Original Article

Analysis of downward referral willingness and its influencing factors among recovery phase hospitalized patients in tertiary hospitals


Yang Chen1, Xiaowen Wang1, Xiangjin Kong2*, Lu Zhao, Yumeng Gao2, Xinyuan Xu2

1Dalian Health Development Center, Dalian 116011, Liaoning Province, China

2College of Humanities and Social Sciences, Dalian Medical University, Dalian 116044, Liaoning Province, China


Objectives: This study aimed to understand the willingness of hospitalized patients in stable conditions and in the recovery phase in tertiary hospitals toward downward referral (DR) and to explore its influencing factors. Methods: A questionnaire survey was conducted among hospitalized patients from five tertiary general hospitals. Data were analyzed and processed using SPSS17.0 software. Results: The rate of knowledge regarding the hierarchical medical system and DR willingness were low. Per capital household annual income, the form of medical insurance, and the medical treatment decision-making process are the main factors influencing patients’ willingness. Conclusion: To improve patients’ DR willingness, it is necessary to further strengthen public understanding of the hierarchical medical system and to build the service capabilities of primary medical institutions.


Keywords: Tertiary hospital; Health care-seeking willingness; Two-way referral; Patients


Address for correspondence:

Xiangjin Kong, E-mail: kongxj110@126.com

For reprints contact: reprints@sppub.org

Received 09 October 2021; Accepted 05 January 2022; Available online 25 July 2022



Two-way referral is key to the establishment of a hierarchical medical system.[1] Rational willingness and behavior of patients are prerequisites for the successful implementation of two-way referral. At present, patients’ general inclination toward higherlevel hospitals leads to the long-standing problem of a high rate of upward referrals and a low rate of downward referrals (DR).[2] In this study, a questionnaire survey was conducted among patients hospitalized in tertiary hospitals to analyze their DR willingness and its influencing factors, and evidence was provided for the improvement of the hierarchical medical system and relevant policies.


Participants and Methods

As sample institutions, this study included five tertiary general hospitals located in four cities in Liaoning province. Patients receiving inpatient treatment in the Gastroenterology, Neurology, Cardiology, Respiratory, and Endocrinology departments and who were in the stable recovery phase participated in the study. Based on the existing domestic and foreign literature, the “Questionnaire on patients’ healthcare-seeking willingness and behavior (for downward referral)” was designed, and then revised and improved after expert consultation. The reliability and validity of the questionnaire were tested through a face-to-face survey with the participants. The questionnaire consisted of two parts relating to the socio-demographic characteristics of the participants and DR willingness. A total of 500 questionnaires were distributed, of which 498 were returned, with an effective recovery rate of 99.6%.


Statistical analysis

The database was established using EpiData 3.1 software, and data analysis was conducted using SPSS 17.0. The socio-demographic characteristics of patients, their awareness of the relevant policies regarding the hierarchical medical system, and their DR willingness were analyzed with descriptive statistics. Chi-square test was used for influencing factors of DR willingness of patients in the recovery phase, and a P-value of less than 0.05 (P < 0.05) was considered statistically significant. Influencing factors with statistical significance were analyzed with logistic regression analysis.


Socio-demographic characteristics of participants

Among the 498 participants, 312 (62.7%) were female, 349 (70.1%) were aged over 35, 352 (70.7%) had a high school or technical school education or higher, 329 (66.1%) were urban residents, 203 (40.8%) had a per capita household annual income of less than CNY 20,000, 300 (60.2%) were covered by the Urban Employee Basic Medical Insurance and the Urban Residents Basic Medical Insurance, and 142 (28.5%) were covered by the New Rural Co-operative Medical System (Table 1).


