Biomedical Research

Journal Banner

Development of an evaluation index system for the humanistic qualities of medical practitioners in China

Xiaowei Wu1*, Changkun Luo2, Jianjun Huang3, Dixiong Xu3, Jidong Huang3, Lin Mei1, Jiancai Zhai1,Junguo Chen3 and Yungui Wang2*

1Office of Educational Administration, Third Military Medical University, Chongqing, PR China

2Third Military Medical University, Chongqing, PR China

3Department of Training, Third Military Medical University, Chongqing, PR China

*Corresponding Authors:
Xiaowei Wu
Office of Educational Administration
Third Military Medical University, PR China

Yungui Wang
Third Military Medical University
Chongqing, PR China

Accepted on February 24, 2017

Visit for more related articles at Biomedical Research


Background: A comprehensive evaluation model for the assessment of the humanistic qualities of medical practitioners is not currently available in Mainland China; thus, we aimed to construct one using the Analytic Hierarchy Process (AHP) method.

Methods: We determined the weight of 37 indicators of the humanistic evaluation system.

Results: We found that the consistency ratio index provided by each expert upon grading was less than 0.1. Among all the indicators in the scheme layer, “reverence for life” had the highest weight (0.066), whereas “bioethics” and “ability to express empathy” were 0.053 and 0.047, respectively. The results of the AHP model were better than the results of the mean method model.

Conclusion: The model of evaluation that was established based on AHP is an objective and effective system and also represents a suitable guide for the selection and evaluation of personnel.


Analytic hierarchy process, Medical practitioners, Medical humanistic qualities, Evaluation index system.


Medical humanistic qualities are characteristically based on common human qualities expressed in the medical field. These qualities are an essential requirement for medical professionals and are critical to the medical field. It is well accepted that, for a medical practitioner, medical humanistic qualities are as equally important as medical skills.

With the rapid development of new and improved medical technologies, dramatic changes have occurred in doctor-patient relationships since the 1960s and 1970s [1]. In recent years, increasingly frequent incidents of doctor-patient conflicts and disturbances indicate that the basic doctor-patient relationship is becoming tense, at least in Mainland China. In 2006, the Chinese Medical Doctor Association conducted a questionnaire-based survey in 350 hospitals in China. The data revealed that the occurrence of medical disturbances increased from 89.58% in 2004 to 97.92% in 2006. During this period, the number of medical professionals injured increased from 203 times per year to 332 times per year, with a direct economic loss that increased from RMB 9.87 million to RMB 14.48 million (Rights Protection Committee of Chinese Medical Doctor Association, 2007) [2]. This survey also revealed that many patients do not trust hospitals and doctors, which results in lower patient satisfaction. Therefore, the importance of medical humanistic qualities in clinical practice has widely been acknowledged by the society.

In the United States, the humanistic qualities of clinical physicians have traditionally been evaluated from a patient’s perspective during routine clinical encounters. In this way, the objective is to assess the effects of a clinician’s emotional content and communication skills on treatment outcomes rather than to evaluate the clinician comprehensively. For example, Carol assessed the effects of doctor-patient communication during clinical encounters on the progression of disease by correlating the physician-patient interaction with blood pressure control. The results suggested that effective communication and an emotional connection positively affected treatment outcomes [3]. Timothy evaluated both clinical physicians’ ability to communicate with patients and their humanistic qualities with the assistance of three trained observers, each of whom rated 20 videotaped clinical encounters [4]. The evaluation methodology employed in this study also relied on the paradigm of personalization and intuitive direction. In an even worse situation, the evaluation of humanistic qualities of physicians in Mainland China has not been adequately studied. Therefore, a comprehensive evaluation system based on scientific methods is needed to be developed for the Chinese context.

Exploring the construction of the Humanistic Quality Evaluation System (HQES) for medical practitioners and identifying its basic elements are particularly important for the development and evaluation of the humanistic qualities among medical practitioners. The HQES is a multilevel and complex data system that comprises numerous factors that are mutually associated and controlled. The HQES not only requires an indepth analysis of the hierarchy evaluation of each expert but also combines the advice of every expert to make a final decision. Therefore, the Analytic Hierarchy Process (AHP), which is used in this study, is an effective method to meet these requirements [5,6].

Materials and Methods

We used the AHP method that was slightly modified based on the previous studies [7,8]. The data in this study were collected from 31 provinces over the entire Mainland China, including Beijing, Tianjin, Hebei, Henan, Shan’xi, Shandong, Liaoning, He Longjiang, Nei Menggu, Shanghai, Zhejiang, Jiangsu, Anhui, Guangxi, Fujian, Jiangxi, Hubei, Hunan, Guangdong, Hainan, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shan’xi, Gansu, Xinjiang and Qinghai. The data collection period was from January of 2014 to December of 2015. We declare no conflict of interests in this study.

