A Study on the Classification of the Transport Needs of Patients Seeking Medical Treatment in High-Density Cities Based on the Kano Model

Abstract

Against the background of traffic conflicts arising due to the highly concentrated population in high-density cities, this study aims to systematically identify the core transport needs of patients awaiting medical treatment; based on the theory of the Kano model, we construct a measurement system relating to patient transport needs when awaiting medical treatment that encompasses multiple levels. Taking 10 large general hospitals in Guangzhou as samples, this study collected data through questionnaires and auxiliary interviews, using the importance–sensitivity analysis method to accurately measure the degree of patient needs for each influencing factor of the transport environment for medical treatment. The study found that, among the primary needs (core basic needs), the perfection of public transport (which directly affects the convenience of medical care) is the core need with the highest degree of demand. Among the second-level needs (refined categorised demand levels), specifically relating to important attributes (essential needs), priority attention should be given to patient diversion, hospital–city connection, and corridor settings. As concerns the high value-added one-dimensional attributes (desired needs), focus should be placed on controlling health and safety distances and guiding the flow of medical treatment, while for high glamour attributes (glamour needs), primary consideration should be given to crowd distribution, stopping and resting, and direct access to the ground floor. The group difference analysis (grouped by emotional state, transport mode, and group type) showed that the first-level demand sensitivity ranking was highly consistent, and the second-level demand for urban connectivity, convenient transfer, and direct underground access were also common priorities. This study is the first to introduce the Kano model into the analysis of high-density urban healthcare transport systems, providing a clear basis for the grading of demand for the design of the transport environment for patients’ medical care. This is of great practical value for alleviating congestion and improving the resilience of emergency response in mega-cities in relation to medical care.

1. Introduction

Good traffic organisation is essential to ensure smooth access and enhance the patient experience [1,2,3,4,5]. In the aftermath of the COVID-19 pandemic, healthcare management organisations require hospitals to have the ability to switch between acute and emergency situations, with the need to coordinate the organisation of functions and traffic control in times of disaster. Optimising the design of hospital transport environments has also been shown to help strengthen infectious disease prevention and control and reduce public health risks [6,7]. The problem is particularly acute in high-density cities, where the urbanisation rate of China’s resident population reached 67% by 2024; the over-concentration of the population in China’s 22 mega-cities will dramatically amplify the pressure on the allocation of transport resources for healthcare [8,9]. In the face of multiple challenges such as land resource constraints, an ageing population, and surging outpatient volumes, there is an urgent need to construct a full-factor evaluation system in relation to the transport environment for patient care, as well as to identify the key influencing factors in high-density urban environments, in order to promote efficient and orderly access to healthcare for patients.

Currently, research on the indoor and outdoor traffic environment of hospitals focuses on the following three aspects: (1) Research on the influence factors of wayfinding efficiency on traffic environment. Wayfinding efficiency is a key factor in shaping the traffic environment of hospitals, and its influencing factors cover indoor signage, landscape environment, and spatial form [2,10]. However, researchers have focused more on exploring the impact of wayfinding efficiency on patients’ emotions about medical care [11]. Additionally, some studies have conducted a more in-depth exploration, systematically integrating and carefully categorising the relevant influencing factors, particularly in relation to adding key elements such as the spatial organisation of traffic and the characteristics of nodes [12]. On this basis, this study further clarifies the interactions between multiple factors such as wayfinding subjects (e.g., patients and family members), the behaviour of others and hospital management, and the traffic for medical care [13]. (2) A study of the factors influencing the transport environment on the patient experience. Patient experience requires that the hospital transport environment be designed for convenience, comfort, and healing [14,15]. The accessibility of the hospital’s transport environment should be considered in conjunction with the variability of patients’ needs due to differences in their illnesses, financial incomes, and age groups [16]. Existing studies have focused on the impact of public transport environment creation on patient psychology [17] or on the creation of impact indicators focusing on overall satisfaction with patient transport needs in the context of the healthcare experience [18]. (3) A study of the factors influencing accessibility to healthcare on the transport environment. Accessibility provides scientific support for the optimal allocation of medical facilities and has a profound impact on public health [19,20]. Along with the trend towards granularity in the new data environment, accessibility studies based on quantitative methods are increasingly focusing on meeting the core transport needs of patients [21]. In order to objectively identify the physical accessibility issues faced by patients in the healthcare process and to highlight the urgency of environmental improvements [22], studies have been conducted to establish the social link between transport organisational structure and accessibility to healthcare by deconstructing and analysing patient transport to healthcare facilities [23,24].

Existing research emphasises the need for hospital design to take into account the multidimensional factors of patient needs for medical transport and to integrate them into the environment; however, the methodology mainly relies on qualitative analysis, with insufficient techniques for combining qualitative and quantitative aspects. The research content focuses on the perspective of the medical process, with insufficient in-depth consideration of the differences between the needs of the patient body at the physiological, psychological, and behavioural levels [25,26]. The key research gaps are the lack of priority stratification (essential/desired/attractive attributes) relating to transport demand for patient medical care in high-density cities, as well as the failure to establish a correlation model between demand sensitivity and resource allocation. To this end, this study aims to use the Kano model to classify 25 transport demand attributes, determine their demand sensitivity ranking by calculating the Better–Worse coefficient, and propose a phased optimisation strategy in conjunction with the degree of importance, in order to expand the relevant theoretical and technical approaches.

2. Research Design

In this study, a two-dimensional cognitive model of the current status of patient demand characteristics for medical transport and their satisfaction is established through the Kano theoretical model. The theory was formally proposed by Noriaki Kano, a professor at Tokyo Institute of Technology and a famous quality management expert, in the 1980s in the journal; the model was inspired by Herzberg’s two-factor theory. The Kano model is based on user needs and focuses on the change in respondent satisfaction through the forward and backward questions of the same need type, i.e., having and not having a certain need. It primarily focuses on the non-linear relationship between product performance and user satisfaction, and quantifies the type of requirement needed in order to assist with decision-making [27]. Compared with traditional methods, SERVQUAL/IPA relies on the assumption of a linear correlation, which reduces needs to a proportional ‘importance–satisfaction’ relationship. The Kano model reveals the asymmetric mechanism of ‘absence triggers strong dissatisfaction’ and ‘presence brings surprises’ through five-dimensional classification (absence triggers dissatisfaction, expectation, charisma, etc.). Secondly, although AHP/GIS is good at weight allocation and spatial analysis, it ignores the dynamics and group differences in the nature of demand (e.g., elderly patients regard resting chairs as a basic need, while young people are categorised as aspirational), whereas the Kano model quantifies the Better–Worse coefficients of different groups of people through positive and negative questionnaires to provide direct guidance for the optimal allocation of resources. Thirdly, its proposed law of ‘shifting demand attributes’ (e.g., during an epidemic, the reservation system shifts from glamour to basic) can provide a decision-making framework for responding to public health emergencies.

