A study on requirement optimization of intelligent health care information system for rural elderly based on improved FKM-TOPSIS method

1.Introduction

With the advent of aging in China, older people’s physical and psychological healthcare problems are becoming increasingly severe, particularly in rural regions, where healthcare conflicts are more evident owing to the limits of rural areas. It is worth noting that data from the Seventh Population Census shows that China has a migratory population of 376 million, with more than 80% of the mobile population coming from rural areas, leaving about 40 million older people behind in rural areas, accounting for 37% of the rural elderly population. However, as many young rural people migrate to cities for work, working and caring for older adults who cannot care for themselves cannot be combined, leading to little workspace for many children caring for their parents [1]. Rural elderly left behind without family care often go to the village health center to deal with their illnesses simply or suffer from them, leading to irreversible consequences over time [2]. Although governments at all levels have implemented large-scale poverty alleviation measures for rural areas, such as the construction of nursing homes, because it is a poverty alleviation project, many townships share a single home for older people. Hence, the facilities and staffing are relatively simple and can only provide the most basic living, not even medical facilities. Thus the medical and pension issues of older people left behind in rural regions remain unsolved [3]. The older people left behind in rural areas need more acceptance of innovative products due to their literacy, age, and lack of smart devices, particularly in intelligent payment and smart medical care, which is a fatal blow to the rural older people. In this situation, developing intelligent monitoring wearable devices and information systems for the old rural left behind can effectively allow children away from home to stay informed about their parent’s health and lifestyle.

We must thoroughly understand the present demands of older rural residents before designing the information system and APP for smart wearable devices. User demands determine the path of product innovation and design, and meeting those needs entails raising user satisfaction. Propose the combined FKM-TOPSIS approach in this research. The product questionnaire is quantified and improved by the FKM approach, and the attribute categorization of user needs is more accurate than the outcomes of the conventional Kano model analysis. However, because this model’s questionnaire asks for positive and negative aspects, different people may have different perspectives. As a result, various uncertainties, such as the variety of evaluation indicators, the subjectivity of indicator weights, and the ambiguity of evaluation information, impact the evaluation results. And TOPSIS, on the other hand, is a multi-objective and multi-attribute efficient evaluation method in operations research that primarily evaluates the importance of product quality elements and makes full use of the user demand evaluation information obtained in the FKM model to rank the proximity of several evaluation schemes to the idealized object. Explore each demand for smart wearable devices as a solution in this study. SI and DSI are utilized as evaluation indices for each solution to assess the relative benefits and drawbacks of the available solutions. Combining the two models generates fresh possibilities for study into intelligent product design.

Additionally, we design the dynamic product interaction prototypes on the PC and mobile-based of the information system of smart wearable devices to confirm the objectivity of the user requirement satisfaction weighting using test data.

2.Literature review

Identifying consumer demands is the beginning point for the idea development phase of the new product development process. It directly influences consumer happiness and product quality [4].

On the one hand, given the many needs of rural elderly for intelligent products, research in the field of smart wearable products has shifted to focusing on the needs of a set of intelligent products, decreasing the cost of the elderly in rural regions, and popularizing intelligent wearable technology products [5]. Based on differences in exploratory perspectives, research on intelligent wearable technologies to help rural elderly’s lives at home and abroad is mainly distributed in three aspects: (1) The research carriers for intelligent wearable devices are primarily specific people or groups, such as the usage of wearable items for diabetes patients for blood glucose measurement, insulin infusion, and other autonomous treatment [6]. Such as Smart devices are implanted in a disabled person’s body to provide intelligent support and let them interact more effectively with their surroundings [7]. Smart wearable devices were employed in studies to remotely monitor the monitoring data of elderly individuals with atrial fibrillation [8]. However, a single smart wearable device is a significant challenge for rural older adults due to the economic pressure, laborious operation process, and the fact that a single device can only solve a single problem. So improving the quality of life first requires solving medical and health problems affecting all body parts. (2) Investigate the acceptability of intelligent wearable devices by older persons, for example, by developing a theoretical model to examine the factors that impact more senior adoption of innovative technology devices. According to Stephen M. Golant, intelligent wearable technologies have low knowledge of privacy protection for older adults [9]. Some academics have utilized two algorithms to increase privacy protection for smart device app users and increase older people’s adoption of intelligent devices. Both are research on the subjectivity of older persons’ reception of intellectual products [10,11]. With the emergence of rural issues such as fewer children, aging, remote rural locations, relatively poor infrastructure, and an insurmountable “digital divide” for older adults, the emergence of intelligent wearable products has alleviated the physical and psychological problems of the older adults, for example, smart devices and cloud diagnosis can help treat rural older adult’s diseases [12]. It belongs to the objective reality constraints that must accept intelligence to ensure the quality of life of rural older adults. (3) According to an analysis of older adults demands in rural areas, whether in China or other nations, considering such needs progressively seriously. In Tianjin, China, Lin Hongyan et al. examined a sample of elderly rural people. They discovered that depression was a severe issue among rural older persons, accompanied by poor sleep and other mental problems requiring immediate attention [13]. The chief physician of a test conducted by a joint geriatric rehabilitation department of a tertiary care hospital showed that the mental state of the older adults who regularly participated in socialization and outside communication was significantly better than those who did not [14]. Kshatri et al. used regression models to analyze rural elderly people with multiple chronic diseases in India and found that their frailty and dependence became more intense with age [15]. Kuo et al. discovered that the rural elderly had a greater risk of death from heart disease when comparing urban and rural Taiwan areas to investigate variations in variables impacting mortality [16]. Referring to the research literature above, it is known that physical and mental health issues continue to be the top priorities for rural older adults.

