Research on Vertical Professional Collaborative Evaluation Tools of Healthcare System Based on the Tight County Healthcare Alliance in China

Development of the questionnaire

Description of the evaluation framework dimensions

Based on the “Evaluation framework of human-centered integrated healthcare system”, the theoretical presupposition of evaluation dimensions and elements of vertical professional integration in healthcare system is designed to include 4 dimensions, each corresponds to several elements (Figure 1). Firstly, professional division of labor and cooperation refers to the division of labor and cooperation mechanism at different levels and between different professions (prevention, rehabilitation, nursing, medical treatment, etc.), such as establishment of interdisciplinary service teams, signing of cooperation agreements on the basis of projects, and the communication mechanism among professionals. It consists of three elements, namely cross-professional coordination mechanism, cross-professional cooperation mode, and the closeness of cross-professional cooperation. Secondly, the incentive and constraint mechanism refers to that between different levels and different specialties, such as the benefit distribution mechanism and the responsibility sharing mechanism among specialties. It also consists of two elements, the incentive mechanism of cross-professional cooperation and the constraint mechanism of cross-professional cooperation. Thirdly, inter-professional service norms and standards refer to the formulation of targeted, guiding and standardized inter-professional service norms and standards to guide different levels and professions to cooperate with each other based on their respective capabilities, roles, responsibilities and obligations; the key element here is inter-professional collaboration of specifications and standards. Fourthly, compatibility of cross-professional values and cultures refers to the consistency of health-related concepts, values and cultural cognition at different levels and in different professions; the two elements fall into this category are the value concept of “human-centered” shared among professionals and the degree of trust between professionals [5].

Figure 1

Data Collection

The paper-based questionnaire collected on-site was conducted among 8 township health centers December 2020, covering all 3 county-level public hospitals in D county, H Province(anonymous processing). A total of 450 doctors, nurses, laboratory technicians and other health professionals in medical and health institutions were interviewed using stratified cluster equal proportion random sampling. Altogether 450 questionnaires were distributed and 420 were effectively recovered, making the effective recovery rate 93.33% after careful examination of each copy on the spot to filter out those with wrong filling or missing items. we maintained research ethics standards by: (1) providing participants with project information sheets, (2) obtaining verbal consent, and (3) ensuring complete data anonymization.

Analysis Method

Exploratory factor analysis: when the eigenvalue > 1 is the valid common factor, the factor load >0.35 is the meaningful item. When the correlation coefficient between the score of the item and the total score of the dimension of the item is greater than 0.2 and the correlation coefficient between the score of the item and the total score of other dimensions is less than 0.2, it is a meaningful item. Items that do not meet the above requirements can be considered for deletion as meaningless.

Reliability analysis: If the Cronbach’s α of the whole and each dimension is greater than 0.7, the reliability is good. If composite Reliability (CR) is greater than 0.7, and Average Variance extraction (AVE) is above 0.6, they can support further structural equation model analysis.

Structural equation model: P value of standardized regression coefficient being less than 0.05 indicates a statistical significance between X and Y; The model fitting index must meet the following conditions, namely Root Mean Square Error of Approximation (RMSEA) < 0.10, Comparative Fit Index (CFI) > 0.9, Normed Fit Index (NFI) > 0.9, Non-Normed Fit Index (NNFI) > 0.9. The above analysis is based on SPSS19.0.

General information of respondents

A total of 420 medical personnels were interviewed, including 215 males, accounting for 51.2% of the surveyed population; 234 of them were with a degree of bachelor or higher, accounting for 55.7% of the total; 159 of the interviewed were with intermediate titles, accounting for 37.9%, and 51 of them were with deputy titles and above, accounting for 10.5% of the total.

Exploratory factor analysis

After the Bartlett spheroid test, KMO = 0.898 > 0.5, indicating that the data were suitable for exploratory factor analysis. Five common factors with eigenvalue greater than 1 and cumulative contribution rate being 72.23% were extracted with principal component and maximum variance method. Items 3, 2, 6, 7, 9, 10, 12, 13, 14, 15, 16, 17, 18, 20, 21, 22, 23 were retained because their factor loads were all greater than 0.5, and those of non-shared factors were all less than 0.5, meanwhile, the correlation coefficient of the total score of the shared factors was greater than 0.2 and that of the total score of the non-owning common factor is less than 0.2. Among them, items 18, 20, 21, 22 and 23 belong to common factor F1, items 6, 7, 9 and 10 belong to common factor F2, items 1, 13, 14 and 15 belong to common factor F3, items 2 and 3 belong to common factor F4, and items 16 and 17 belong to common factor F5 (Table 2). Item 1 and item 4 were abandoned due to their disqualified factor load of the common factor and that of the non-common factor (both were greater than 0.5) items 2, 5, 8, 11, 16 and 19 were abandoned because their correlation coefficient of the total score of the common factors was less than 0.2, and their total score coefficient of the common factors was greater than 0.2, which made them meaningless items (Table 2).