Current Kano modelling research techniques focus on the following three levels: (1) Single Kano model application: the use of the Kano model to assess customer satisfaction by Rozaq [28] and Kumar identifies important attributes of assistive products for people with motor disabilities [29], while the work of Hu suggests optimising commercial space in the community [30]. (2) Kano modelling combined with other methods:Lu,Jiang,Jia, Chen and Wang combined the Kano model with the two-dimensional quality method, service quality, IPA, QFD, and semi-structured interviews to study visitor centre satisfaction, leisure and fishery service quality, underground atrium design, customer service, and seating comfort, respectively [31,32,33,34,35]. (3) Development of a new framework based on the Kano model: Keyvanfar developed an adaptive behavioural satisfaction index [36], Liu proposed an S-Kano model [37], and Yang developed a BERT-TCBAD-Kano-based requirement analysis methodology [38].

In healthcare service research, Wan-I Lee used the Kano model to analyse the relationship between healthcare service quality and patient satisfaction in Taiwan, finding that patients were most concerned about doctors’ nursing care [39]. Susanne Gustavsson incorporated the patient’s multiple role perspectives into the Kano application, combining the views of patients, relatives, and healthcare professionals to identify a wide range of patient needs [40]. D.C. Ferreira used the Kano model and its improved version to assess the satisfaction of Portuguese patients with the National Health System and medical appointment services in public hospitals [41]. The above studies show that the Kano model has a wide range of applications in hospital health system satisfaction improvement, as well as being applicable to the study of patient needs in transport environments. Advantages of the Kano model in the study of patient needs in the transport environment include the following: (1) Quantification of the attributes of the needs. The Kano model can quantify the qualitative needs of patient transport, scientifically assessing the extent of the needs. (2) Identification of the priorities of the needs. The Kano model can classify patient transport needs into essential needs, desired needs, glamour needs, etc., so as to accurately identify and prioritise the needs. (3) Expanding the depth of the study. The Kano model, combined with the Delphi method, QFD, two-dimensional quality method, and other research tools, can continue to expand the depth of the study of the healthcare transport environment.

The technical route of the research firstly establishes a composite evaluation index system by identifying the demand attributes of the transport environment and completing the Kano model demand questionnaire, which was determined to be comprehensive and targeted according to the pilot study and expert group discussion. Then, the data results are classified according to the initial screening of demand types. Finally, each demand attribute is classified and evaluated in a step-by-step manner using the Better–Worse coefficient of demand, importance, sensitivity, and group type difference analysis (Figure 1).

2.1. Research Sample

Guangzhou (

Table 1. Statistics on relevant information of Guangzhou’s districts.

Name of the Districts	Area (km2)	Population (Unit: 10,000)	Population Density (10,000 people/km2)	Urbanisation Rate (%)	Level 3 Hospital (Units: One)	Number of Targeted Treatment Hospitals (Units: One)	Sample Distribution (Units: One)
Yuexiu	33.80	96	2.840237	100	20	5	4
Tianhe	96.33	223.8	2.323264	100	8	2	2
Haizhu	90.40	176.83	1.956084	100	7	1	1
Liwan	59.10	113.3	1.91709	100	7	0	0
Baiyun	795.79	366.68	0.460775	81.41	14	2	2
Panyu	786.15	282.29	0.359079	90.81	5	0	0
Huangpu	484.17	122.21	0.252411	94.37	5	1	1
Huadu	970.04	172.87	0.178209	70.82	3	0	0
Nansha	803	96.79	0.120535	75.02	2	1	0
Zengcheng	1616	158.67	0.098187	74.81	7	0	0
Conghua	1974.50	73.26	0.037103	53.89	1	0	0

), one of the mega-cities in China and the capital of Guangdong Province, has a resident population of 18.978 million. Although the hospital scale and the number of beds per capita are close to the level of developed countries, and the spatial distribution of tertiary-level hospitals has shown a balanced trend (based on the standard deviation ellipse analysis of GIS) (Figure 2), the combination of on-site research has revealed a problem relating to the high degree of concentration of patient access to healthcare, and the complexity of conflicts in the traffic environment remains serious. This study focuses on a sample of 10 large general hospitals with an average daily outpatient volume of more than 8000, which is highly concentrated to reflect the typical traffic dilemma of medical care under the superimposed effect of high urbanisation rates and ultra-high densities. (1) High urbanisation and population pressure in the sample area: in the sample, four hospitals are located in areas where the urbanisation rate has reached 100%, which means that all the space has been included in the built-up urban area system. The average population density is as high as 23,700 people per square kilometre, far exceeding the world’s recognised standard for densely populated areas; the areas described by the two hospitals in the sample are the city’s core areas with the largest population bases. This extreme urbanisation and high population density is the core background factor that leads to a high concentration of demand for healthcare, far exceeding that of the average area. (2) Core transport environmental problems faced by the sample hospitals: In total, 8 of the 10 sample hospitals are located in the historical old central urban area and are generally facing problems such as constraints on spatial resources (tight land use area), external traffic congestion (insufficient carrying capacity of the roads), and internal environmental pressures (crowded space for medical consultation and waiting). (3) Public health emergency attributes of the sample hospitals: The selected samples are all designated hospitals for the treatment of COVID-19 and are responsible for the emergency response to major public health emergencies; the vulnerability of their transport environments is even more pronounced in extreme cases (Figure 2).

Kim pointed out that the improvement in the transport environment for medical care contributes to the psychological healing of patients, emphasising the need to build transport links between the external and internal spaces of hospitals [42]. Arakawa Martins further subdivided the exterior space of the hospital into seven functional zones (site entrance zone, vehicular and pedestrian zone, parking zone, building entrance zone, internal and external connectivity zone, external activity zone, and service zone) [43]. In summary, the samples selected for this study fully reflect the typical characteristics of patient transport needs in terms of the number of beds, site area, building density, average daily outpatient volume, and other core indicators, and are representative of the study (see Appendix A).

2.2. Survey Questionnaire Setting

2.2.1. Factors

Based on the connotative characteristics and evaluation directions of the transport needs of patients in high-density urban areas, this study systematically constructs an evaluation system based on the principles of ‘combining comprehensiveness and typicality, and balancing scientificity and feasibility’ in the selection of indicators. The logic of the study is progressive, including the following:

a.

Demand analysis and principle anchoring, focusing on the intrinsic connection between patient experience attributes and material attribute indicators, clarifying the evaluation objectives and selection criteria.

b.

Field research and qualitative mining through hospital visits and semi-structured interviews with patients’ families to clarify the pain points and key conflicts.

c.

Literature evidence-based initial screening indicator set, integrating existing research results to form a basic indicator pool.

d.

Expert multidimensional validation and optimisation, from the importance of indicator selection, feasibility, sensitivity (using a 5-point Likert scale), three-dimensional Delphi assessment, and the use of Kendall’s harmony coefficient to test the consensus degree of the indicators, to ensure that the system is scientific and reliable.

The pathway connects the four components of the evidence chain—theory–practice–literature–expert—determining the selection of indicators from theoretical derivation, empirical modification, and quantitative verification.