Establishment of the evaluation index system

A four-layer HQES (Figure 1) was established by adopting multiple research methods, e.g., literature studies, a historical review, expert interviews, and survey questionnaires (Table 1).


Figure 1: Summary graph for the four-layer humanistic quality evaluation system. This system includes target layer (O), criterion layer I (A1, A2, A3), criterion layer II (B1 to B11) and project layer (C1 to C37).

  Expert I Expert II Expert III Expert IV Expert V Expert VI Expert VII Expert VIII Expert IX Expert X Comprehensive weight Weight order
C1 0.021 0.05 0.073 0.09 0.014 0.044 0.039 0.037 0.043 0.053 0.044 4
C2 0.014 0.013 0.073 0.045 0.042 0.022 0.039 0.03 0.008 0.042 0.027 15
C3 0.034 0.153 0.146 0.045 0.042 0.022 0.039 0.024 0.023 0.033 0.053 2
C4 0.024 0.035 0.036 0.045 0.049 0.053 0.007 0.005 0.014 0.009 0.018 26
C5 0.01 0.004 0.036 0.022 0.025 0.018 0.014 0.008 0.076 0.016 0.025 20
C6 0.015 0.035 0.073 0.022 0.025 0.018 0.014 0.015 0.072 0.03 0.034 11
C7 0.011 0.023 0.087 0.027 0.01 0.032 0.017 0.013 0.016 0.048 0.025 19
C8 0.017 0.004 0.029 0.014 0.017 0.008 0.017 0.013 0.016 0.048 0.018 25
C9 0.007 0.002 0.029 0.014 0.004 0.008 0.017 0.013 0.016 0.048 0.016 28
C10 0.025 0.004 0.007 0.015 0.019 0.005 0.005 0.004 0.029 0.007 0.013 35
C11 0.01 0 0.012 0.015 0.006 0.004 0.007 0.004 0.008 0.007 0.007 37
C12 0.025 0.001 0.018 0.015 0.024 0.006 0.012 0.005 0.024 0.011 0.014 34
C13 0.013 0.003 0.018 0.015 0.024 0.007 0.015 0.012 0.026 0.024 0.016 30
C14 0.025 0.005 0.018 0.015 0.032 0.005 0.009 0.012 0.018 0.024 0.016 29
C15 0.027 0.074 0.011 0.032 0.025 0.039 0.005 0.024 0.005 0.028 0.025 18
C16 0.019 0.014 0.006 0.016 0.025 0.021 0.012 0.012 0.003 0.014 0.012 36
C17 0.038 0.015 0.011 0.016 0.025 0.015 0.02 0.017 0.006 0.02 0.018 27
C18 0.013 0.004 0.011 0.016 0.025 0.016 0.034 0.017 0.007 0.02 0.014 33
C19 0.028 0.018 0.048 0.027 0.014 0.023 0.039 0.061 0.02 0.036 0.036 9
C20 0.018 0.071 0.048 0.027 0.043 0.011 0.097 0.038 0.02 0.023 0.038 6
C21 0.023 0.014 0.048 0.027 0.043 0.011 0.061 0.096 0.02 0.057 0.047 3
C22 0.061 0.015 0.022 0.04 0.033 0.061 0.057 0.038 0.028 0.045 0.038 7
C23 0.038 0.003 0.014 0.02 0.033 0.061 0.023 0.038 0.014 0.045 0.027 16
C24 0.096 0.001 0.009 0.02 0.033 0.061 0.036 0.038 0.014 0.045 0.036 8
C25 0.019 0.04 0.004 0.04 0.007 0.036 0.016 0.055 0.013 0.033 0.029 14
C26 0.033 0.01 0.002 0.04 0.012 0.031 0.045 0.033 0.006 0.02 0.022 23
C27 0.039 0.049 0.005 0.04 0.004 0.043 0.032 0.019 0.009 0.011 0.024 22
C28 0.047 0.004 0.01 0.04 0.01 0.072 0.023 0.007 0.005 0.004 0.015 32
C29 0.024 0.117 0.023 0.022 0.049 0.063 0.071 0.016 0.189 0.01 0.066 1
C30 0.012 0.013 0.003 0.022 0.049 0.031 0.018 0.02 0.063 0.013 0.024 21
C31 0.012 0.013 0.003 0.022 0.049 0.031 0.011 0.026 0.063 0.017 0.025 17
C32 0.038 0.101 0.011 0.027 0.089 0.018 0.02 0.031 0.025 0.02 0.036 10
C33 0.024 0.009 0.017 0.027 0.03 0.036 0.031 0.031 0.056 0.02 0.03 13
C34 0.061 0.033 0.027 0.013 0.03 0.009 0.049 0.062 0.022 0.04 0.043 5
C35 0.031 0.004 0.006 0.022 0.004 0.02 0.01 0.024 0.004 0.016 0.016 31
C36 0.016 0.006 0.002 0.022 0.016 0.016 0.016 0.039 0.008 0.025 0.019 24
C37 0.031 0.038 0.003 0.022 0.016 0.026 0.025 0.061 0.015 0.039 0.034 12