Moreover, in rural areas, the proportion of older adults who are defrauded is very high compared to those in urban areas, as Deliema et al. (2020), whose proportion of rural older adults who are tricked has gradually increased in recent years and depression has become an essential factor affecting the probability of being defrauded [17]. Empirical evidence by Wei et al. (2021) also found that individual financial literacy significantly reduces their likelihood of being tricked, and compared to the urban older adults, the rural low literacy level of older adults resulted in them being the most affected by fraud [18]. Manea et al. concluded that a significant risk factor for chronic diseases is a lack of exercise and using simple wearable devices to record data on changes in the body [19]. The results showed that moderate leisure activities could reduce the risk of chronic diseases. Combining these current conditions suggests that with the advent of aging, older adults’ daily life, recreation, and spiritual needs seriously affect their physical and mental health [20] and some literature shows that compared to urban areas, rural elderly are more willing to age at home [21]. All these phenomena provide a broad commercial market for technological products such as intelligent assistive devices and smart homes [22].

On the other hand, the Kano model is a standard method to identify customer requirements, and the use of the Kano model helps designers identify customers’ different functional needs and enables them to control quality and satisfaction more accurately in the product design and development process [23]. Many scholars have been expanding the integration method of the Kano model with other methods. For one thing, the traditional Kano model research methods are improved, such as the FKM (Fuzzy Kano Model) model, which quantifies and improves the five satisfaction levels of the questionnaire’s demand attribute classification the Kano model to ensure a more accurate attribute classification of user demand [24]. This method applies to all products that need to explore user needs and satisfaction types. For another, the Kano model is widely combined with various ways represented by TOPSIS to form a closed loop of mapping from customer requirements classification and definition to product architecture, which provides technical support and methodological guidelines for design. For example, To efficiently solve the evaluation problem, Shan Du developed the Kano-DEMATEL-TOPSIS method and applied it to logistics service supply chain (LSSC) management [25]. Chen et al. proposed a hybrid analytical method integrating the fuzzy Delphi method (FDM), Kano two-dimensional quality model, and theory of innovative problem solving (TRIZ) was proposed [26].

Previous studies on the needs of rural older adults have failed to integrate the two issues of “intelligent products” and “life needs of the rural elderly”. The information systems of intelligent wearable devices developed based on the psycho-physiological needs of rural older adults cannot only alleviate the health problems of the rural older adults but also play a comforting spiritual role for the older adults with its on-demand functions, and the rural older adults have a tremendous demand for chronic disease management, so the use of smart wearable products for primary monitoring and health care can prevent malignant changes of chronic diseases. Smart wearable devices in their intelligent mode, with less operation, less interaction, automatic all-around monitoring, and simple digital literacy for rural elderly, can be used [27–29]. Based on the above considerations, this paper uses the FKM-TOPSIS method to conduct a study on the actual needs of smart wearable monitoring devices for rural older adults, to effectively enhance their participation in intellectual development and enjoy the dividends of Internet development.

3.Integration analysis of the demand for innovative wearable products for rural older adults

Although the development of smart wearable devices is intended to solve the problem of older people in rural areas, the majority of users of its information system are village officials, village doctors, children of older people, and other participants in the direct care of older people, and their needs for the functions of the smart wearable products are highly consistent with those of older people. In this study, we investigate village authorities, village physicians, senior people, and their children as a group and analyze their common needs for smart wearable devices.