(1)

Semi-structured interviews

Through on-site interviews with ten representative samples, empirical evidence was provided for the construction of a medical transportation indicator system. First, the frequently mentioned negative evaluation items in the table (such as ‘distance from the city centre’ (11:3), ‘ground-level parking spaces’ (4:16), and ‘traffic signage’ (5:11)) directly point to the need to incorporate core indicators such as ‘location accessibility,’ ‘static traffic capacity,’ and ‘clarity of the guidance system’ into the system. Secondly, the prominent positive evaluations of ‘skywalks’ (18:12) and ‘medical streets’ (14:5) validate the idea that ‘indoor–outdoor circulation efficiency’ should be established as a key soft indicator. Furthermore, negative feedback on specific scenarios such as the ‘passenger drop-off system’ and ‘direct access to medical care from the basement’ (5:8, 3:8) requires that indicators be refined to operational levels such as ‘vertical traffic connectivity’ and ‘passenger drop-off area turnover capacity.’ This statistical table, by mapping real demands and pain points, provides data anchors and validation foundations for constructing a hierarchical (macro-level location–meso-level circulation–micro-level nodes), problem-oriented medical transportation indicator system (see Appendix B).

(2)

Evidence-based literature

Currently, research on the medical traffic environment involves many disciplines such as transportation, environmental behaviour and psychology, and management; an efficient evaluation index system has not yet been established. Pei research on the integration of the urban road network, hospital services, and emergency management has shown that in the face of hazardous events, the hospital traffic and healthcare system needs to have resilience characteristics [44]. Ghosh’s study determined accessibility, quality of care, safety, and services to be major influencing factors when patients choose hospitals [45]. Lather and Bayraktar sari et al. studies explore the spatial layout of hospital units and how this affects the efficiency and effectiveness of healthcare delivery [46,47]. Douglas and Jiang et al. noted that personal space, a welcoming atmosphere, a supportive environment, and access to external areas and recreational facilities are essential to a patient’s health and wellbeing [48,49]. Several studies have shown significant differences between public and private transport in terms of accessibility to healthcare, which need to be incorporated into hospital accessibility models, and their impact should be accurately measured [19,50,51]. The study of Yazdani found that the use of air transport can significantly improve evacuation requirements when epidemics are detected in hospitals [52]. Furthermore, the study of Neustatel and Ji suggested that parking problems in hospitals lead to delayed patient access, stress, and poor access experiences [53,54]. Zhang found that the size of the outpatient clinic and the number of hospital beds did not have a significant effect on the distribution of footfall according to a survey of 2443 patients in three hospitals [55]. Chen proposed that the traffic environment of outpatient buildings should focus on spatial proximity, station building relationship, and reduction in pedestrian flow interference [56]. Wei study points out the importance of designing pedestrianised transport environments for infection control [57]. Moreover, Ji solves the vehicular traffic problem involved in the delivery of medical supplies in home healthcare services by optimising the model [58]. Zou’s simulation of hospital evacuation found that hospitals should prioritise the evacuation of wheelchair users to improve the efficiency of emergency evacuation [59]. Meanwhile, Trigo and Fumihito explored transport measures to optimise emergency ambulance patient treatment times using GPS data [60,61]. Lv solved the problem of emergency vehicle networking data, improving the speed and reliability of emergency vehicle response, as well as the efficiency of emergency treatment on the way to the doctor [62]. Madelin used simulation models to optimise the internal distribution network of a hospital, suggesting specific improvements such as reducing the frequency of deliveries and optimising routes [63]. Liu and Xin optimised a medical waste transport route model using Anylogic 8.2.3 Personal Learning Edition human flow simulation software and a hybrid dynamic approach, respectively, in order to ensure the fast and safe transport of medical waste during an outbreak [64,65]. Boucherie, Richard developed a number of complex discrete event simulation tools to optimise existing logistics processes in and around hospitals [66].

Based on the characteristics of the medical transport demands of high-density urban areas, the pain points of patient demand were analysed through 10 samples of field research, and a pool of indicators was initially constructed by combining these with literature-based evidence, before being validated using the Delphi method, which measured the importance of the selected indicators (w = 0.635, χ2 = 228.597, p < 0.000), as well as the feasibility (w = 0.759, χ2 = 273.391, p < 0.000) and sensitivity (w = 0.654, χ2 = 235.401, p < 0.001) of significant consistency. Finally, a multilevel evaluation system containing 4 first-level indicators and 25 s-level indicators was determined in order to achieve the closed-loop optimisation from demand mining to the calibration of the indicator selection (

Table 2. Indicators of transport environmental needs for sentinel hospitals.

Level 1 Requirements	Level 2 Requirements	Explicit Explanation	No.
Public transport (A)	Traffic type	Mode of transport travelling to and from the hospital (e.g., special buses for medical appointments)	A1
	Traffic frequency	Intervals between metro or bus journeys	A2
	Transport station	Number of routes served by public transport stops	A3
	Public evacuation	Improvement of pedestrian circulation at hospital entrances and exits	A4
	City connection	Consideration of urban transport network design, public transport access, etc.	A5
	Air accessible	Establishment of an emergency aircraft apron to cater for special medical situations	A6
Private transport (B)	Long-term parking	Adequate aboveground and underground car parking spaces to meet patient parking needs for medical treatment	B1
	Short-term parking	Provision of dedicated lanes for stop-and-go motor vehicles	B2
	Direct underground	Patients can take the lift directly from the underground car park to the outpatient hall on the ground floor.	B3
	Convenient pick-up and drop-off	Patients can take a direct bus to the lift entrances and exits of the hospital’s outpatient and inpatient departments	B4
Pedestrian transport (C)	Pedestrian environment	The pedestrian and landscape environments of hospitals should be interconnected	C1
	Pedestrian–vehicle separation	Separate footpath and motorway routes	C2
	Patient–doctor triage	Separate entrances for doctors and patients	C3
	Patient triage	Separate entrances and exits to and from hospitals for patients with non-communicable diseases and patients with communicable diseases	C4
	Distance control	Strict control of sanitary spacing between walking paths and pollution sources	C5
	Rest stop	To meet the needs of patient activities, rehabilitation training, and family members’ waiting and resting needs	C6
	Sunshade and heat insulation	Suitable canopy and gallery design can provide sunshade and heat insulation	C7
	Guide	Clearly signposted areas make it easy for patients to reach different parts of the hospital	C8
	Security check	Patients are required to undergo a security check to enter the hospital (including body temperature, carry-on items, etc.)	C9
	Public evacuation	Ensure that all personnel can evacuate to a safe area in a rapid and orderly manner	C10
Medical transport (D)	Infectious disease area independence	Ensure the effectiveness of prevention and control measures to reduce the risk of disease transmission	D1
	Emergency and urgent care flow line	Ensure rapid and effective emergency medical assistance and patient transfer	D2
	Separate clean flow line	Separate clean flow lines to prevent cross-contamination	D3
	Waste flow lines are independent	Separate placement of pollutant flow lines facilitates control of infection sources	D4
	Supply flow line independence	Ensure efficient distribution of materials to meet the efficient operation of logistics support	D5

).

2.2.2. Questionnaire Setting

The questionnaire is divided into two parts. The first part comprises the basic information of the user group, including gender, age, education, income, reason for coming to the hospital, mode of transport to the hospital, travelling time needed to come to the hospital, and emotion of medical treatment, etc. The second part is the Kano questionnaire, which corresponds to the secondary indexes, with a total of 50 questions set up in both the positive and the negative questions (see Appendix C and Appendix D); this takes into account the fact that the questions are accurate in five dimensions of classification precision (matching the five dimensions of needs), logical symmetry (forward and reverse question anchoring), decision support efficiency (Better–Worse quantification), patient friendliness (low cognitive load), and data purity (avoiding neutrality traps), especially in relation to the healthcare transport study, which is faced with differentiated groups such as ageing and emergency; therefore, a 5-point Likert scale was used. Importance analysis was used to determine the importance of transport demand attributes; these attributes were further divided, with the options of the questionnaire scale corresponding to the secondary demand, as shown in

Table 3. Scale design.