Table 1. Project layer weight and comprehensive weight of each expert.

Layer one is the Target Layer (O). This layer consists of quantitative rankings of the evaluation indexes of the humanistic qualities of medical practitioners.

Layer two is Criterion Layer I (A). This layer is the primary evaluation index factors of the HQES, which includes medical humanistic knowledge, medical humanistic ability and medical humanistic spirit-marked as A1, A2 and A3, respectively, in sequence.

Layer three is Criterion Layer II (B). This layer consists of 11 evaluation index factors included in the three primary evaluation index factors and are marked as Bi (i=1, 2, .., 11) in sequence.

Layer four is the Project Layer (C). This layer consists of 37 evaluation index factors derived from Criterion Layer II (B) and marked as Cj (j=1, 2, .., 37) in sequence.

Construction of the judgment matrix

The judgment matrix was constructed by comparing the influence of all the factors on the same layer with the corresponding factors of the upper layer. This means that each factor at the same layer was respectively compared with the corresponding factors (at the upper layer) in pairs according to the evaluation indexes of the HQES.

To compare the influence of the n factors (Y1, Y2, …, Yn) on X (at the upper layer), the proportion of the n factors in X was determined. dij was used to describe the influence ratio of any two factors (for example, Yi and Yj) to X, and the proportion scale of dij (I, j=1, 2, …, n) was marked from 1-9. After this procedure, the judgment matrix D=(dij)n × n was obtained for pair comparisons.

For this example, the formula would obviously be

Dij>0, dji=i/dij, dii=1 (I, j=1, 2, .., n) → (1)

Confirmation and consistency check of weight

The formula for the general judgment matrix Dwas DW=λmax W in which λmax was the maximum characteristic root of D, and W was the corresponding eigenvector of λmax. W can be approximately taken as a weight vector of D after normalization.

A consistency check was an important approach to evaluate the reasonableness of the judgment matrix. The evaluation method was as follows:

CR=CI/RI → (2)

In this formula, CI=λmax-n/n-1, RI was a random consistency index, and the matrices with different orders corresponded to different values. In the case of CR<0.10, the consistency of the matrix was acceptable.

Determination of the weight vector WA of Criterion Layer I (A) to the Target Layer (O) Judgment matrix D0= (dij)3 × 3 was constructed according to the principle of constructing judgment matrices, and D0 was the straight reciprocal matrix of Order 3. The value of λmax and its corresponding eigenvector were solved, along with normalization for this eigenvector. After this process, the weight WA of Criterion Layer (A) to the Target Layer (O) was obtained.


After obtaining the value WA of ten experts with Equation 3, the consistency ratio was calculated with Equation 2. As a result, the CRA of all experts was less than 0.1, which suggests that WA can be used as a weight vector because this constructed judgment matrix satisfies the consistency requirements.

Determination of the combination weight of criterion layer II (B) to the target layer (O)

Judgment matrices D1= (dij)4 × 4, D2=dij)4 × 4 and D3= dij)3 × 3 were constructed by following the principle of constructing judgment matrices and by corresponding to the factor indexes of criterion layer IA. The weights P1B=(P11b, P12b, P13b, P14b) T, P2B=(P21b, P22b, P23b, P24b) T and P3B=(P31b, P32b, P33b,P34b) T of criterion layer II to criterion layer I were then calculated. The combination weight WB of criterion layer II to the target layer IA was


Because CI(B)=CI(B) W0 and RI(B)=RI(B) W0 can be obtained from the above formula, the index of the combined consistency ratio was

CRB=CRA+CI(B)/RI(B) → (5)

The calculated CRB values of all experts were less than 0.1, which complies with the consistency requirement in constructing the judgment matrix. Therefore, WB can be applied as the weight vector of criterion layer II.

