1.INTRODUCTION
Hospitals are the most important health clinic in every country where facing many difficulties as a services organization to improve quality of service. Patient’s satisfaction with hospital services is one of the most important indices of effectiveness and promotion of quality of services at hospitals. Paying attention to patients’ comments helps to improve methods, processes and outcomes; and consequently increase their satisfaction with all the aspects of healthcare services of hospitals. Therefore, identifying and prioritizing the components affecting patients’ satisfaction is crucial in order to improve management planning and decisionmakings and consequently increasing the level of patients’ satisfaction.
1.1.History
In a study entitled management priorities in healthcare clinic of public and private hospitals in Latin America, it was founded that public hospitals faced with the problem of more referrals of patients and need to better management of capacity, facilities and equipment, waiting time and working capital. While, private hospitals face with not applying full capacity of their existing problems, marketing, pricing and demand forecasting (Ketelhöhn and Sanz, 2016).
In another study entitled “Improving Quality in Healthcare service: What makes the satisfaction of a patient?” Among the patients who referred healthcare centers in nine public hospitals between 2006 and 2009 in the United State of America, good correlation coefficients were observed in employees’ behavior with patients and professional capabilities (Más et al., 2016).
A study was conducted on the patients of healthcare centers in Detroit entitled “What’s really important for customers of healthcare services and what plans should be concerned to provide these services” and the following results were obtained (Health Care Consumers, 2016).
According to customers:

Minimum waiting time to visit the doctor, laboratories, hospital waiting rooms and so on.

Physician or other healthcare providers do not rush in examining and visiting.

Physician or other healthcare providers listen to me when I talk.

Physician or other healthcare providers explain clearly what they are doing and what I have to do.
In a study entitled perceived quality of patients in the emergency department and identify areas that need the quality improvement, components like the friendly atmosphere in the workplace, clean environment, supportive environment, combined equipment and skills, specifications of employees were noted as effective components in perceived quality (Muntlin et al., 2006). In a study to evaluate the quality of hospital clinical services, assurance has been introduced the most important component and tangible at the lowest level of importance (Wisniewski and Wisniewski, 2005).
A study was conducted to assess the quality of physiotherapy services, the assurance had the highest priority and tangible was in the lowest priority from perspective of consumers (Adrienne and Emma, 2002). A study on the perceptions and expectations about the quality of dental care in Greek showed that expectations about empathy and assurance is a higher priority (Karydis et al., 2001).
In a study on satisfaction and expectations of patients at Singapore hospitals, the importance of the quality of was different from perspective of consumers, so that assurance has the utmost importance, and responsiveness component was the least important (Lim and Tang, 2000). In another study examining the perceptions of service quality with the expectations of patients, assurance was of utmost importance and components of tangibles, reliability, responsiveness and empathy were in the next ranks (Dotchin and Oakland, 1994).
2.THEORETICAL FOUNDATIONS of RESEARCH
2.1.Satisfaction and Quality of Services

Customer satisfaction is defined as the differences between customer’s expectations and the perceived quality (Oliver, 2010).

Customer satisfaction is the consumers’ feeling or attitude towards the product or service after being used (Jamal and Naser, 2002).

Satisfaction is associated with subjective assessment of feelings (Caruana, 2002).

Customer satisfaction is defined as the difference between the ideal and the actual characteristics of a product or service (Pfaff, 1977).

Quality is customers’ judgment about the superiority or advantage of an object (Parasuraman et al., 1991).

The quality of service means compliance services with customer’s expectations (Van duong et al., 2004).

Quality has no meaning other than what the customer really wants. In other words, a high quality product is adapted with customer’s requirements and demands (Crosby, 1984).

Perceptions reflect customer’s evaluation of the quality of provided services and expectations are customer’s demands of service providers (Chin and Pun, 2002).

In terms of quality systems 2000, quality is defined as all the features that meet customer’s requirements (Kotler, 2000).