Table 1: Socio-demographic characteristics of participants.
Socio-demographic characteristics Number Proportion (%) Socio-demographic characteristics Number Proportion (%)
Sex Male 186 37.3 Occupation Government or public institution 113 22.7
Female 312 62.7 Salaried employee 107 21.5
Age <18 17 3.4 Farmer 68 13.7
18–35 132 26.5 Student 94 18.9
35–45 94 18.9 Self-employed 63 12.7
45–60 146 29.3 Other 53 10.6
≥60 109 21.9 Per capita household annual income (CNY) <20000 203 40.8
Place of registration Urban area 329 66.1 20000–29999 118 23.7
Rural area 169 33.9 30000–39999 117 23.5
Education <Junior high school 146 29.3 ≥40000 60 12
High school/technical school 96 19.3 Insurance Urban Employee Basic Medical Insurance 150 30.1
College 175 35.1 Urban Residents Basic Medical Insurance 150 30.1
University or higher 81 16.3 New Rural Co-operative Medical System 142 28.5
Other 56 11.2


Patient awareness of the hierarchical medical system

The results indicated that 49.8% of patients were “aware that the Chinese government is implementing the hierarchical medical system,” whereas the other 50.2% indicated their “lack of awareness.” A total of 50.6% of the patients stated their “awareness of the differential medical insurance reimbursement policy (the health claim settlement ratio for hospitalization in primary medical institutions is higher compared to higher-level hospitals),” whereas the other 49.4% indicated their “lack of awareness.”


Patients’ downward referral willingness

Regarding the question, “if your illness is stable and you are in the recovery phase, would you be willing to be referred downward to a primary medical institution to continue your treatment?”. 67.1% of patients expressed their “unwillingness” and only 32.9% expressed their “willingness.” Regarding “the main reasons for not accepting DR to primary medical institutions after the illness condition became stable,” among 334 patients who indicated their unwillingness, 53.9% chose “distrust in medical technologies of primary medical institutions,” 24% chose “fear of relapse,” 11.1% chose “poor healthcare environment and equipment condition,” and 11% chose “other reasons.”


Analysis of factors influencing patients’ downward referral willingness

Single-factor analysis of personal socio-demographic characteristics influencing patients’ downward referral willingness

The results showed that the proportion of male patients (73.1%) who were unwilling to accept DR was higher than the proportion of female patients (63.5%); the 35–45 age group had a higher proportion of patients unwilling to accept DR (71.3%) compared to other age groups; the proportion of patients from urban areas (72.9%) who were unwilling to accept DR was higher than those from rural areas (55.6%); the proportion of patients with a university degree or higher (77.8%) who were unwilling to accept DR was higher than other education groups; and the proportion of patients working in government/public institutions (80.5%) who were unwilling to accept DR was higher than those with other occupations. There were statistically significant differences among the participants regarding sex, place of registration, education, occupation, and per capita household annual income (Table 2).


Table 2: Results of single-factor analysis of personal socio-demographic characteristics influencing patients’ downward referral willingness.
Socio-demographic characteristics Willing Unwilling χ2-value P-value
Number Proportion (%) Number Proportion (%)
Sex Male 50 26.9 136 73.1 4.92 0.030
Female 114 36.5 198 63.5
Age <18 5 29.4 12 70.6 2.381 0.666
18–35 50 37.9 82 62.1
35–45 27 28.7 67 71.3
45–60 47 32.2 99 67.8
≥60 35 32.1 74 67.9
Place of registration Urban area 89 27.1 240 72.9 15.176 <0.001
Rural area 75 44.4 94 55.6
Education <Junior high school 49 33.6 97 66.4 10.344 0.016
High school/technical school 26 27.1 70 72.9
College 71 40.6 104 59.4
University or higher 18 22.2 63 77.8
Occupation Government or public institution 22 19.5 91 80.5 19.521 0.002
Salaried employee 29 27.1 78 72.9
Farmer 27 39.7 41 60.3
Student 42 44.7 52 55.3
Self-employed 24 38.1 39 61.9
Other 20 37.7 33 62.3
Per capita household annual income (CNY) <20000 87 42.9 116 57.1 23.859 <0.001
20000–30000 39 33.1 79 66.9
30000–40000 19 16.2 98 83.8
≥40000 19 31.7 41 68.3


