Big Data Analytics Implementation and Practices in Medical Institute Libraries of Pakistan

1 Introduction

The effects of information and communication technology (ICT) and the rise of the digital age have stimulated the creation of a vast array of digital items in different parts of the world (Ahmad et al. 2019). In this digital age, there is an information flood and explosion since data is being produced at an exponential rate because of the development of the internet, mobile technology, sensors, and cloud computing (Giannakis et al. 2019). Over the past two decades, big data has been extensively addressed in a variety of sectors (Djafri et al. 2018), however, until now, experts have been unable to agree on a single definition of big data. It is unclear what amount of data will be considered big; nonetheless, some data scientists have indicated that big data (BD) has data in the Exabyte range, while others have discussed zettabytes/yottabytes (Anna and Mannan 2020). Various specialists have discussed the same concept based on their unique points of view and facts, leading to big data being defined as “high-volume, high-velocity, and/or high-variety information assets that necessitate innovative and cost-effective analyzing data methods to facilitate decision making and future predictions” (Zhang et al. 2018). Elsewhere, the subject of who initially created the term “Big Data” has also been contested. According to reports, John Mashey invented the term “Big Data” during a table discussion at Silicon Graphics In-charge (SGI) in the mid 1990s (Diebold et al. 2012). However, some experts believe Michael Cox and David Ellsworth invented the term Big Data in 1997 at an IEEE conference (Wang et al. 2018).

Big data can be categorized into three categories: volume, velocity, and diversity (Gandomi and Haider 2015). According to Kubick: “Big data has been used in data sets, and these data sets have become very large. The use of traditional database management systems has been incapacitated. Their data sets are larger than the ability of common software tools and storage systems to collect, store, manage, and process data at a reasonable time” (Kubick 2012).

Similarly, Halamka (2014) declared, “Big Data analytics (BDA) mean an organization’s ability to fully leverage its digital assets, which collectively include a large amount of data and more.” Herland (2019) also suggested that new technologies should be applied to the health sector to reduce the gap between patients and doctors. By implementing modern technology in the health sector, patients will get proper treatment at a minimal cost (Islam et al. 2019). Elsewhere, Teets and Goldner (2013) have said that library data and big data are inextricably linked, with rising library materials/items contributing to big data.

Library professionals recognize the significance of BDA in library operations (Zhan and Tan 2020). As a result, Ambigavathi and Sridharan (2018) claimed that BDA is bringing substantial changes in the ease with which data may be transformed into important insights. Libraries have been impacted by BD in two ways: first, the system’s capacity and operational needs are overwhelming due to the large volume, speed of the concerned resources, and selection. Second, because BDA methodologies and algorithms are complex procedures, BDA is entirely an IT-based endeavor (Kamupunga and Chunting 2019). To implement these unavoidable changes in current libraries, they need be turned into library 4.0, which means a watchful library that can assess massive amounts of data and give customers the required information in the shortest period of time (Anna and Mannan 2020). Furthermore, Noh (2015) explained that Library 4.0 means a library that has the ability to handle BD for better use. This is the primary responsibility of library professionals, to analyze the existing data of libraries for the purpose of achieving and fulfilling the institutional strategic goals (Islam et al. 2020).

In this study, a medical institute is defined as any institution that provides health-related programs such as Bachelor of Medicine and Bachelor of Surgery (MBBS), Bachelor of Dental Surgery (BDS), Nursing, Doctor of Physical Therapy (DPT), and Homoeopathy, among others. It is also affiliated with a medical university recognized by the country’s national medical council/commission. The Pakistan Medical and Dental Council (PMDC), now known as the Pakistan Medical Commission (PMC), is the certifying organization for medical institutes in Pakistan. In Pakistan, medical libraries refer to libraries housed at medical institutions and teaching hospitals. In the country, there are 369 medical institutes and teaching hospitals (PMDC 2019).

