*4.2. Stages of This Research*

The stages of development of the SHSCPM model by combining the balanced scorecard with DEMATEL and ANP, and using Super Decisions software version 2.8.0 for ANP data processing, were based on the respondents' perceptions. The research stages were as follows:


#### *4.3. Determination of BSC Strategy Map by DEMATEL*

The DEMATEL used for design of strategy map. Survey result was processed for the DEMATEL steps. Survey to identify level of influence between perspectives and indicators with 0–4 scale (0 = no influence, 1 = low, 2 = normal, 3 = strong, 4 = very strong). The BSC strategy map was designed by using DEMATEL with following steps:

a. Building direct relation matrix (*A*) based on average of influence value from ai to aj by survey:

$$A = \begin{bmatrix} a\_{ij} \end{bmatrix}\_{n \ge 0}$$

$$\begin{bmatrix} 0 & a\_{12} & \dots & a\_{1n} \\ a\_{21} & 0 & \dots & a\_{2n} \end{bmatrix}$$

⎤


$$\mathbf{X} = \mathbf{A} \times \mathbf{s}$$

$$s = \min\left[\frac{1}{\max\sum\_{j=1}^{n} a\_{ij}}, \frac{1}{\max\sum\_{i=1}^{n} a\_{ij}}\right].$$

c. Building total relationship matrix (T), T = X(I − X) −1 ; I is a identity matrix. Calculating of importance level and influence level of the perspective and indicator. Element of T <sup>=</sup> *tij nxn*, *<sup>i</sup>*, *<sup>j</sup>* <sup>=</sup> 1, 2, ... , *<sup>n</sup>*; where, *<sup>i</sup>* <sup>=</sup> rows, and *<sup>j</sup>* <sup>=</sup> columns; D and R represent direct and indirect relationships from rows and columns:

$$\mathbf{D} = \left[\sum\_{j=1}^{n} t\_{ij}\right]\_{\text{xx1}}, (i = 1, 2, \dots, n)$$

$$\mathbf{R} = \left[\sum\_{i=1}^{n} t\_{ij}\right]\_{\text{xx1}}, (j = 1, 2, \dots, n)$$

where, (D + R) indicates of importance level and (D − R) indicates of influence level.

	- Calculation of the average of T: *X* = <sup>T</sup> *ij* ; where, *<sup>i</sup>* <sup>=</sup> the sum of the row and *<sup>j</sup>* <sup>=</sup> the sum of the column.
	- Reduction of all T with *X* or (T − *X*), the value of the significant matrix indicates the level of the relationship

The direct relation matrix is shown in Table 2, and Table 3 shows the total relationships matrix. Then, the level of importance and influence is shown in Table 4. Finally, the significance matrix has illustrate of relationships between perspectives and indicators shown in Table 5.

Based on the level of importance (D + R), the important indicators on the SHSCPM are quality of service, stakeholder satisfaction, customer satisfaction, flexibility, collaboration with supplier, standard of service, innovations, and organization behavior. These are incorporated in the customer perspective and innovation and learning perspectives. After that, all of indicators incorporated into the innovation and learning perspective have the most influence compared to other indicators, with the values of all of the influence levels (D − R) being positive (+).

Finally, the indicators incorporated into the customer perspective and innovation and learning perspectives are closely related with human resources, so the existence of human resources is important in the SHSCPM.

Directrelationmatrixasinitialmatrixbasedonexpert




**Table 4.** Level of importance and influence.

#### *4.4. The ANP Structure Model*

The ANP structure model was designed by BSC strategy map. The ANP is used to determine weight of perspectives and indicators performance. The ANP processes using survey of expert respondents. Survey has used to perform a pair-wise comparison between indicators with a scale of 1–9. Survey results were processed by ANP steps. The validity of the pair-wise comparison has considered the inconsistency value. If value of inconsistency <1, so the pair-wise comparison is valid. The ANP uses an initial matrix derived from average value of the survey result.

Weight calculation of perspectives and indicators on ANP can be process by inputting average values of pair-wise comparison into software of super decision version 2.8.0. The value of the consistency ratio can also be seen directly in the super decision software after values of pair-wise comparison inputted into the software system.

The structure model illustrates the inner dependence and outer dependence in this study shown in Figure 2. Inner and outer dependence were designed by relationships between perspectives and indicators as significant matrix. Inner dependence is relationship between indicators based on the same perspective. Outer dependence is relationship between indicators with other indicators based on the different perspectives.