Single-factor analysis of impersonal socio-demographic characteristics influencing patients’ downward referral willingness

The results indicated that the proportion of patients covered by the Urban Employee Basic Medical Insurance (UEBMI) (88.0%) who were unwilling to accept DR was higher than those covered by both the Urban Residents Basic Medical Insurance (URBMI) and the New Rural Co-operative Medical System; the proportion of those having knowledge of the hierarchical medical system and the differential medical insurance reimbursement policy (72.2% and 69.8%, respectively) who were unwilling to accept DR was higher than those lacking that knowledge; the proportion of those with experience of receiving medical treatment from a primary medical institution (67.9%) who were unwilling to accept DR was higher than those lacking that experience; the proportion of patients making their own treatment process decisions (78.7%) who were unwilling to accept DR was higher than those whose medical treatment was decided by family members or through co-determination by the patient and his/her family members. Patients with different types of medical insurance, awareness of the hierarchical medical system, and different decision-making processes showed statistically significant differences in their DR willingness (Table 3).


Table 3: Results of single-factor analysis of impersonal socio-demographic characteristics influencing patients’ downward referral willingness.
Socio-demographic characteristics Willing Unwilling χ2-value P-value
Number Proportion (%) Number Proportion (%)
Medical Insurance Urban Employee Basic Medical Insurance 18 12.0 132 88 61.595 <0.001
Urban Residents Basic Medical Insurance 65 43.3 85 56.7
New Rural Co-operative Medical System 71 50.0 71 50.0
Other 10 17.9 46 82.1
Knowledge of the hierarchical medical system Yes 69 27.8 179 72.2 5.839 0.017
No 95 38.0 155 62.0
Knowledge of the medical insurance reimbursement policy Yes 76 30.2 176 69.8 1.776 0.215
No 88 35.8 158 64.2
Receipt of treatment from primary medical institutions Yes 123 32.1 260 67.9 0.501 0.498
No 41 35.7 74 64.3
Treatment-receiving decision-making process By myself 33 21.3 122 78.7 32.961 <0.001
By family members 38 61.3 24 38.7
Co-determination 86 32.5 179 67.5
Other 7 43.8 9 56.3


Logistic regression analysis of the factors influencing patients’ downward referral willingness

The factors influencing DR willingness, as listed above, that showed statistical significance in the single-factor analyses (i.e., sex, place of registration, education, occupation, per capita household annual income, medical insurance, knowledge of the hierarchical medical system, and medical treatment decisionmaking process) were regarded as independent variables. Patients’ DR willingness was the dependent variable for establishing the logistic regression model. The analysis results indicated that per capita household annual income, medical insurance, and decisionmaking process were the three factors that had significant effects on patients’ DR willingness; patients covered by the New Rural Co-operative Medical System and URBMI, and those whose treatment process was decided by their family members were more unwilling to accept DR; and patients with a per capita household annual income of higher than CNY 30,000 were more willing to accept DR (Table 4).


Table 4: Logistic regression analysis of the factors influencing patients’ downward referral willingness.
Variables B Sig. Exp (b) 95% C.I. of EXP(b)
Min. Max.
Sex
  Female 0.835
  Male 0.050 0.835 1.051 0.658 1.678
Place of registration
  Rural area 0.444
  Urban area -0.253 0.444 0.777 0.406 1.484
Education
  University or higher 0.497
  <Junior high school 0.362 0.386 1.436 0.634 3.25
  High school/technical school 0.465 0.275 1.593 0.691 3.671
  College 0.041 0.912 1.042 0.505 2.151
Occupation
  Other 0.227
  Government and public institution(s) 0.166 0.707 1.180 0.497 2.802
  Salaried employee 0.004 0.992 1.004 0.447 2.257
  Farmer 0.900 0.051 2.460 0.995 6.083
  Student -0.01 0.980 0.990 0.449 2.182
  Self-employed -0.163 0.705 0.850 0.366 1.974
Per capita household annual income
  ≥40000 0.037
  <20000 -0.333 0.376 0.717 0.343 1.499
  20000–30000 -0.083 0.834 0.920 0.423 2.001
  30000–40000 0.659 0.018 1.933 0.846 4.414
Medical insurance
  Others 0.000
  Urban Employee Basic Medical Insurance 0.279 0.577 1.322 0.496 3.524
  Urban Residents Basic Medical Insurance -1.302 0.002 0.272 0.117 0.631
  New Rural Co-operative Medical System -1.603 0.001 0.201 0.078 0.52
Knowledge of the hierarchical medical system
  No 0.794
  Yes -0.06 0.794 0.941 0.598 1.482
Decision-making process
  Others 0.000
  By myself 0.306 0.604 1.358 0.427 4.324
  By family members -1.332 0.032 0.264 0.078 0.895
  Co-determination -0.108 0.849 0.898 0.296 2.721
  Constant 1.812 0.074 6.125