Jiangsu Province served as the overall sample and using a three-stage sampling approach to sample the population to determine the user demand for information systems for smart wearable devices for older individuals living in rural areas. The PPS unequal probability sampling technique was used in the first stage to sample eight Jiangsu Province prefecture-level cities based on the fraction of each town in Jiangsu Province; the random sampling method was used in the second stage. 36 villages were chosen randomly from a list of all the villages in the eight prefecture-level cities to represent the sample. Still, they had to satisfy two requirements simultaneously: they had to have a nursing home and a village medical and health facility. In the third stage, the research team contacted the village committee to obtain a list of committee members, village doctors, and older adults residing in nursing homes. Based on this list, they selected the village committee members, village doctors, children of older adults, and older adults who choose randomly as the final sample. A total of 110 samples—including 20 village committee members, who made up about 18% of the total sample, ten village doctors, who made up about 9% of the total sample, 43 older adult children, who made up about 39% of the total sample, and 35 older adults, who made up about 34% of the total sample—were ultimately collected for this study. We conducted an interview study on the daily needs of rural older adults, and these interviewees were potential users of intelligent wearable products. The researcher randomly interviewed the children of older people and others after arriving in the rural areas, and interviews with village doctors and village council members were pre-arranged. The research proceeded quite well since these problems are more prevalent and perceived easily daily.

The interviews with the village doctors focused on the health of older adults in rural areas, including the number of older adults with underlying diseases (hypertension, hyperlipidemia, hyperglycemia, etc.), the level of older adults’ knowledge of their illnesses, the complementary treatments that were available for the older adults with underlying conditions, and first aid for the older adults. It took around half a day to complete.

The interviews with village committee officials covered topics such as the number of older adults in the village, their physical and psychological conditions, how many people have experienced fraud in the past and how resolved it, what measures the village committee has taken to deal with older adults in an aging society, how the spiritual life of older adults in rural areas, and so on. It lasted approximately half a day.

The interviews with the children of older adults focused on what concerns the children have about the older adults’ problems, whether they plan to be involved in older adults in the future, which older adults’ assistance products they want to popularize, what kind of elderly care model the children of older adults believe their parents prefer, what special needs they have for the future of specialized elderly care, and so on. It took around 2–3 h.

The randomly selected older adults are primarily from rural areas, and the research takes about 4-5 hours. It asks questions like what kind of elderly care mode they prefer, what aspects of life they believe are more difficult right now, what types of help the older adults most need right now, what kind of expectations they have for their old age, etc.

To verify the study’s adequacy and objectivity [30]. We interviewed experts specializing in the information system for smart wearable items and other relevant practitioners using the Delphi expert survey method. These professionals comprised front-line intelligent product salespeople, product managers from firms that manufacture intelligent products, academics with backgrounds in information systems from universities, as well as physicians and nurses. After several rounds of gathering expert opinions on the issues raised and following repeated solicitation, generalization, modification, evaluation, and classification, resulting in a platform architecture system, the research process needs to be informed of each requirement’s feasibility and technical difficulty in addition to determining the reasonableness of the system’s user requirements. And evaluated and categorized each functional attribute and resulting in a platform architecture system that includes five categories of Medical Services (MS), Communication Functions (TC), Anti-fraud System (FS), Leisure and Entertainment (ET), and Sports and Health (SH), with a total of twenty-five functional requirements, a complete understanding of the full picture of the information system of smart wearable products. As shown in Table 1.

4.User research and data collection

Due to their lack of knowledge and restricted vision, it is difficult to fully and adequately identify the demands of the rural old, and older adults have a wide variety of needs. This article employs a questionnaire and in-depth user interviews to acquire original consumer demand information. The fuzzy clustering approach is then used to classify the original user demand information to get new user demand information, thereby reducing the effects of user demand information clutter on user demand analysis.

4.1.Survey questionnaire setting

Noriaki Kano, a Tokyo Institute of Technology professor, established the Kano model as a method for categorizing user demands based on user satisfaction [31,32]. The Kano model allows for qualitative and quantitative study of user requirements, exploring the breadth and depth of user demands, comprehension of the user needs hierarchy, and identification of critical factors to increase satisfaction. The Kano model organizes user requirements into five categories based on the non-linear connection between product characteristics and user satisfaction. M means Must-be Demand: customer satisfaction does not rise when providing this demand but considerably reduces when it is not; O means Performance Demand: user satisfaction will increase when providing these demands, whereas it will fall when it is not; A means Attractive Demand: it falls under an unanticipated user demand; if not provided, user satisfaction won’t be affected, yet, if it is provided, user satisfaction will be significantly higher; I means Indifferent Demand: user satisfaction will not vary whether this demand is met or not, and users do not care; and R means Reverse Demand: users do not have this demand, and if provided it, user satisfaction will decrease. As shown in Fig. 1.