International Standards Organization defines the quality as: the features and characteristics of the product or service that has the ability to meet customer’s requirements.
2.2.Hospital
The main mission of hospitals is to provide high quality healthcare services for patients and meet their needs and expectations. Fulfilling this important mission requires institutionalized quality in hospitals.
National System of Health in England (NHS) defined quality of healthcare as following: “Quality is providing the right services to the right people through appropriate and practical methods for middlelevel population and in compliance with ethics and humanity” (Haddad et al., 2000).
3.THEORETICAL FRAMEWORK
3.1.Multi Attributes Decision Making Techniques (MADM)
Multi Attributes Decision Making Techniques with various and sometimes contradictory indices are widely used in decision makings. This technique has been used by various decision makers, because of its potential to reduce the complexities in decisionmaking, combined use of qualitative and quantitative indices, creating a structured framework in the decision making problems and ultimately ease of use. This technique formulates the problems of decision making in the form of a matrix as in Table 1 and the required analysis is done on them.
In this matrix, A_{i} represents option i, X_{j} represents index j, and x_{ij} is index value j for option i.
In this study, three most commonly used techniques in the field of multiattribute decision namely TOPSIS and SAW and ELECTRE techniques have been used. Also, description of Entropy methods, Copeland Method and SERVQUAL model will be presented.
3.2.Shannon Entropy Method
This method has been presented by Shannon in the late 40’s. Entropy method is used to obtain the relative weights of the criteria. Entropy reflects the amount of uncertainty in the content and will be determined based on the data dispersion, importance and weight of each index, and procedures are as follows:

First Step: formation of decision matrix:
This step is explained in Table 1.

Second Step: convert decision matrix to non Scale matrix (normalized decision matrix)(1)(2)

Third Step: calculate the entropy (E_{j}) according to the following formula:

Fourth Step: Calculation of uncertainty for each parameter:
D_{j} for each index is calculated using the following formula that reflects the uncertainty of the data for the index j.(3)

Fifth Step: Calculate the weight of each index:
3.3.SAW
SIMPLE ADDITIVE Weighted; i.e. SAW is one of the easiest methods of multiattribute decisionmaking. In this technique, by estimating the weight (W) of indices, the most appropriate option (A^{*}) was calculated as follows.(6)
3.4.TOPSIS
TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method was proposed by Hwang and Yoon (1981). In this method, index of m will be evaluated by the option of n. This technique is based on the concept that the choice should has the least distance with the ideal positive solution (best case, possible, ${A}_{i}^{+}$ ) and maximum distance from the ideal negative solution (worst case, possible, ${A}_{i}^{}$).
TOPISIS problem solving method includes six steps as follows:

First Step: convert decision matrix to nonscale matrix (normalized decision matrix):(7)

Second Step: forming nonscale weighted decision matrix:(8)
$${V}_{m\times n}={\text{N}}_{\text{m\xd7n}}\hspace{0.17em}\cdot \hspace{0.17em}{\text{W}}_{\text{n\xd7n}}$$(8)where, V is a nonscale weighted matrix and W is diagonal matrix of weights of indices and has been weighted using Shannon Entropy Method.

Third Step: Determine ideal positive and negative alternative option using the following formula:
For ideal positive option (A^{+}) and negative ideal (A^{−}) we have:(9)(10)(11)
$$\begin{array}{c}{A}^{+}=\left\{\left(\text{max}\hspace{0.17em}{v}_{ij}{I}_{j}\in J\right),\hspace{0.17em}\left(\text{min}\hspace{0.17em}{V}_{ij}{I}_{j}\in {J}^{\prime}\right)\right\}\\ =\left\{{V}_{1}^{+},\hspace{0.17em}{V}_{2}^{+},\hspace{0.17em}\cdots ,\hspace{0.17em}{v}_{n}^{+}\right\}\end{array}$$(9)$$\begin{array}{c}{A}^{}=\left\{\left(\text{min}\hspace{0.17em}{v}_{ij}{I}_{j}\in J\right),\hspace{0.17em}\left(\text{max}\hspace{0.17em}{v}_{ij}{I}_{J}\in {J}^{\prime}\right)\right\}\\ =\left\{{V}_{1}^{},\hspace{0.17em}{V}_{2}^{},\hspace{0.17em}\cdots ,\hspace{0.17em}{V}_{n}^{}\right\}\end{array}$$(10)$${J}^{\prime}=\left\{j=1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}\cdots ,\hspace{0.17em}n\right\}\hspace{0.17em}\stackrel{j}{\to}\text{cost}$$(11)where ${v}_{j}^{+}$ and ${v}_{j}^{}$ respective indicate positive and negative ideal solution for index j.