Transport Needs	Traffic Type
Form of the question	Forward: Would you be satisfied with an increase in transport travel to and from the hospital (e.g., special buses for medical appointments)?
	Backward: Would you be dissatisfied with a reduction in the number of transport trips to and from the hospital (e.g., special buses for medical appointments)?
Quality Attribute Survey	Satisfied ☐          Deserved ☐          Indifferent ☐          Tolerated ☐          Dissatisfied ☐
Importance survey	Very important ☐          Important ☐          Neutral ☐          Unimportant ☐          Very unimportant ☐

.

The survey used a combination of paper and electronic questionnaires. Before the formal distribution of the questionnaire, its quality was controlled using a pre-test questionnaire (n = 30). The average completion time of a single questionnaire was 8.6 min, the rate of ambiguity of understanding the reverse questions was less than 5%, the Cronbach’s S value was 0.802, and the value of the KMO metric was 0.831. Additionally, Bartlett’s test of sphericity was 0.000, which passes the reliability and validity test. The study was officially released in May–June 2024, and the research date was selected as Monday/Saturday, with time periods of 8:00–11:30 and 14:30–17:00, respectively. The main object of the study was the group of medical practitioners located in the 10 sample hospitals. Inclusion criteria were as follows: age greater than 18–70 years old, no abnormalities in limb function, no language communication barriers, and direct consent and voluntary participation in the study. Exclusion criteria were as follows: completion of the questionnaire in less than 5 min, observation of the wrong respondents, and the use of the field random sampling method. The number of questionnaires collected should be 10 times the number of items on the Kano scale, and the total number of questionnaires should not be less than 250 [67]. The number of questionnaires distributed in each hospital was 40–50. A total of 465 questionnaires were distributed, excluding some missing questionnaires and 49 invalid questionnaires. In total, 416 valid questionnaires were recovered, and the statistical values are shown in

Table 4. Statistical summary of respondents’ information (n = 416).

Items	Classifications	Proportions	Items	Classifications	Proportions
Genders	Man	57.06%	Presence of COVID-19 infection	Exist	83.64%
	Woman	42.94%		None	16.36%
Personnel type	Visit a doctor	70.22%	Transport to medical treatment	Taxi	26.01%
	Hospitalisation	2.25%		Town bus	13.05%
	Visit a sick patient	2.86%		Metro	28.13%
	Accompany	12.92%		Walk	1.56%
	Other	11.75%		Private car	29.69%
Academic qualifications	Specialist and below	50.84%	Length of stay	≤1 h	21.25%
	Undergraduate	36.35%		1–3 h	60.56%
	Master’s degree	12.05%		3–6 h	14.06%
	Doctorate and above	1.77%		≥6 h	4.13%
Incomes	≤CNY 5000	23.44%	Travelling time to the hospital	≤0.5 h	46.88%
	CNY 5000–10,000	53.13%		0.5–1 h	43.75%
	CNY 10,000–20,000	14.06%		1–2 h	6.25%
	≥CNY 20,000	9.38%		≥2 h	3.13%
Age	18–30 years	50.00%	Medical mood	Positive emotions	38.86%
	31–40 years	32.81%		General emotions	11.40%
	41–50 years	9.38%		Negative emotions	49.74%
	>50 years	7.71%	—	—	—

.

The analysis of the statistical information of the group of people who use the medical transport environment shows that 57.06% of men use the environment and 42.94% of women use the environment, which is comparable to the ratio of the two. Among the surveyed participants, individuals holding a bachelor’s degree or higher accounted for 50.17%. Furthermore, those aged 18–40 years constituted 82.81% of the cohort, with 83.64% of this group reporting prior infection with COVID-19. The participant cohort demonstrated a sound understanding of the healthcare transportation environment under both routine and emergency scenarios. The proportion of people who need to travel to the hospital for ≤0.5 h and 0.5–1 h is 46.88% and 43.75%, respectively, and most of the patient groups choose private car, underground, and taxi as the mode of transport for medical treatment, indicating that the underground connection, parking spaces for private cars, and temporary parking for taxis should be considered as key concerns. In the group classification, patients accounted for 70.22%, indicating that the design of the hospital traffic environment should be prioritised from the perspective of the patients’ traffic demand for medical treatment. The stay time of medical groups in ≤1 h and 1–3 h accounted for 81.81%, indicating that the hospital traffic environment mainly serves the short-stay groups. The proportion of negative emotions in patients’ emotions is 49.74%, reflecting that the hospital transport environment should consider the healing nature. Florence Nightingale pointed out that a positive medical environment plays a very important role in accelerating the recovery of patients.

2.3. Statistical Methods

According to the Kano model theory (

Table 5. Kano attribute classification.

User Requirement		Reverse Issue
		Satisfied (5)	Deserved (4)	Indifferent (3)	Tolerated (2)	Dissatisfied (1)
Forward issue	satisfied (5)	Q	A	A	A	O
	deserved (4)	R	I	I	I	M
	indifferent (3)	R	I	I	I	M
	tolerated (2)	R	I	I	I	M
	dissatisfied (1)	R	R	R	R	Q
M	Must-be quality	Patient satisfaction does not increase significantly when such needs are met; however, when such needs are not met, patient satisfaction decreases dramatically.
O	One-dimensional quality	Patient satisfaction increases when such needs are met and, otherwise, decreases.
A	Attractive quality	Patient satisfaction is not reduced when this type of need is not met; however, when it is met, patient satisfaction is greatly enhanced.
I	Indifferentiated quality	Patient satisfaction does not change significantly whether such requirements are met or not.
R	Reverse quality	Patients do not have this need, and patient satisfaction decreases when it is met.
Q	Questionable quality	This result does not usually occur unless the question is unreasonable or the respondent does not understand the question.

Remarks: 1. Double-neutral answer processing rules: If both the positive- and negative-oriented questions are answered as ‘indifferent’ (3), there is no difference in taste attributes in Cabinet (I). 2. Contradictory answers: If both the positive- and negative-oriented questions are answered as ‘satisfied’ (5), the result is considered doubtful and will be excluded.

), the Hospital Transportation Environment Needs Questionnaire consists of 4 dimensions and 25 attribute elements in the form of Kano questions, with positive and negative phrasing for each question. The following 5-point Likert scale was used: satisfied = 5, deserved = 4, indifferent = 3, tolerated = 2, and dissatisfied = 1. The importance survey also used a 5-point Likert scale—very important = 5, important = 4, neutral = 3, unimportant = 2, and very unimportant = 1. The Kano model divides the need quality attributes into six categories, including must-be quality (M), one-dimensional quality (O), attractive quality (A), indifferent quality (I), reverse quality (R), and question quality (Q).