Fig. 1.

Kano model user requirements classification chart.

The original Kano model questionnaire used five degrees of possibility to answer questions such as “Like”, “Deserved”, “Indifferent”, “Tolerated”, and “Disliked”. Some rural older adults’ subjective consciousness does not accurately represent their will while making decisions, resulting in ambiguity in demand survey data from some consumers. Thus it is suggested to utilize fuzzy interval values to describe the users’ sentiments instead, which is more accurate for ambiguous situations. Therefore, the FKM model questionnaire quantifies and improves the five possibility degrees of user needs [33,34], and the total score of the five levels of options is set to 1 and represented by a number between 0 and 1, making the attribute classification of user demands more precise. Table 2 depicts the initial Kano questionnaire, and Table 3 illustrates the improved FKM questionnaire.

Table 2

Example of KANO model questionnaire

Design elements	Number	Elemental analysis	User requirements	Like	Deserved	Indifferent	Tolerated	Disliked
Medical services (MS)	MS1	Call family doctor	Provide		□	□	□	□
			Not provide	□	□	□	□	□
	MS2	Review reminder	Provide	□	□	□	□	□
			Not provide	□	□	□	□	□
	MS…	……	…	…	…	…	…	…
	MS9	Two QR codes and one nucleic acid result	Provide	□	□	□	□	□
			Not provide	□	□	□	□	□

Table 3

Example of FKM questionnaire

Design elements	Number	Elemental analysis	User requirements	Like	Deserved	Indifferent	Tolerated	Disliked
Medical services (MS)	MS1	Call family doctor	Provide	0.7	0.2	0.1	0	0
			Not provide	0	0	0.1	0.3	0.6
	MS2	Review reminder	Provide	0.6	0.3	0.1	0	0
			Not provide	0	0.1	0.1	0.2	0.6
	MS…	……	…	…	…	…	…	…
	MS9	Two QR codes and one nucleic acid result	Provide	0.3	0.5	0.1	0.1	0
			Not provide	0	0.2	0.1	0.4	0.3

5.Functional satisfaction analysis based on the TOPSIS method

TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) is a multi-generic decision-making system that uses a multi-generic ranking mechanism to approximate idealized solutions [36]. The researchers used the TOPSIS method to fully utilize the information obtained in the FKM model for evaluating the user requirements of the information system of intelligent wearable devices for rural older adults and ranking several evaluation solutions based on their proximity to the idealized objects. Each demand in this work serves as an evaluation solution, and SI and DSI are evaluation indices for each scheme. As a consequence, comparing the relative benefits and drawbacks of existing programs and the assessment findings can more precisely depict the gap between the evaluation objects. Following that, the programs are ranked based on the size of the cumulative evaluation value. And in this paper, the weights of the two indices are set to be equal or 0.5 for each. The following are the steps in the TOPSIS approach for making decisions on functional satisfaction.

Step 1, construct the evaluation matrix of users’ satisfaction with each function.

In the above matrix, n is the number of functions of intelligent wearable products, n = 25; m is the number of evaluation indicators, m = 2.

In step 2, the functional satisfaction evaluation matrix is weighted and normalized to form a weighted normalized decision matrix, i.e.

In Eq. (3), 𝜔_j is the weight of the evaluation index, each is 0.5.

Step 3, calculate the weighted normalized evaluation matrix’s positive and negative ideal solutions.

Step 4, calculates the Euclid distance of positive and negative ideal solutions.

Step 5, calculates the relative closeness of each function.

According to Eqs (3)–(5), the normalized user demand satisfaction matrix and its positive and negative ideal solutions are obtained, and the results are shown in Table 7.

The Euclid distance and the relative closeness Vi values of positive and negative ideal solutions for each user demand satisfied were calculated using Eqs (6)–(8). Based on the normalized user satisfaction decision matrix and positive and negative ideal solutions, the results are shown in Table 8.

Based on the relative proximity of user satisfaction as calculated by the TOPSIS method. each function demand is prioritized in the following order: MS5 >ET2 >TC2 >MS3 >MS8 >MS6 >MS1 >MS2 >MS7 >ET4 >TM2 >SH4 >MS9 >TM3 >TC1 >SH2 >TC3 >TM1 >SH1 >MS4 >SH3 >ET3 >ET1 >SH5 >ET5.