Fourth Step: calculating distance of options by positive and negative ideal options using the following formula:(12)(13)
$${d}_{\text{i}}^{\text{}}{\left\{{\displaystyle {\sum}_{j=1}^{n}{\left({v}_{ij}{v}_{j}^{}\right)}^{2}}\right\}}^{\frac{1}{2}},\hspace{0.17em}\left(i=1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}\cdots ,\hspace{0.17em}m\right)$$(12)$${d}_{\text{i}}^{+}{\left\{{\displaystyle {\sum}_{j=1}^{n}{\left({v}_{ij}{v}_{j}^{+}\right)}^{2}}\right\}}^{\frac{1}{2}},\hspace{0.17em}\left(i=1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}\cdots ,\hspace{0.17em}m\right)$$(13)That ${d}_{\text{i}}^{+}$ and ${d}_{i}^{}$ respectively indicate the ideal positive and negative solution for option i.

Fifth Step: calculate the relative closeness of each of the options to ideal solution and is presented by C_{i}.(14)

Sixth Step: prioritizing the options
In this stage, every option with larger C_{i} is better.
3.5.Elimination et Choice in Translating Reality (ELECTRE)
ELECTRE was introduced in the late 1980s. Basis of this concept is superior rating relationship. i.e. does not lead to a ranking of options, but may eliminate options. The steps of this technique include:

First Step: convert decision matrix to nonscale weighted matrix(15)

Second Step: forming a nonscale weighted matrix(16)
where, V is a nonscale weighted matrix and W is diagonal matrix of weights of indices and has been weighted using Shannon Entropy Method.

Third Step: identifying the coordinated and uncoordinated set:
In this stage, all options are evaluated than all indices and set of coordinated and uncoordinated matrices is formed.
If the index has positive aspects: ${s}_{k,\hspace{0.17em}L=\left\{j{x}_{kj\le {x}_{1j}}\right\}}$
If the index has a negative aspect: ${D}_{k,\hspace{0.17em}L=\left\{j{x}_{k,j\le {x}_{l,j}}\right\}}$

Fourth Step: calculating coordinated and uncoordinated matrix:(17)
$$N{I}_{k,\hspace{0.17em}L}=\frac{{\text{max}}_{j\in {D}_{k,L}}\left{V}_{kj}{V}_{lj}\right}{{\text{max}}_{j\in j}\left{V}_{kj}{V}_{lj}\right}$$(17)Coordinated matrix is a square m×m matrix in diameter with no element. Other elements of the matrix are obtained by total weights of indices belonging to the coordinated set. Uncoordinated matrix as a m×m matrix in diameter with no elements and other elements of this matrix are nonscale weighted.

Fifth Step: Create effective coordination matrix using the following equation:(18)

Sixth Step: Create effective inconsistent matrix(19)

Seventh Step: determine the overall effective matrix H
${h}_{k,\hspace{0.17em}l}={f}_{k,l}\times {g}_{k,l}$
This matrix reflects the superiority of different strategies to each other: i.e. if , 1 = h_{k,l} can be said A_{K} preferred on A_{L}.