Need for public understanding of the hierarchical medical system

The results indicated that the awareness rate of the hierarchical medical system and the differential medical insurance reimbursement policy was 49.8% and 50.6%, respectively. This indicated that the awareness rate was relatively low, mainly due to insufficient publicity regarding the system, relevant government policies, and health care delivery systems. An understanding of and familiarity with the hierarchical medical system and the effective use of medical resources is a prerequisite for the effective implementation of the hierarchical medical system, which is also a basis for patients to change their DR willingness. In the authors’ opinion, efforts should be made in the following three ways to promote the awareness rate of the hierarchical medical system and relevant policies: (1) publicity for the rationale behind implementing the hierarchical medical system should be strengthened by the government in multiple ways and through multiple measures; (2) higher-level general hospitals (especially tertiary hospitals) must change their operating principle of providing both outpatient and inpatient services and must take responsibility for publicity regarding the hierarchical medical system and relevant policies. First, they should actively suggest that patients in the stable recovery period be referred to lower-level institutions; second, patients with common and chronic diseases who attend the hospital directly should be informed of preferential policies for receiving treatment in primary medical institutions to encourage them to seek treatment in those institutions first in the future; (3) primary medical institutions should make full use of the advantages of being geographically closer to patients[3] and knowledge of the hierarchical medical system should be promoted and popularized through health education.


Need for the development of basic standards of medical resources in primary medical institutions

The results showed that more than two-thirds of the participants stated their unwillingness to be referred to lower-level or primary medical institutions even though they were in the stable recovery phase. This indicates that patients in tertiary hospitals have a relatively low DR willingness and more than 50% of them were unwilling to accept DR due to “the distrust of medical technologies of primary medical institutions.” These results are consistent with the study of Xiaofeng Liu et al.[4] on patient willingness for two-way referral and awareness. This is also the primary reason why patients prefer higher-level hospitals when seeking medical treatment. In other words, the main reason for patients to not accept DR is their distrust of the diagnoses and treatment available in primary medical institutions, and their concerns that receiving treatment from primary medical institutions will lead to a missed diagnosis and prolonged recovery. From this perspective, if the conditions of medical resources in primary medical institutions are not improved and there is no satisfactory basic standard for healthcare resources, it is difficult to change people’s preference for seeking medical treatment from higher-level hospitals. With only “policy guidance” to encourage “voluntary” DR willingness, the effect will be limited.[5]


Need for improving differential medical insurance reimbursement policy

The analysis results indicated that, although patients with different sexes, places of registration, education backgrounds, per capita household annual incomes, medical insurance, awareness of the hierarchical medical system, and decision-making processes showed statistically significant differences in their DR willingness, the logistic regression analysis showed that personal socio-demographic characteristics are not the main influencing factors of DR willingness, which is inconsistent with previous studies on patients’ DR willingness. For example, Runming Zhou et al.’s[6] study suggested that the main factors influencing patients’ DR willingness include patient’s awareness of the two-way referral policy, the status of their family doctor, and their first choice of hospital; while Kuo Gao et al.’s study proposed that the main influencing factors of DR willingness relate to the hospital, including the medical technology level of primary medical institutions, service attitudes, drug categorization, and referral procedures.[7]