Determination of the combination weight WC of the project layer to the target layer

Judgment matrix Dci (i=1, 2,.., 11) was constructed based on the principle of constructing judgment matrices. Each judgment matrix corresponds to the factor index of criterion layer II. The combination weight vector WC of the project layer to the target layer was obtained after calculating the combination weight with the judgment matrix


In this formula, Sijk refers to the weight of index Ck under the index weight wb ij of criterion layer II.

From the formula, CIC=CI(C)WB and RIC=RI(C)WB can be obtained. CI(C) and RI(C) refer to the corresponding consistency index vector and random consistency index vector, respectively, of the judgment matrix in Criterion Layer II; thus, the combined consistency ratio index of the Project Layer was


The calculated CRB values of all experts were less than 0.1, which indicates that the judgment of each expert concerning the entire evaluation system complied with the requirements of consistency. Weight vector WC of the Project Layer can be used as the final weight of each expert. The results are shown in Table 1.

Calculation of comprehensive weight

Because different experts may have different understandings regarding a problem, expert weight can be established according to their levels of understanding concerning this problem. When the distances between the combined consistencies index CRi c of expert i and 0.1 are farther apart, the reliability of the score and the reliability of the weight are higher. Therefore, expert weight was defined as


The final comprehensive weight vector of the Project Layer to the Target Layer was obtained according to the layer analysis weight Wi C of expert i calculated with Equation 6 and the expert weight Wi E calculated with Equation 8. The results are shown in Table 2.

  Mean value Correlation analysis Paired sample t test
Evaluation results of analytic hierarchy process 8.58 r=0.996 t=7.38
Evaluation results of mean value method 8.48 p =3.45 × 10-58 p=8.71 × 10-10

Table 2. Comparison of the evaluation results of the analytic hierarchy process and mean value method.


Results and Discussion

From the weights of the 37 indexes in the Project Layer, “reverence for life” had the highest weight (0.066), followed by “bioethics” and “ability to express empathy”. “Basic principles of medical ethics”, “privacy protection”, “ability to empathize”, “ability to build trust”, “ability to communicate in an influential way”, “empathic concern” and “medical morality” were rated among the top ten. “Perception of traditional and modern medical history” had the lowest weight.

“Reverence for life”, the core, premise and basis of medical humanistic qualities, ranks highest in this field. Albert, a contemporary thinker and one of the greatest humanists of the last century, proposed that expressing a devout attitude towards life is a vital characteristic and behavior style [9]. Accordingly, “reverence for life” is given the highest comprehensive weight

The weights of the three indexes of “ability to empathize” in the Project Layer all ranked in the top 10. Considering the job specifications of medical practitioners, the ability to empathize undoubtedly plays an important role in the doctor-patient relationship. Whether medical practitioners can observe and feel the pain that patients are experiencing determines which emotions they reveal and attitudes they express. The attitude of medical practitioners further influences how patients judge whether they have received sufficient understanding and concern and determines the successful establishment of mutual trust between doctors and patients. Many studies also emphasize the important role that the ability to empathize plays in establishing and maintaining a good doctor-patient relationship. Charles et al. have indicated that the ability to empathize is the spiritual basis of healing relationships [10]. Halpern has also suggested that in internal medicine, accurate medical judgments made between patients and doctors reflect a higher level of empathy, which always leads to improved nursing care and treatment [8].

Regarding the knowledge of medical ethics, two items were rated among the top ten, and their effects on the humanistic qualities of medical practitioners cannot be ignored. Both the rapid development of medical technologies and the changing life experiences of people have created new requirements for medical practitioners in the area of medical ethics, especially in organ transplantation, hospice care, euthanasia, etc. Therefore, medical ethics is inevitably one of the important factors that affect the humanistic qualities of medical practitioners. David et al. have emphasized that medical ethics and humanities education are essential to professional development in medicine [11].

Concerning “clinical communication skills”, two items were rated among the top ten and were notably influencing factors. A good doctor-patient relationship is the basis for diagnosis and treatment, and clinical communication skills are a prerequisite to accurate diagnosis and treatment. Good communication skills are essential to guide patients to accurately describe their symptoms and to understand and accept medical advice. Good communication also promotes mutual cooperation among medical practitioners, the effective management of emergencies and the proper resolution of doctor-patient conflicts. Robyn et al. have found that unsuitable doctor-patient communication was the key reason for a poor doctor-patient relationship [12]. Studies by Fatima and Abdulla have demonstrated that professionalism and communication skills are two significant parts of the entire essence of a doctor's character [13].

To further verify the applicability of the comprehensive evaluation model, 56 persons who had passed the medical licensing examination were interviewed. They were evaluated on 37 indexes of the Project Layer, and the full mark of each index was 10. The comprehensive evaluation model for the humanistic qualities of medical practitioners is as follows:


In this formula, ωk refers to the comprehensive weight in Table 1 and refers to the score of index k of evaluation object t.