This study combines the Better–Worse coefficient calculation method proposed by Bergar et al. in order to conduct a quantitative classification analysis of user group demand. Compared with methods such as utility weighting or joint analysis, which focus on the relative weights or combined utility of attributes, this method can directly quantify the asymmetric effects of demand, objectively distinguish between demand types, and support resource allocation decisions. For a certain demand (n), statistics on the proportion of $M_n, O_n, A_n, I_n$, and through the proportion of $B_n$ and $W_n$, are derived, respectively, with the following formulas:

$$B_n={\displaystyle \frac{\left(O_n+A_n\right)}{\left(M_n+O_n+A_n+I_n\right)}} \overline{B}={\sum }_{n=1}^m{B}_n$$

$$W_n={\displaystyle \frac{\left(O_n+M_n\right)}{\left(M_n+O_n+A_n+I_n\right)}} \overline{W}={\sum }_{n=1}^m{W}_n$$

where the Better–Worse coefficient indicates the degree of user satisfaction or dissatisfaction when increasing or decreasing a particular type of need. $B_n$ is the rate of increase in satisfaction for the group of users who satisfy this need, and $W_n$ is the rate of decrease in satisfaction for the group of users who do not satisfy this need. Based on the value of the Better–Worse coefficient on the horizontal and vertical axes, the demand types are classified into four quadrants (Figure 3).

3. Results of the Study

The results of the Kano questionnaire relating to the need to use the medical transport environment were tested for validity and reliability using SPSS 26.0. The magnitude of several of these test values for the overall questionnaire is shown in the table below. The overall Cronbach S value is 0.791, of which the Cronbach S value for the positive questions is 0.877, and the Cronbach S value for the negative questions is 0.902; the questionnaire reliability is at a good level. For the validity test, the KMO measure values of the questionnaire overall, positive-oriented questions, and negative-oriented questions are between 0.8 and 0.9, and the validity is at a reasonable level; the significant probability of Bartlett’s sphere test statistic value is 0.000, which is less than 0.01, and has a reasonable correlation (

Table 6. Questionnaire validity and reliability tests.

Element	Item	Cronbach’s Alpha	KMO	Bartlett
Forward issue (n = 416)	25	0.877	0.868	0.000
Reverse issue (n = 416)	25	0.902	0.874	0.000
Overall issue (n = 416)	50	0.791	0.838	0.000

).

3.1. Identification of Need Attributes

The attribute classification of transport demand through Better–Worse (

Table 7. Summary of results of the Kano model analysis (n = 416).

Transport Requirement	A	O	M	I	R	Q	Classification Result	Better	Worse	B-W
A1	41.57%	16.29%	6.18%	32.58%	1.69%	1.69%	A	59.88%	−23.26%	A
A2	8.43%	30.34%	38.76%	19.66%	1.12%	1.69%	M	39.88%	−71.10%	M
A3	18.54%	42.13%	24.16%	15.17%	0.00%	0.00%	O	60.67%	−66.29%	O
A4	29.21%	36.52%	11.24%	19.10%	1.12%	2.81%	O	68.42%	−49.71%	A
A5	10.67%	35.39%	25.84%	23.03%	3.37%	1.69%	O	48.52%	−64.50%	M
A6	31.46%	26.97%	15.73%	25.84%	0.00%	0.00%	A	58.43%	−42.70%	A
B1	9.55%	12.67%	41.01%	28.65%	3.37%	4.74%	M	24.18%	−58.42%	M
B2	24.72%	30.90%	13.48%	27.53%	0.56%	2.81%	O	57.56%	−45.93%	A
B3	34.83%	28.09%	11.80%	22.47%	1.69%	1.12%	A	64.74%	−41.04%	A
B4	12.92%	30.90%	37.08%	17.98%	0.00%	1.12%	M	44.32%	−68.75%	M
C1	15.17%	35.96%	21.91%	25.28%	1.12%	0.56%	O	52.00%	−58.86%	O
C2	19.10%	32.02%	26.97%	21.91%	0.00%	0.00%	O	51.12%	−58.99%	O
C3	19.10%	9.55%	13.48%	56.74%	1.12%	0.00%	I	28.98%	−23.30%	I
C4	5.62%	26.40%	53.37%	14.61%	0.00%	0.00%	M	32.02%	−79.78%	M
C5	14.61%	42.13%	16.29%	25.84%	0.56%	0.56%	O	57.39%	−59.09%	O
C6	37.08%	32.02%	7.87%	22.47%	0.00%	0.56%	A	69.49%	−40.11%	A
C7	8.43%	24.72%	43.26%	23.03%	0.00%	0.56%	M	33.33%	−68.36%	M
C8	17.42%	37.64%	17.98%	26.40%	0.00%	0.56%	O	55.37%	−55.93%	O
C9	16.85%	20.79%	16.85%	42.70%	2.25%	0.56%	I	38.73%	−38.73%	I
C10	5.62%	23.60%	43.82%	24.16%	1.12%	1.69%	M	30.06%	−69.36%	M
D1	11.80%	21.35%	38.76%	23.60%	1.69%	2.81%	M	34.71%	−62.94%	M
D2	11.24%	21.91%	30.34%	30.34%	1.69%	4.49%	I	35.33%	−55.69%	M
D3	17.98%	23.03%	14.61%	43.26%	0.56%	0.56%	I	41.48%	−38.07%	I
D4	39.89%	23.03%	11.80%	25.28%	0.00%	0.00%	A	62.92%	−34.83%	A
D5	17.42%	24.16%	13.48%	44.38%	0.00%	0.56%	I	41.81%	−37.85%	I

Note—A: attractive quality; O: one-dimensional quality; M: must-be quality; I: indifferentiated quality; R: reverse quality; Q: questionable quality.

) and the adoption of mixed-effects logistic regression to quantify the impact of clustering found an intra-group correlation coefficient ICC of 0.12, as well as a 95% confidence interval value of 0.05 to 0.18, which is less than 0.2, indicating that the difference in the transport demand of the sample hospitals in relation to seeking medical treatment is not significant. The problematic demand (Q) accounted for less than 5%, and all demands meet the attribute classification requirements. Among them, there are a total of eight items relating to must-be needs (M), including A2, A5, B4, C4, C7, C10, D1, and D2, which are the attributes that must be possessed by the hospital transport environment, or else they will directly affect the efficiency of the patient groups’ access to medical treatment. There are a total of five items relating to one-dimensional needs (O), comprising A3, C1, C2, C5, and C8, which are the patient groups that think the content should be optimised and improved. Additionally, there are a total of seven items relating to attractive needs (A)—A1, A4, B2, B3, A6, C6, and D4—indicating that the hospital transport environment can be provided, greatly improving the patients’ experience of medical treatment. Indifferent need (I) comprise a total of four items—C3, C9, D3, and D5—for which the patient group has no special need for such functions.

3.2. Importance Analysis

Yang [68] introduced the importance degree into the Kano model, using a Likert scale to investigate the importance degree of each element, defining the attributes at a value higher than the average value of the importance degree as high-importance attributes, while the attributes with values lower than the average value of the importance degree are classified as low-importance attributes. These four attributes of the Kano model are subdivided into eight attributes—must-be attributes (M) contain important attributes (M-1) and required attributes (M-2); one-dimensional attributes (O) contain high-added-value one-dimensional attributes (O-1) and low-added-value one-dimensional attributes (O-2); attractive attributes (A) contain high attractive attributes (A-1) and low attractive attributes (A-2); and indifference attributes (I) contain potential attributes (I-1) and irrelevant attributes (I-2).

