt templates, and nursing follow-up reminder templates. S4 Push Control: Set push control parameters according to the attenuation level F and execute push. The push control parameters include push frequency, single interaction duration, push time window and interaction termination condition. The single interaction duration is set to no more than 2 minutes, the number of pushes per day is set to no more than 3, the push time window is set to 06:00 to 22:00, and the interaction termination condition includes the number of consecutive non-completions or rejections reaching a preset threshold. S5 Feedback Update: Collect interactive feedback data of the target object to the push notification information. The interactive feedback data includes task completion rate, number of consecutive rejections and emotion self-rating. When there are consecutive failures or rejections ≥ 2 times, automatically switch to low load notification mode, reduce the push frequency and update the notification type for the next cycle. In the low load notification mode, the number of words in the notification information is reduced and the push frequency is reduced to 50% of the original frequency.

2. The method of claim 1, wherein, In step S2, the calculation rules for the frailty level F are as follows: a score of 1-2 on the FRAIL scale indicates pre-frailty, and a score of 3-5 indicates frailty. The calculation rules for the depression risk level D are as follows: when using the PHQ-9 scale, a score of 5-9 indicates low risk, a score of 10-14 indicates medium risk, and a score ≥15 indicates high risk; when using the GDS scale, a score of 6-10 indicates low risk, a score of 11-15 indicates medium risk, and a score ≥16 indicates high risk.

3. The method of claim 1, wherein, In step S4, the rule for setting the push frequency based on the frailty level F is as follows: for individuals in the early stage of frailty and with a high risk level of depression, push notifications are sent 3 times per day; for individuals in the early stage of frailty and with a medium risk level of depression, push notifications are sent 2 times per day. For individuals in the early stages of frailty who are at low risk of depression, a push notification will be sent every two days. For individuals who are frail and have a high risk of depression, push notifications will be sent twice daily. For individuals who are frail and have a medium or low risk of depression, a message will be sent once a day.

4. The method of claim 1, wherein, In step S4, for objects with a weakening level, the duration of a single interaction is further limited to no more than 90 seconds, and the interval between two pushes is no less than 4 hours; for objects with a weakening level in the early stage of weakening, the interval between two pushes is no less than 3 hours.

5. The method of claim 1, wherein, In step S3, the retrieval and matching rules for the prompt information include: when the sleep duration is <6 hours or the number of nighttime awakenings is ≥3 times in the objective activity data, priority is given to retrieving low-load behavior activation prompt templates for sleep improvement from the psychological support prompt content library; when the daily step count is <2000 steps, priority is given to retrieving low-load behavior activation prompt templates for activity encouragement; when the score of social-related items in the PHQ-9 scale is ≥2 points or the loneliness-related items in the GDS scale are answered affirmatively, priority is given to retrieving companionship encouragement prompt templates or family communication prompt templates.

6. The method of claim 1, wherein, Step S3 further includes generating prompts for different receiving ends: prompts for patients include brief emotional reassurances and single-step behavioral suggestions, with a word count of 50-100; prompts for family members include an overview of the patient's current condition and companionship suggestions, with a word count of 100-150; and prompts for caregivers include patient compliance data and follow-up suggestions, with a word count of 150-200.

7. A device for generating psychological support prompts to improve depressive symptoms in elderly and frail patients, characterized in that, include: The data acquisition module is used to collect frailty assessment data, objective activity data, and depression status scale data of target subjects aged ≥65 years, with FRAIL score ≥1 and PHQ-9 score ≥5 or GDS score ≥6 through the terminal. The objective activity data includes at least one of the following: sleep duration, number of nighttime awakenings, daily steps, and activity duration. The status assessment module is used to calculate the frailty level F based on frailty assessment data and the depression risk level D based on objective activity data and depression status scale data. The prompt generation module is used to retrieve matching prompt information from the psychological support prompt content library based on the frailty level F and the depression risk level D, and generate a push plan. The psychological support prompt content library includes emotional comfort prompt templates, companionship and encouragement prompt templates, low-load behavior activation prompt templates, family communication prompt templates, and nursing follow-up reminder templates. The push control module is used to set push control parameters and execute push according to the attenuation level F. The push control parameters include push frequency, single interaction duration not exceeding 2 minutes, daily push frequency not exceeding 3 times, push time window from 06:00 to 22:00, and interaction termination conditions. The feedback update module is used to collect interactive feedback data such as task completion rate, number of consecutive rejections, and emotion self-rating. When there are ≥2 consecutive failures or rejections, it automatically switches to low-load prompt mode and updates the prompt type for the next cycle.

8. The apparatus of claim 7, wherein, The push control module includes a parameter adaptive adjustment unit, which is used to set different single interaction duration thresholds and push intervals for the early stage of attenuation or attenuation according to the attenuation level. For attenuated objects, the single interaction duration is limited to no more than 90 seconds and the push interval is no less than 4 hours.

9. The apparatus of claim 7, wherein, The prompt generation module includes a multi-terminal content generation unit, which is used to generate prompt messages with different content and lengths for the patient terminal, family member terminal, and nursing terminal respectively. The prompts for the patient terminal are 50-100 characters long, the prompts for the family member terminal are 100-150 characters long, and the prompts for the nursing terminal are 150-200 characters long.

Images