6.PC and mobile-based product prototype testing for smart wearable device information system

With the aid of the prototyping software “AXURE”, we quickly created a dynamic product interaction prototype for the PC and mobile-based of the smart wearable device information system to verify the scientific nature of the satisfaction weighting of user requirements [37,38]. As shown in Figs 2, 3, the product prototype resembles a finished item. It dynamically links the entire system’s interaction, usage process, and specific contents to create a complete set of production process systems.

The system’s interface prototypes may interact dynamically according to the operation flow, giving users a realistic operating experience without requiring numerous scripting. For example, the main functional data display interface on the computer-based and the heart rate data page on the mobile-based—heart rate detail page—triggered the 120 emergency calls page after a sudden abnormality in the data. As shown in Figs 2, 3, and other processes, these prototypes provide a clear, logical relationship for the system development, ensuring the product’s usability. After generating the system prototype, we shared the prototype through the AXURE software to retrieve the link and QR code, which was sent to those who had previously participated in the research, including village officials, village doctors, children of older people, and older adults with smartphones, who clicked on the link or scanned the QR code on their cell phones to access the system. Researchers gave testers test accounts, performed simulations using the test data to provide users real-world access to the product, and clicked the “Like” button for each favorite demand. Following the completion of the test, monitor the number of “Like” for each user’s demand by the prototype’s back-end and rate the satisfaction level of all user requests. The more increased the “Like” number, the higher the user’s contentment with the demand.

Fig. 2.

Screenshot of the PC-based display interface of the information system prototype of intelligent wearable devices for rural older adults.

Fig. 3.

Partial screenshot of the mobile-based interface of the information system prototype for smart wearable products for older adults.

Combined with Fig. 4, by the amount of user “Like” for each demand, we can see that the most important functions with the demand are concentrated in emergency assistance for rural older adults. Such as 120 call (MS5), Call Emergency Contacts (ET2), Positioning System (TC2), Blood Pressure Testing (MS8), Heartbeat Pulse Monitoring (MS6), etc. These functions can effectively reduce the health risks of older adults in critical emergencies, and improving these functions is therefore significant. As a result, improving the usability and quality of the aforementioned top-ranked functionalities is a significant issue for smart wearable product developers. On the contrary, non-emergency life care tasks, such as Weather Forecast (SH5), News Radio Broadcast (ET1), Opera Music Playback (ET3), Energy Consumption (SH3), Sleep Monitoring (MS4), Real-time Pedometer (SH1) and so on, are substantially less relevant to users’ demands. Although these duties contribute to the health monitoring of older people, they are not extraordinarily vital and are essentially auxiliary support functions. Even the absence of these functions does not cause severe discomfort to older adults in some circumstances. Therefore, improving these functions does not significantly impact the enhancement of user satisfaction with smart wearable devices in general.

Fig. 4.

Screenshot of user demand “Like” data of the information system’s prototype for smart wearable products for rural elderly.

7.Conclusion

This paper constructed a service system for elderly-friendly smart wearable products by dividing the functional types into five major categories, comprising twenty-five functional requirements. On this basis, the Kano model analyzes the demand levels of this product’s functional services to theoretically dig deeper into the demand centers of rural elderly groups. To use the TOPSIS method to calculate the relative closeness of user satisfaction and the priority weighting of functional demands to uncover the core demands and hidden needs of older people in rural areas for age-friendly smart wearable products to promote the more profound development of these products. This study can give critical information for planning, developing, or optimizing the demand services of age-appropriate smart wearable products by addressing the diverse needs of the aged population for intelligent wearable devices. The paper’s disadvantage is that: while the research group is primarily rural older adults, the specific situations of rural older adults are not subdivided, such as rural elderly left behind, elderly depending on their children, entirely self-care elderly, crippled elderly, etc. The demands of older people are different in different contexts. Therefore, to deepen the refinement of the user group in this work and the follow-up study, we may build a research questionnaire based on the focus on the needs of different groups.

Furthermore, while this paper addresses the classification of functional categories, it does not pay specific attention to disparities in demand for age-appropriate smart wearable products by gender. As a result, the follow-up study should further refine the distinct demand groups of the rural older people, achieve product function differentiation and customization, and always pay attention to the breakthrough of new age-appropriate technology and the dynamic changes in the new demand of the rural older people.