Summarising the results of the Kano model, B-W coefficients, and importance analysis (Figure 4), the 25 requirements are distributed as follows: nine belong to the must-be attributes, of which six are important attributes, namely, A5, C4, C7, C10, D1, and D2, while three are required attributes, namely, A2, B1, and B4; five belong to the one-dimensional attributes, of which there are four high-added-value one-dimensional attributes, namely, C1, C2, C5, and C8, as well as one low-added-value one-dimensional attribute—A3; seven are attractive attributes, of which four are high attractive attributes and three are low attractive attributes; four are indifferentiated attributes, of which two are potential attributes (D3 and D5) and two are irrelevant attributes (C3 and C9).

3.3. Sensitivity Analysis

Transportation attributes were categorised based on the Kano model and B-W coefficients; however, it is not possible to determine the transport need elements that require improvement, as well as the priority of these improvements. Combined with Yang’s research on the perception of importance, sensitivity ($R_n$) [69] is introduced to calculate the quantitative indicators of each need. The content of Table 8 is divided into four quadrants (Figure 5). The mean value of the better coefficient ($\overline{B}$) is 0.48, while the mean value of the Worse coefficient ($\overline{W}$) is 0.52. The scores of the values are used to horizontally compare the priority of each need in the same quadrant.

$$R_n=\sqrt{{B_n}^2+{W_n}^2}$$

Included among these, n is a need and $R_n$ is the sensitivity of that need; the larger the value, the more sensitive the patient group is to that attribute.

According to the values of the quadrant coordinates (Figure 6), the highest sensitivity in the first quadrant (attractive need, A-q) is A4 (public circulation), which indicates that the main entrances and plazas of the site should be reasonably organised in the hospital traffic environment in order to reduce the waiting time for patients and visitors to enter the hospital. The highest sensitivity in the second quadrant (one-dimensional need, O-q) is A3 (transport station), which indicates that taking buses is also an effective way to reach the hospital in a more developed area and increasing the number of bus routes passing through the hospital will significantly improve the experience of medical care for the patient population. The third quadrant (indifferentiated need, I-q) has the highest sensitivity at D5 (supply flow line independence), which indicates that the provision of medical supplies in the transport environment will have an indirect impact on the patient group, but the direct need is not significant. The highest sensitivity in the fourth quadrant (must-have need, M-q) was B4 (convenient pick-up and drop-off), suggesting that the lack of smooth and convenient transport routes to medical care in the private transport environment will significantly reduce the satisfaction of the patient group.

Figure 5. Better–Worse four-quadrant plot.

Figure 5. Better–Worse four-quadrant plot.

Figure 6. Sensitivity analysis.

Figure 6. Sensitivity analysis.

According to the results of the quadrant classification (Table 8), and based on the priority ordering of the Kano model attribute classification ‘M-q > O-q > A-q > I-q’, the sorting of each requirement category can be achieved, i.e., the final result is that must-be needs (C4 > A5 > C7 > C10 > D1 > D2 > B4 > A2 > B1) are prioritised over one-dimensional needs (C5 > C8 > C1 > C2 > A3) over attractive needs (A4 > C6 > B3 > D4 > B2 > A6 > A1) over indifferentiated needs (D5 > D3 > C9 > C3).

Table 8. Sensitivity ranking.

Serial Number	B-W	Sensitivity	Arrange in Order	Serial Number	B-W	Sensitivity	Arrange in Order
A3	O	0.8986	1	C10	M	0.7559	14
C4	M	0.8597	2	B2	A	0.7364	15
A4	A	0.8457	3	A6	A	0.7237	16
C5	O	0.8237	4	D4	A	0.7192	17
B4	M	0.8180	5	D1	M	0.7188	18
A2	M	0.8152	6	D2	M	0.6595	19
A5	M	0.8071	7	A1	A	0.6424	20
C6	A	0.8024	8	B1	M	0.6175	21
C8	O	0.7870	9	D5	I	0.5640	22
C1	O	0.7854	10	D3	I	0.5630	23
C2	O	0.7806	11	C9	I	0.5477	24
B3	A	0.7665	12	C3	I	0.3719	25
C7	M	0.7605	13	—	—	—	—

Table 8. Sensitivity ranking.

Serial Number	B-W	Sensitivity	Arrange in Order	Serial Number	B-W	Sensitivity	Arrange in Order
A3	O	0.8986	1	C10	M	0.7559	14
C4	M	0.8597	2	B2	A	0.7364	15
A4	A	0.8457	3	A6	A	0.7237	16
C5	O	0.8237	4	D4	A	0.7192	17
B4	M	0.8180	5	D1	M	0.7188	18
A2	M	0.8152	6	D2	M	0.6595	19
A5	M	0.8071	7	A1	A	0.6424	20
C6	A	0.8024	8	B1	M	0.6175	21
C8	O	0.7870	9	D5	I	0.5640	22
C1	O	0.7854	10	D3	I	0.5630	23
C2	O	0.7806	11	C9	I	0.5477	24
B3	A	0.7665	12	C3	I	0.3719	25
C7	M	0.7605	13	—	—	—	—

In conjunction with the importance needs ranking analysis, the ranking priority of each transport need element is shown in Figure 7. Among all requirements, the essential attribute C4 (patient triage) indicates that mixing patients with different health conditions significantly increases the exposure of susceptible populations to pathogens, while mixed medical traffic delays timely treatment for critically ill patients and exacerbates medical congestion. One-dimensional attribute C5 (distance control) should be focused on, whereby clear guidance to facilities can help patients find their destination and shorten the distance to their medical treatment. Meanwhile, charismatic attribute A4 (public diversion) should be considered because hospitals experience an intensive flow of people. Considering the psychological and behavioural state of the patients, as well as the relationship between the patient’s entrance and exit, car parks and distribution plazas should be reasonably designed. At the same time, this study also found that the user group pays attention to the setting of pedestrian traffic in the hospital traffic environment, indicating that a good pedestrian environment can improve the efficiency of medical care and improve the experience of medical care.

Statistics were calculated for each sub-component $R_n$ of need under the first level of need (

Table 9. Comprehensive ranking of needs.

Level 1 Requirements	Mean	Max	Min	SD	Ranking of Secondary Requirements
Public transport (A)	78.88%	89.86%	64.24%	9.17%	A5 > A2 > A3 > A4 > A6 > A1
Private transport (B)	73.46%	81.80%	61.75%	7.36%	B4 > B1 > B3 > B2
Pedestrian traffic (C)	72.75%	85.97%	37.19%	14.23%	C4 > C7 > C10 > C5 > C8 > C1 > C2 > C6 > C9 > C3
Medical transport (D)	64.49%	71.92%	56.30%	6.99%	D1 > D2 > D4 > D5 > D3

First-order need averages are ranked: A > B > C > D.