In general, the most common application is to take the mean value of the evaluation indexes. However, the indexes vary in terms of the degree of importance compared with the overall system; therefore, the mean value method is not scientifically rigorous. The comprehensive evaluation of 56 subjects with AHP and the mean value method were compared below. The results are shown in Table 3.

Medical humanistic knowledge (A1) Knowledge of medical ethics (B1) Basic principles of medical ethics (C1)
Clinical diagnosis and treatment ethics (C2)
Bioethics (C3)
Knowledge of medical psychology (B2) Basic theory of medical psychology (C4)
Stress and mental disorders (C5)
Psychological intervention and patient psychology (C6)
Knowledge of health law (B3) Management of physicians and medical institutions (C7)
Management of disease control and diagnosis treatment (C8)
Drug and blood management code (C9)
Knowledge of philosophy, history, behavior and sociology (B4) Medical pattern (C10)
Traditional and modern medical history (C11)
Social and cultural aspects of health and disease (C12)
Doctor-patient interaction (C13)
Behavior and health (C14)
Medical humanistic ability (A2) Psychological adjustment ability (B5) Adaptability (C15)
Emotional stability (C16)
Anti-frustration ability (C17)
Growth ability after stress (C18)
Empathic ability (B6) Empathic ability (C19)
Emotional empathic ability (C20)
Ability to express empathy (C21)
Clinical communication ability (B7) Ability to build trust (C22)
Participatory communication (C23)
Influential communication skills (C24)
Clinical thinking ability (B8) Comprehension ability (C25)
Discrimination ability (C26)
Processing ability (C27)
Reflection ability (C28)
Medical humanistic spirit (A3) Humanistic philosophy (B9) Reverence for life (C29)
Respect for patients (C30)
Understanding of patients (C31)
Medical ethics (B10) Medical morality (C32)
Personal qualities (C33)
Protection of privacy (C34)
Sense of responsibility (B11) Self-responsibility (C35)
Team responsibility (C36)
Social responsibility (C37)

Table 3. Evaluation index system for the medical humanistic qualities of medical practitioners.

As shown in Figure 2, the evaluation results of AHP and the mean value method differed greatly. A paired sample t test was conducted for the two evaluation results of the 56 subjects. The results are shown in Table 2.


Figure 2: Comparison of evaluation results between AHP and the mean value method. This plot shows the evaluation results ranging from 0 to 10 for 56 interviewees who had passed the medical licensing examination. Each data point refers to the evaluation value (Y axis) for each interviewee (X axis for the interviewee number). Blue data points indicate the results from the method using the Analytic Hierarchy Process (AHP), while green circles indicate the results from the mean method model (Mean).

Table 2 indicates that (1) the evaluation results of AHP were higher than those of the mean value method, (2) these two results were significantly correlated, and (3) the results of a paired sample t test were t=7.38 and p<0.05, which indicates that the two methods had significant differences. Although the two methods correlated, the evaluation results of AHP were higher than those of the mean value method and better reflected the differences of the indexes. Therefore, AHP was a more scientific evaluation method; (4) according to Equation 9, once 37 results of the evaluation objects were received, the results of the comprehensive evaluation could be obtained.


Our work provides a comprehensive evaluation system that, for the first time, combines three essential elements together: knowledge of medical humanistic qualities, the ability to practice medical humanistic qualities and the spirit of medical humanistic qualities. This combined evaluation model is objective, effective and practical to evaluate the humanistic qualities of medical practitioners and is suitable to guide the selection and evaluation of personnel in practice, at least in Mainland China.

Future Direction

Some limitations of this study and future directions should be mentioned. First, the use of a questionnaire for data collection in the field is known to be often affected by information bias of part of the subjects. However, this problem was minimized by formal training provided by the field investigators prior to the commencement of the study, in addition to other quality control measures that were meticulously implemented. Second, the data collection was restricted to Mainland China. Whether the implication can be applied to other countries requires future investigation. Third, more definitive conclusions could be drawn by increasing the number of samples. Fourth, this is just an evaluation system. More future work is needed to implement this system into the medical education system. Altogether, further large-scale population surveys and interviews that are not restricted to China should be performed to gain a deeper understanding of the importance of humanistic qualities in health management and to provide a more reliable theoretical basis to improve the health management system.


This study was supported by a grant from National Health and Family Planning Commission (NMEC-2014-2). We would like to thank Dr. Yi D for participating in the data processing and Dr. Shen X for a critical reading of the manuscript.