Reliability analysis

The model had good convergence, which supported further structural equation model analysis. The overall reliability was good, with Cronbach’s α = 0.846. The reliability of all dimensions was generally good, and the Cronbach’s coefficients of common factors F1, F2, F3, F4 and F5 were 0.929, 0.893, 0.822, 0.643, 0.81, respectively, and the Cronbach’s α of the other common factors were greater than 0.7 except F4. The component reliability (CR) of common factors F1, F2, F3, F4 and F5 were all greater than 0.7, being 0.920, 0.880, 0.880, 0.818 and 0.787 and the average coefficient of variation extraction (AVE) were all greater than 0.6, being 0.697, 0.647, 0.650, 0.693 and 0.648.

Structural equation model

In the initial model (M0), of F1, F2, F3 and F4 were retained for their P values of the standard regression coefficients of the degree of professional integration were all less than 0.05. Items 18, 20, 21, 22 and 23 were retained for their corresponding P values of F1 standard regression coefficients were all less than 0.05. Items 6, 7, 9 and 10 were retained for their corresponding P values of the standard regression coefficients of F2 were all less than 0.05. Items 12, 13, 14 and 15 were retained for their corresponding P values of the standard regression coefficients were all less than 0.05. And items 2 and 3 were retained for their corresponding P values of the standard regression coefficients of F4 were all less than 0.05. of Meanwhile, F5 (corresponding items 16 and 17) was abandoned because its P value of the standard regression coefficient was greater than 0.05, indicating no impact on the degree of professional integration.

In the revised model (M1), F1, F2, F3 and F4 were retained because their P values of the standard regression coefficients for the degree of professional integration were all less than 0.05, and those of the corresponding items and factors were all less than 0.05. The model fitting indexes of the revised model (M1) were all better than those of the initial model (M0) (Tables 3 and 4).

The dimension of “value compatibility and trust” was essentially to measure the degree of the consensus of medical and health service providers at different levels on collaborative relationships

The common factor F1 was renamed as “value compatibility and trust”, referring to the consistency of health-related values, culture and behavior norms of between different levels and different professional medical and health service providers, including the two elements of value compatibility and trust, which in essence reflected the degree of consensus reached by different levels and professional medical and health service providers on the subjective meaning of vertical professional integration as a social action, which was conducive to the formation of the legitimate order of collaboration [7].

First, value compatibility was an important aspect of value integration. This element was originally named “subjective norm compatibility”, whose essence was the subjective cognition and consensus of different action subjects to the rules, and was a sociological order of tradition and habit. Value compatibility can form moral constraints, so that individuals perceive social environment pressure when deciding whether to perform a specific behavior, reflect the influence of important others or groups on individual behavior and are influenced by normative beliefs and obedience motives, and moral constraints formed by value compatibility are considered to be important factors affecting doctors’ autonomous professional behavior [89]. Item 18 measured perception of rationality of existing norms and standards of interprofessional collaboration by different professional medical and health service providers, which originally belonged to the dimension of “norms and standards of interprofessional collaboration”, but was eventually included in the dimension of “value compatibility”, indicating that the perception of normative rationality was part of the degree of value integration. This, similar to the recognition of the “human-centered” value concept measured with items 22 and 23, fell into the category of “value compatibility”, while showing consistency with the results of relevant studies on doctors’ referral behavior, indicating doctors’ referral behavior being altruistic [10].

Next, the degree of interprofessional trust. Trust is the basis of cooperative behavior. In a broad sense, trust refers to a belief or state of mind in which the donor believes that the fiduciary will do its best to fulfill its commitments and will not exploit the weakness of the donor to seek improper benefits. In the previous theoretical presupposition, the trust relationship between medical and health service providers of different levels and specialties was also regarded as an important element in the concept and value dimension that affected vertical professional integration [11]. From the empirical results of this study, items 20 and 21 measured the degree of trust between different levels and different professions, and further indicated that there was a direct correlation between the degree of trust and vertical professional cooperation behavior, and the possibility of vertical professional integration behavior can be speculated by measuring the degree of trust relationship between different levels and different professions, which was consistent with the previous theoretical presupposition.

“Communication and coordination mechanism” was an important prerequisite for the cooperation between different parities whose essence is the flow of information elements and its guarantee mechanism

Common factor F2 was renamed as “communication and coordination mechanism” to reflect the flow of information elements and its guarantee mechanism for collaboration between medical and health service providers at different levels and indifferent areas of expertise, including two elements of health information support system and organizational guarantee.

Health information support system was supposed to be interconnected Item 6 “Do you refer to diagnoses and treatment plans of other medical institutions in the medical community?”, item 9 “Have you ever conducted collaboration with doctors from other medical institutions in the medical community on the medical information platform?”, and item 10 “Have you ever obtained patient’s previous diagnosis and treatment information from the medical information platform?” evaluated and measured the interconnection of information through the use of health information system by medical staff, which further explained that the interconnection of health information system was an important factor to promote vertical professional integration. This was similar to some previous study results, which adopted the theory of planned behavior and proposed that information interconnection had a positive influence on perceptual behavior control, and indirectly promoted two-way referral intention through perceptual behavioral control [1213].