). According to the order of the mean sensitivity value, the first level of need can be obtained as follows: public transport (78.88%) > private transport (73.46%) > pedestrian transport (72.75%) > medical transport (64.49%); this reflects the higher importance of public transport in the demand for transport to healthcare compared to other levels of demand. Urban connectivity (A5) indicates that priority should be given to the hospital–urban interface in relation to the patient’s medical transport needs, avoiding thinking about the patient’s transport organisation only from within the base, understanding the patient’s medical transport needs in a fragmented way, and ignoring the continuous spatial perception of a person in medical treatment. This will lead to a decrease in the efficiency of wayfinding. Private transport (73.46%) and walking transport (72.75%) have similar mean values, and the standard deviation of the sensitivity of the second-level subcomponents of private transport is the smallest, and has the least fluctuation, which indicates that the user groups have similar patient experience needs for private transport in the transport environment of medical treatment. The convenience of pick-up and drop-off (B4) has a higher sensitivity ranking, which indicates that user groups are more concerned about driving routes and parking needs in a private transport environment. The standard deviation of the sensitivity of the secondary subitems in the pedestrian transport environment is the largest and it fluctuates the most, which may be due to the different medical conditions of the patients and the differences in cognition, leading to the existence of large differences in the transport needs of the patients. The pedestrian traffic environment is more concerned with patient triage (C4), indicating that hospitals should strictly control the traffic flow of patients with infectious diseases to prevent cross-infection. Medical traffic is significantly lower than the first three items, indicating that the user group pays less attention to flow lines that are not directly related to them, with the highest sensitivity for infectious disease area independence (D1). This conclusion further complements the conclusions of C4.

3.4. Group Variability Analysis

This study divided the statistically obtained data into three groups—medical mood, mode of transport to the hospital, and group type—and used SPSS26.0 to test the validity and reliability of the questionnaire results of the demand in the three groups (Figure 8). The significant probability of Bartlett’s sphere test statistic values was 0.000, which was less than 0.01, and the results were at a more reasonable level.

The top three rankings of the sensitivity $R_n$ of each of the secondary indicators within the group were analysed for their demand preferences (

Table 10. Group differential needs ranking.

Groups	Categories	Public Transport	Private Transport	Pedestrian Traffic	Medical Transport
Medical mood	positive emotion	A5 > A2 > A3	B3 > B4 > B2	C4 > C5 > C2	D2 > D4 > D1
	general emotion	A5 > A2 > A4	B3 > B4 > B2	C1 > C4 > C2	D1 > D2 > D4
	negative emotion	A2 > A4 > A3	B4 > B2 > B3	C5 > C7 > C4	D1 > D4 > D2
Transport mode	private car	A5 > A2 > A3	B3 > B1 > B2	C4 > C5 > C2	D2 > D4 > D1
	town bus	A5 > A2 > A4	B1 > B2 > B3	C4 > C5 > C2	D1 > D2 > D4
	taxi	A2 > A4 > A3	B1 > B3 > B2	C5 > C7 > C4	D1 > D4 > D2
	metro	A5 > A2 > A3	B1 > B3 > B2	C4 > C5 > C2	D2 > D4 > D1
	walk	A5 > A2 > A3	B1 > B3 > B2	C4 > C5 > C2	D2 > D4 > D1
Group type	seek medical advice	A2 > A5 > A4	B4 > B3 > B2	C5 > C4 > C7	D4 > D2 > D1
	chaperone	A2 > A5 > A4	B3 > B1 > B2	C10 > C5 > C1	D1 > D2 > D4
	visit a sick patient	A2 > A4 > A1	B1 > B2 > B4	C2 > C5 > C8	D1 > D2 > D3
	hospitalisation	A5 > A1 > A2	B4 > B5 > B1	C1 > C9 > C2	D3 > D4 > D5

). The sensitivity rankings of public transport demand, private transport demand, and walking transport demand were generally consistent across the groups, with a more significant demand for urban connectivity (A5), convenient transfers (B4), direct underground access (B3), patient diversion (C4), and emergency and urgent care flow lines (D2).

The medical visit mood group has a positive mood, while the general mood groups have a higher degree of need for city connectivity than negative mood patients; the negative mood patients favour the degree of need for transport frequency due to the fact that the high frequency of departures reduces the waiting time of patients, avoids misdiagnosis due to traffic delays, and relieves anxiety. In private transport, the negative emotion patient group is more concerned about the convenience of pick-up and drop-off and the importance of companionship than the other emotion patients, as the former is more concerned about direct underground access. In pedestrian transport, the patient group is more concerned about patient triage; in particular, after the COVID-19 pandemic, people’s awareness of infectious diseases has increased. For medical transport, the positive mood group is more concerned about emergency transport than the other two, and the latter is more concerned about the independence of infectious disease areas. This indicates that there is a greater demand from the user group for smooth and convenient transport flow routes for medical appointments involving acute illnesses.

The level of need for hospital–city connectivity was more pronounced in the groups of modes of transport to the hospital. This suggests that hospitals should have a friendly and continuous spatial interface with the city to reduce the waiting time and walking distance of patients and visitors to the hospital entrance and to reduce traffic congestion around the hospital. In relation to private traffic, private cars are more concerned about underground direct access than other modes of transport, indicating that private cars have higher requirements for the traffic environment of the hospital than other modes of transport; additionally, the hospital should set up an underground traffic connection area, which can be connected to the underground commercial street and should provide access to the diagnostic and treatment units through escalators and lifts. The pedestrian traffic group has a higher degree of need for patient triage, distance control, and pedestrian–vehicle separation, especially in the event of a health emergency, where patient separation and distance control are important methods of preventing and controlling the spread of viruses. In specialised medical transport, emergency and urgent care circulation is a common concern for the patient group, probably because of the need for rapid response and timely treatment in emergency medical situations.

Among group types and categories, patient groups in the public transport group were generally concerned about the frequency of transport, public evacuation, and the degree of need for the frequency of transport, indicating that the patient groups relied more on using public transport for medical treatment. In relation to private transport, the degree of need for convenient transfer is highest in the consultation and hospitalisation group, probably because this group is directly associated with the hospital and due to the idea that convenient transfer will reduce the discomfort in the process of medical treatment. Combined with the interviews, it was found that a small number of patients came from remote rural areas to seek medical treatment, and the travelling time required to come to the hospital was more than 2 h. The psychological pressure to seek medical treatment was high; therefore, convenient transport connections could improve spatial accessibility. On the other hand, it also indicated the existence of the phenomenon of the unbalanced distribution of medical resources. In pedestrian traffic, the patient group has a higher degree of need for distance control, especially the consultation group. In relation to medical traffic, the accompanying and visiting groups have a higher degree of need for the independence of the infectious disease area, because these groups have a higher need for public activity space to meet the needs of resting, walking, and socialising.

4. Discussion

The differences between this study and existing studies of patients’ transport needs for medical treatment are mainly reflected in the following three aspects. (1) Previous studies have explored the important elements of patient transport needs for medical treatment more from a single factor perspective of the transport environment for medical treatment, and have not explored patient variability from the integrated transport system perspective of the city–hospital outdoors–hospital indoors [70,71]. (2) Relative to traditional mathematical and statistical methods [72,73], the Kano model can more comprehensively determine the priority of patient transport needs for medical treatment and optimise the allocation of resources to patient transport for medical treatment by city managers and hospital administrators. (3) The Kano model can regularly reassess the phased priority evolution of patients’ medical transportation needs and make timely adjustments. Unlike previous computational models, which primarily focused on the analysis of static data [74], this research method can swiftly capture differences in patient healthcare transportation needs arising from changes in urban and hospital transportation. When optimising the actual indicators, both material foundation attributes and patient experience attributes must be optimised simultaneously. This is because the former serves as the material foundation for the latter (e.g., parking facilities influence waiting times), while the latter reveals deficiencies in facilities (e.g., ‘route efficiency’ indicates insufficient signage). Given the differing developmental foundations, priorities, and directions across regions, management requirements and control measures for hospitals must also be tailored to local conditions, as well as being adjusted and refined in accordance with the construction goals and requirements of the specific location. Additionally, the research methodology framework established in this paper holds significant implications for studying the transportation needs of users of other building types within cities.