In general, the financial capability of patients covered by the URBMI and the New Rural Co-operative Medical Scheme is lower than those covered by the Urban Employee Basic Medical Insurance and commercial health insurance schemes. However, the analysis results showed that patients covered by the URBMI and the New Rural Co-operative Medical Scheme were less willing to accept DR compared to those covered by the Urban Employee Basic Medical Insurance and other types of insurance (such as commercial health insurance). The reasons underlying this result need to be studied further. This result at least suggests that financial capability is not a main influencing factor of DR willingness on the one hand and, on the other hand, the existing differential medical insurance reimbursement policy does not produce an effective incentive effect.[8] Countries with mature medical insurance systems all regard the differential medical insurance reimbursement policy as an essential means to guide patients toward a scientific and organized health-seeking process. Therefore, the Chinese government should further improve the medical insurance reimbursement policy to give play to its function as leverage for guiding patients to seek healthcare rationally during the establishment of the hierarchical medical system.[9] In addition, patients whose treatment processes were decided by their family members were less willing to be referred downward. This might be related to patients’ lack of involvement in the medical decision-making process and overdependence on family members’ decisions.


Solving the difficulties surrounding DR of patients in the stable recovery phase in general hospitals is one of the key steps for developing the hierarchical medical system. The implementation of the hierarchical medical system depends not only on the standardization of the system and the cooperation of medical institutions, but also on the approval and support of the public.[10] It is important to point out that developing a hierarchical medical system cannot fundamentally change people’s choices when seeking medical care. During the improvement of the hierarchical medical system and relevant policies, one should consider not only the successful implementation of the system, but also how the system will influence patients’ DR willingness.


Source of Funding

Funding program: the 2020 research project on socioeconomic development “Empirical Research on Health Care-Seeking Willingness and Behavior of Patients under the Hierarchical Medical System” (2020lslktwzz-018) in Liaoning province.


Conflict of Interest

None declared.


  1. Wang Y, Xiao YL, Cao Y, Zhan TH. [The strategy of promoting the hierarchical medical system under the regional medical association model. ]Jiangsu Healthcare Admin 2016;27:24–26.
  2. Shen Y, Huang WR, Ji SS, Yu J, Li MJ. [Systematic evaluation of current situation, effects and problems of two-way referral in 1997-2017 China.] J Chin General Pract 2018;21:3604–3610.
  3. Zhang T, Chen L, Lin W. [Development Status and Countermeasures of Primary Medical Institutions in Guangdong Province.] Soft Sci Health 2018;32:16–22.
  4. Liu XF, Yang Z, Dai AQ, Li JM, Zhang JW. [The patient awareness of two-way referral under the cooperation of urban hospitals and its causes.] Chongqing Med J 2015;44:2662–2663.
  5. Kong X. [Analysis of the problems and obstacles in the establishment of hierarchical medical treatment system from the perspective of medical resource allocation.] Med Phil 2018;39:66–69.
  6. Zhou R, Cui F, Yao W. [Willingness for downward referral among patients with chronic diseases in Guangzhou and its influencing factors.] Med Society 2019;32:26–30.
  7. Gao K, Gan X. [Behavior decision of two-way referral amount patients in China and its influencing factors.] Chin General Pract 2015;18:3393–3395.
  8. Wang HF, Liu F, Liao XC. [Adapt to the new pattern of the hierarchical medical system and innovate medical insurance payment methods.] Chin Med Insurance 2015;6:12–15.
  9. Jiang Y, Yang J. [The realization of the hierarchical medical system optimization based on the influencing factors of medical treatment selection.] Chin Hosp 2015;19:15–17.
  10. Li Q, Zhou M, Wang Y, Tian LQ. [Analysis on patient medical treatment selection and patients’ personal factors after the implementation of the hierarchical medical system.] Modern Hosp Management 2019;17:39–42.