*Sustainability* **2019**, *11*, 3626

**Figure 2.** Inner and outer dependence.

Figure 2 makes two phenomena clear: the innovation and growth perspective is the most influential because it has an influence on all of the other perspectives, and the customer perspective is the most important perspective because it is influenced by all the other perspectives. Besides, performance indicators by inner and outer dependence can explain using examples: Demand indicator (X1) on the financial perspective has inner dependence with effectiveness (X2), efficiency (X3), profit (X4), and revenue (X5), and then, indicator of demand (X1) has outer dependence with indicators on the customer perspective (i.e. quality of service (X6), delivery (X7), customer satisfaction (X8), patient loyalty (X9), and stakeholder satisfaction (X10)). Furthermore, the organization behavior indicator (X29) on the innovation and learning perspective has inner dependence with capacity and professionalism (X24), innovation (X25), training and education (X26), research and development (X27), and health and safety (28). Besides, the organization behavior indicator (X29) on the innovation and learning perspective has outer dependence with all of indicators on the other perspectives.

## **5. Results and Discussion**

This study uses Super Decisions software version 2.8.0 for data processing of the ANP. The data processing is the result of pair-wise comparison between performance perspectives and/or performance indicators derived from the inner dependence and outer dependence. Hence, the values of the pair-wise comparison are based on the preferences of the expert respondents.

#### *5.1. Data Collection*

Pair-wise comparison by the expert respondents' perceptions is used to collect the requirement data. Then, the data has been calculated to determine weights of the perspectives and indicators. Table 6 shows the results of pair-wise comparison for the performance perspectives.


**Table 6.** Normalized weight of pair-wise comparison between the performance perspectives.

Validation of the pair-wise comparison result is based on the value of the inconsistency ratio. For pair-wise comparison of the performance perspectives, this value is 0.079, where 0.079 ≤ 0.1, which means that this perception by the respondents is valid.

#### *5.2. Influence Analysis between Perspectives Based on Indicators Relationship*

The influence between perspectives can be determined based on the relations between indicators. The Super Decisions software processed the values of influence between performance perspectives shown in Table 7.


**Table 7.** Influence between performance perspectives.

Table 7 shows that the performance indicators incorporated in the perspectives of innovation and learning had an influence on all of the perspectives, including a self-influence. The performance indicators in the information perspective, operational perspective, and financial perspective too have the greatest influence on the customer perspective. Besides, the customer perspective has the highest value of self-influence. From these phenomena, the customer perspective is seen to be the most important in the SHSCPM.

#### *5.3. Weight of Performance Perspectives and Performance Indicators*

Based on the Super Decisions processing, the indicator weights on the cluster (perspectives), indicator weights in the SHSCPM system, and the perspective weights can be calculated. Table 8 shows the calculation of the weights of the perspectives and indicators.


The indicator weights on the clusters in Table 8 indicate that profit and demand (patient) are very important for the financial perspective, while quality of service and stakeholder satisfaction are very important for the customer perspective. Then, collaboration with the supplier is very important for the operational perspective. Finally, the information perspective and innovation and learning perspective are not important indicators, because all of the indicators in this cluster have a value of zero (0).

Beside, big five dominant indicators in the SHSCPM are quality of service (0.2296), profit (0.1442), stakeholder satisfaction (0.1440), patient loyalty (0.1234), and customer satisfaction (0.1212). Furthermore, based on the performance perspectives, the customer perspective has the greatest weight compared with the other perspectives and is, therefore, the most influential on the performance of the SHSCPM, with a weight of 0.6331. Next, the financial perspective has a weight of 0.3021, followed by the operational perspective, with a weight of 0.0648. Finally, the perspectives of information and innovation and learning have no weight, so these perspectives have the least influence on the performance of the SHSCPM.

#### *5.4. Weight of Sustainability Aspects*

The weight of sustainability aspects can be processed by calculating the indicator weights based on the sustainability aspect classification. Table 9 shows the weights of the sustainability aspects based on the indicator weights.