This study has the following limitations, which need to be further improved in future research. First, in terms of methodology, although the Kano model can effectively identify patient need priorities, an increase in the number of positive and negative items in the questionnaire significantly reduces data quality. Therefore, it is necessary to optimise data reliability by streamlining the number of questions. Additionally, analyses based on 10 individual attributes indicate that the impact of individual differences on healthcare transportation needs exhibits variability. Future research may need to incorporate more diverse personal factors, such as health status, disease severity, and socioeconomic status, to enhance explanatory power. Second, in terms of sample representativeness, the current study focuses solely on the ultra-high-density urban environment of 10 hospitals in Guangzhou. Future research could include cases from differentiated regions such as low-density cities and pilot zones for tiered healthcare systems. For example, the former should prioritise enhancing parking facilities and private vehicle shuttle services, while the latter should improve public transport adaptability through cross-regional medical shuttle lines and appointment-based shuttle buses. Thirdly, in terms of research dimensions, the existing two-dimensional analysis framework of demand and importance can reflect basic cognition, but it fails to combine spatial elements such as urban land use mix and multi-modal transportation accessibility to construct a predictive model. Future research could integrate GIS spatiotemporal data analysis technology to establish a dynamic demand prediction system (introducing entropy methods to adjust indicators). Additionally, there is an insufficient exploration of the differentiated needs of special groups (such as patients transferred from other regions). It is recommended to supplement surveys of specific populations through stratified sampling and explore potential influencing factors such as appointment booking time slots, the number of accompanying persons, and emergency medical scenarios, thereby forming a more inclusive transportation service strategy. Future research could expand the regional scope, deepen the analysis of variable correlations, and introduce implementation cost calculation models to extend the conclusions to diverse urban types and complex medical scenarios, thereby providing theoretical support for refined urban medical transportation planning.

5. Conclusions

Based on the Kano model, this study systematically developed a questionnaire relating to patient transportation environment needs, which includes 4 primary indicators and 25 secondary indicators. A field survey was conducted in 10 hospitals in Guangzhou, and the Better–Worse coefficient and importance analysis were combined to define the attributes of each factor and their degree of influence. Through sensitivity ranking calculations, key priority factors were identified. The questionnaire’s reliability and validity were verified, and the study clarified the key areas for optimising transportation environment factors.

(1)

Prioritise the efficiency of public transport connections

Optimising public transport should be the core task of hospital transport environment renovations. Specifically, this goal should be achieved by strengthening seamless connections between hospitals and the city’s main road network and rail transit stations (e.g., by adding shuttle lines and reserving space for metro connections), while reducing investment in low-demand facilities (e.g., direct aerial passageways). Additionally, the efficiency of private vehicle traffic flows should be improved in tandem with ensuring safe pedestrian diversion designs, while strictly controlling the proportion of resources allocated to special medical transportation. It is recommended that the above content be incorporated into the Chinese Comprehensive Hospital Architectural Design Standards, explicitly stipulating that public transportation hubs must be established within a 200 metre radius of comprehensive hospitals, and that this requirement could be used as a bonus criterion in the green building evaluation system.

(2)

Establish a tiered response-based transportation service system

a.

Private transportation: Optimise drop-off and parking management to improve traffic efficiency. Set up clearly marked, time-limited ‘quick entry and exit’ drop-off zones near the hospital entrance and strictly separate them from long-term parking areas (such as underground/remote car parks). Mandatory implementation of intelligent parking guidance systems (real-time display of available parking spaces and optimal routes) should be integrated into the city’s ‘new infrastructure’ smart transportation system to effectively alleviate congestion at entrances (Figure 9).

b.

In terms of pedestrian traffic, ensure safe and comfortable walking and optimise the medical treatment process. Designated treatment hospitals (in accordance with the Chinese Construction Standards for Infectious Disease Hospitals and emergency planning) must strictly implement ‘patient segregation’ and incorporate infection control standards into hospital protocols. General hospitals should optimise pedestrian networks, add continuous sunshades and rain shelters, and install seating areas (integrated with courtyard landscaping). Implement a ‘security check exemption’ system (for non-critical areas and routine outpatient services) to streamline processes, enhance efficiency, and improve the patient experience.

c.

Special medical transportation focuses on core emergency support and strictly controls resource allocation. Only the ‘emergency green channel’ is retained and strengthened to ensure ambulances reach the emergency area with ‘zero delay.’ This channel must have physical isolation, dedicated entrances, and intelligent interconnection (traffic signal priority) features, and be clearly stipulated in the Comprehensive Hospital Construction Standards and local emergency regulations. Strictly control the proportion of other special transportation resources to ensure the fairness and efficiency of public medical resources, as well as prioritising core emergency needs.

(3)

Differentiated responses to the core demands of different groups

In terms of emotional state, priority should be given to improving public transport frequency, optimising convenient pick-up and drop-off services, and strengthening independent signage in infectious disease areas to alleviate psychological stress for patients with high anxiety and negative emotions. For patients with positive/neutral emotions, the focus should be on improving seamless connections between hospitals and urban transport networks, ensuring the smooth flow of dedicated emergency and first aid channels, and optimising direct access for private cars to core treatment areas. In response to transportation behaviour patterns, public transportation must significantly increase route density and peak-hour departure frequencies while optimising station crowd management. Private vehicles require the construction of efficient multi-level transfer zones that are directly connected to treatment units (with adequate vertical transportation), the implementation of appointment-based parking and smart guidance systems, and the provision of sufficient drop-off zones for quick exits. Pedestrian systems must strictly enforce patient-to-patient and pedestrian-to-vehicle separation, optimise pathways and environments, and strengthen distance control during emergencies. Special medical transportation must ensure that emergency and urgent care channels remain unobstructed 24/7 and that infectious disease zones have completely independent circulation routes. For different social groups, high-frequency service coverage must be ensured for public transport dependents. Convenient drop-off zones should be set up at the hospital entrance for outpatient/inpatient groups, while intra-hospital shuttle services should be improved. Scheduled shuttle buses should be established for patients in remote areas (supplemented by long-term strategies for the balanced allocation of regional medical resources); path distance control and safety for walking outpatient groups should be optimised; and the needs of accompanying visitors for independent transportation and resting spaces in infectious disease zones should be met.

(4)

Establish a dynamic demand assessment mechanism

Regularly monitor changes in demand for medical care and transportation modes (with a particular focus on the combined effects of negative emotions and private car use), and identify ‘urban–hospital connectivity,’ ‘internal transfer efficiency,’ and the ‘isolation of infectious disease flow lines’ as indicators for continuous optimisation.

This study provides new research insights for the development of healthcare transportation environments in cities with high urbanisation rates and population densities, validates the methodological feasibility, and systematically assesses the hierarchical structure of transportation demand guided by healthcare experience. The research findings, supported by demand priority data, propose strategies for transportation environment planning and design.