The second element was about organizational guarantee mechanism. Item 7 “Is specific coordination available for you to have inter-professional collaboration with other doctors?” manifested requirement for special personnel to be responsible for inter-professional coordination. This is consistent with the classical theoretical model of professional collaboration, Chronic Care Model (CCM), which clearly stated that dedicated coordination was an important organizational guarantee element, meaning “designating a medical service worker as a case manager to improve the coordination and continuity of the service system” [1415]. At the same time, in the specific practice at home and abroad, the establishment of a special person responsible for the coordination of different medical institutions to ensure the realization of two-way referral is also widely emphasized. For example, in the management of chronic diseases in Germany, a case manager is designated to help build collaboration between different professional medical and health service providers [16]. In practice in China, the Shanghai hierarchical diagnosis and treatment plan clearly requires that “each medical institution within the medical consortium set up a coordination department for referral, and assign special personnel be responsible for the work contact between the institution and the upper and lower medical institutions of the medical consortium and the coordination between departments within the institution [417].

“Incentive and constraint mechanism” included material and non-material factors that affect vertical professional integration behavior

Common factor F3, named “incentive and constraint mechanism” worked as a mechanism to guide and motivate professional cooperation behavior with incentives that influenced medical staff, involving performance appraisal and income, professional title promotion, doctor-patient trust and other aspects, both material and non-material, which was a core dimension to promote vertical interprofessional collaboration [118].

Firstly, economic income and career development are common components of incentive and constraint mechanisms, focusing on material and economic incentives. This is consistent with the presupposition of the previous theoretical framework. that is, material and economic factors such as economic income and career promotion space play an important role in stimulating and guiding the vertical professional integration behavior of different levels and different professional health service providers [19]. Most of the studies have focused on the influence of economic incentives on the behavior of doctors, emphasizing the use of rational and purposeful policy tools such as medical insurance payment, salary and administrative power to play the role of formal institutions in shaping doctors’ group behavior [20].

Secondly, doctor-patient trust is a new component of incentive and restraint mechanism which belongs to interpersonal capital and is a non-material incentive factor. Item 15 “Does engagement in inter-professional collaboration have an positive impact on trust between you and your patients?” measured patient-trust relationship, which previously belonged to the “sharing of people-centered values among professionals” and later incorporated into the “incentive and constraint mechanism”, supplementing the role of previously overlooked non-material incentives. Some existing research results also showed that the doctor-patient trust relationship, as an interpersonal capital in the doctor-patient social relationship, had a positive incentive and promotion effect on improving the doctor’s referral behavior [21]. At the same time, the doctor-patient trust relationship manifested interpersonal capital status between doctors and patients, rather than that between medical and health service providers of different levels and professions. Therefore, compared with the dimension of “shared value concept among professions”, the dimension of doctor-patient trust was more in line with the dimension of “incentive and constraint mechanism” logically, whose content validity was better [22].

The “structure and strength of cooperation relationship” reflected the strength of social network structure and connection between health technicians at different levels and in different professions

The common factor F4 was renamed as “structure and strength of cooperation relationship”, referring to the structure and strength of the social networks among medical and health service providers at different levels and with different specialties, including two elements: cooperative relationship structure and cooperative relationship strength.

The essence of the division of labor and cooperation mechanism is the position, function and cooperation mode of different levels and types of medical and health service providers in the collaborative network, which reflects the structure and intensity of the collaborative relationship. Specifically, the first is the structure of cooperative relationship. Item 2 “What business collaborations have you participated in in the medical community?” originally belonged to “division of labor and cooperation mode” in the dimension of “division of labor and cooperation mechanism” and indicated the positioning and cooperation mode of different levels and professional medical and health service providers in the collaboration network, which was in line with the definition of social network structure, because of what it was named the collaborative relationship structure. This coincided with results of some studies, among which formal collaborative relationships such as referral – training, referral – support, chronic disease – telemedicine, information – organization management, friendship – referral, friendship – telemedicine, friendship – chronic disease, friendship – support were pointed out to have positively affected each other and have been helped with informal friendship networks in formation. Social network structure promoted vertical professional integration [23]. The second is the strength of the collaborative relationship. What came next was the strength of collaborative relationship. Item 3 “What information do you have to contact doctors in other medical institutions within the community?” measured the frequency and closeness of the contact between health technicians, mainly doctors, reflecting the strength of collaborative relationships. This was consistent with the results of previous studies, which pointed out that the stronger the social network connection, the higher the closeness and frequency of connection, which could promote the trust relationship between collaborators, and then positively affected the ability of knowledge integration, including systematization, socialization and collaboration [24].