Eight Step: Remove effective options. Superior option is an option that has the least number of 1 in column.
3.6.Copeland Method
In this method, the number of wins and losses for each index is specified through the formation of pair comparison matrix and pairwise comparison of all the indexes, and the score for each index is calculated and prioritized by mining the number of losses than wins.
3.7.SERVQUAL Model
In the mid1980s, Parasuraman, Berry and Zeithaml (1990) studied the indices of quality of service and quality of service perceived by customers based on the quality of their services. As a result of studies later, 10 components of quality of service (due to the high correlation between some of the components) was reduced to 5 components (Parasuraman, 1990). Including physical and tangible, assurance, responsiveness, guarantee, empathy. Price component is added to this study derived by European Union Satisfaction Model.
4.RESULTS
The is an applied study in terms of objective. method and nature of the study is descriptivesurvey. Statistical population consisted of all the patients who were admitted in one of the clinics of public hospitals in Tehran in 2016.
Data were collected using a developed SERVQUAL questionnaire composed of Likert scale and the item of price. The Statistical sample consisted of 60 subjects with academic education who admitted in the clinic more than once. (It should be noted in these techniques, number of statistical sample is not relevant and expertise of the subject is important.) reliability of the data obtained from the questionnaire were estimated at rate if 0.89 using Cronbach’s alpha and given the standardized questionnaire used in different studies, its validity is confirmed.
Then, decision tree composed of 22 columns (number of indices) and 60 rows (number of subjects) was formed using the collected data and the weight of indices was determined using Shannon Entropy. Then, multiattribute decisionmaking techniques such as SAW, TOPSIS and ELECTRE were used to prioritize the components affecting patients’ satisfaction and finally, the results of different techniques were combined using Copeland Method and a uniform ranking of components was obtained.
In order to prioritize the components affecting patient’s satisfaction, the significance of the differences between these components must first be confirmed. For this purpose, we have:

H0: The importance of all the components are identical.

H1: There is a difference between the least importance of two components.
According to the results of variance analysis (ANOVA), the significance level was less than 0.05; and therefore, difference assumption in at least two components is confirmed. So, we can prioritize the components.
Therefore, importance coefficients of indices were determined using Shannon Entropy and the identified components were prioritized using SAW, TOPSIS and ELECTRE techniques and the results are presented in Table 2. Then, results of different techniques were com bined using Copeland Method and a uniform ranking of components was obtained.
According to data in Table 2, for more explanation of these techniques, different ways that prefer fifteen indices on sixteen indices are two methods including (SAW and TOPSIS) and therefore, only (ELECTRE) method prefers sixteen indices to fifteen indices prefers. Therefore, according to most methods, rather than indices of fifteen to sixteen indices. Therefore, final prioritize of indices in the Copeland column of Table 2.
5.CONCLUSION
The topic of improving patients’ satisfaction with healthcare services is necessary for management planning and decisionmaking. Therefore, aim of this research is to identify and prioritize the components affecting patients’ satisfaction with healthcare services using multiattribute decisionmaking techniques.
So, based on developed SERVQUAL model (based on the European model, cost component was added) approved by professors and experts in this field, 22 components affecting patients’ satisfaction were identified and were completed in a form of questionnaire by 60 patients with academic education who referred more than once to the clinic and after data extraction, the components were prioritized using various multiattribute decision techniques and finally, the results of different techniques were combined using an integrated model and a uniform rating was obtained.
The results of this study indicate that the most important components affecting patients’ satisfaction with healthcare services from the perspective of consumers,’ respectively are “employees with sufficient knowledge and skills,” “modern equipment in clinic” and “peace and security at clients “and” confidence to employees “and” polite and humble employees “and the least important components were respectively” payment term “and” appropriateness of the cost of consumables “and” appropriateness of the cost of services.”
Therefore, according to the ranking of indices in Table 2, “guarantees” and then “physical and tangible” were the first and second priorities to improve satisfaction from the perspective of consumers,’ and then, “confidence” and “sympathy” and “responsiveness” and at last “cost” were noted.
It seems, in public hospitals and the use of insurance services, recipients of health services have no particular problems regarding the payment of fees in clinics, but specialist physicians are expected to present in public hospitals and also is expected to employ polite and humble employees and the latest medical facilities to feel peace, security and confidence to employees and just to think about treatment. Therefore, the following suggestions are presented in accordance to the results of present study:
5.1.Suggestions

1. According to the high visiting of people to public hospitals due to lowcost, some employees in the hospital may deal the patients in bad behavior and this causes patients’ dissatisfaction. Therefore, it is recommended to hold courses for employees on standardized and necessary training to behave patients with respect, kindness and humility and explain the approach of “customer is always ight.” In the hospital and this importance should always be monitored through the people.

2. In public hospitals, specialist, experienced and compassionate physicians should be employed and for this purpose, physicians’ satisfaction should be provided and this importance should be monitored always by the people.

3. Based on the comment of customers, the best employees and physicians should be appreciated in their service and record their names.

4. Public hospitals should be equipped with modern and updated medical facilities to serve the lowincome patients and not impose great cost private hospitals.