#### **Table 9.** Weights of sustainability aspects.

Table 9 shows that the economic aspect has the greatest weight, and the environmental aspect has the lowest weight compared to the others. The social aspect was ranked second after the economic aspect, so this aspect was considered feasible in the SHSCPM. From the weight of the indicators, the performance indicators incorporated in the economic aspects still dominate compared to the others. The performance indicators incorporated in the environmental aspect have small weights, which means that environmental factors receive less attention from all the actors in the healthcare supply chain in Indonesia's Province of East Java.

The weight of the social aspect is 0.241, which means that social factors are highly regarded by actors in the healthcare supply chain in the Province of East Java. This is different from the opinion that social aspects were less explored in sustainable supply chain [14,37].

#### **6. Conclusions**

Integration of BSC with DEMATEL and ANP is a new model for measuring performance of sustainable SHSC. This model has more comprehensive with other models because all of the supply chain actors have involved to determining of performance indicators, strategy map, and weight of the performance indicators.

This study has been using five perspectives with twenty-nine indicators. The performance indicators included intangible characteristics and sustainability aspects. The performance indicators that reflect this intangibility are related to information and human resources, while performance indicators that reflect sustainability are related to economic, environmental, and social factors.

There are three major findings in this study. First, from the BSC and DEMATEL, the indicators incorporated in the customer perspective and innovation and learning perspectives were important, so the indicators incorporated into the innovation and learning perspective were the most influential on other indicators. Second, from DEMATEL and ANP, the innovation and learning perspective had the most influence on other perspectives, but on the other hand, this perspective was not important because it did not affect the performance value. Thus, the customer perspective is the most important because it has a major influence on the performance value. Third, based on the weights of the sustainability aspects, the economic aspect has the greatest weight, and the environmental aspect has the least weight compared to the others. The social aspect was ranked second after the economic aspect, so this aspect was considered feasible in the SHSC performance. Finally, the environmental aspect in the SHSC receives less attention in the healthcare business in East Java Province, Indonesia. Based on DEMATEL and the indicator weights from ANP, the performance indicators incorporated in the financial perspective and customer perspective are drivers of the SHSC's performance. The indicators driving the SHSC's performance consist of profit, quality of service, revenue, customer satisfaction, and stakeholder satisfaction. Besides, the performance indicators incorporated into the economic aspect of sustainability have the greatest effect on SHSC performance, followed by the social aspect and the environmental aspect. Furthermore, this study found a contradiction in the social aspect of sustainability, which received less attention than in some other studies, although this was an important aspect of sustainability after the financial aspect. Finally, human resources, as an intangible asset, are the main factor in the SHSC because they have a significant effect on the improvement of performance, especially from the customer perspective, innovation and learning perspective and the social aspect of sustainability.

Implementation of SHSCPM by integration of BSC with DEMATEL and ANP may help the management of the healthcare business to give more attention to human resources as one of intangible characteristics of a healthcare business, especially for innovation and organization behavior, because its exerts the most influence to other indicators. Beside that, the management of the healthcare business must be maintain customer satisfaction, patient loyalty, collaboration with suppliers, and stakeholder satisfaction as parts of social aspects for performance increasing. The SHSCPM model by combining BSC with DEMATEL and ANP can help management in the healthcare business to manage the company performance with simultaneously attention to intangible and sustainability aspects.

As with most empirical research, this article has limitations. First, the survey method with in-depth interviews was used with just seven expert respondents. So, the study could be improved by using more expert respondents. Second, the current study covered only East Java Province, so the findings cannot be generalized to other provinces in Indonesia.

Future research needs to explore the social and environmental aspects of sustainability in SHSCPM, where a contradiction was found. Furthermore, the SHSCPM system needs a new design with a simulation of system dynamics for predicting SHSC performance in the future based on the strategy map, weights of performance indicators and in the past performance values. The new design can help the healthcare business to prepare better its operations in order to achieve a high of the SHSC performance.

**Author Contributions:** Conceptualization, E.B.L.; Methodology, all authors, writing—original draft preparation, E.B.L.; writing—review and editing, all authors.

**Funding:** Partially of the research was funded by the grant scheme for doctoral dissertation research, Ministry of Research, Technology and Higher Education of the Republic of Indonesia, 2018.

**Acknowledgments:** This article is part of doctoral dissertation research at Department of Industrial Engineering, Institut Teknologi Sepuluh Nopember (ITS), Surabaya, Indonesia.

**Conflicts of Interest:** The authors declare no conflict of interest.