Medical library users save lives. Requiring up-to-date, accurate, and relevant resources to meet their everyday information needs (Ebenezer 1993), medical librarians can use BDA to deliver information to concerned consumers in a very organized manner based on their needs (El-Seoud et al. 2017). Similarly, Islam et al. (2019) claimed that BDA can also be used to provide more effective information to health practitioners, educators, students, researchers, and paramedical workers. The primary goal of this research is to assess the level of understanding, usage, and applicability of BDA in Pakistani medical libraries. Efforts have also been undertaken to identify the present practices, requisite skills, and capacities of these institutions’ medical librarians.

Due to the emergence of modern technologies, the roles and approaches of library professionals have changed significantly, with Anna and Mannan (2020) stating that medical library practitioners are now dealing with a wide range of huge data practices. The researchers have tried to assess the possible implementation of the BDA in medical library centers in the country.

The following are the research objectives of the current study:

1. To determine the degree to which medical librarians perceive ease of use of BDA in medical libraries of Pakistan.

2. To identify the perception of medical librarians about the usefulness of BDA to improve their job performance.

3. To identify the Behavioral Intention (BI) of medical librarians.

4. To determine the skills of these librarians towards applying for BDA.

5. To explore the current practices of Big Data Analytics in these medical institutes.

6. To evaluate the performance of medical librarians.

In the medical library, BDA is an effective tool used to manage past and present data for easy access and use. Using this study, we aim to regulate the level of understanding, current skills, and capabilities of medical librarians at these institutions. Its findings will be a valuable resource for medical librarians, healthcare researchers, high level individuals of health departments, and policymakers associated with the medical field in Pakistan, and their implementation will help improve the existing situation of medical libraries. By providing the latest and required information to medical stakeholders, we can prevent diseases, thereby supporting the establishment of a healthy and prosperous Pakistan.

2 Literature Review

The notion of “Big Data” is not new; however, its definition is constantly changing due to the needs at any moment in time. Ward and Barker (2013) stated that there are almost as many definitions as the number of people you ask. Various scientists have defined it in different ways, but Big Data (BD) can be compared to a collection of items whose size, velocity, kind, and complexity mean that one has to adopt and discover new software and hardware setups to effectively analyze, store, and visualize the data (Liu et al. 2018). Digital data, in all formats and sizes, has been growing at astounding rates since the National Security Agency (NSA) reported over a decade ago that the internet was processing 1826 Petabytes of data daily (Jarvis 2013). Silva et al. (2019) described Big Data as a 3Vs model, namely volume, variety, and velocity. Nevertheless, another definition of BD given by IBM is widely accepted as it altered the 3Vs into a 4Vs model by incorporating veracity as the fourth V of any Big Data Model (Gandomi and Haider 2015).

BDA refers to the technology and processes used to find, store, transfer, analyze, and visualize large amounts of unstructured and structured data (Erevelles, Fukawa, and Swayn 2016). Similarly, Manyika (2011) stated that Big Data analytics might enhance the global economy by being actively used in a variety of fields. BDA can be used to improve efficacy and efficiency, and can be defined as a collection of data management and analytical methodologies and skills for managing massive (Terabytes and Exabytes) and miscellaneous data sets. Furthermore, the BDA framework requires modern data storage, data management, data analysis, and visualization technologies (Chen et al. 2012). Tebboune et al. (2016) stated that BDA has created a paradigm shift in the area of scientific data retrieval after overcoming several problems such as data capture, management, storage, analysis, and visualization.

3 Materials and Methods

3.2 Proposed Framework

Based on the reviewed literature and study of different theories and models about the technology acceptance model, the researcher has developed the following conceptual model and hypotheses. The current study is based on TAM, considering its importance and unique role among the technology acceptance models as given in Figure 1.

Figure 1:

Technology acceptance model (TAM) (Davis 1989).

3.2.1 Perceived Ease of Use (PEoU)

According to several academics, ease of use means the degree to which a person anticipates finding it simple to utilize a particular information system (Davis 1989). Indeed, with this in mind, organizational performance can be enhanced with the help of easy and effortless processes (Brock and Khan 2017). Davis (1989, n.p.) has also defined Perceived Ease of Use as “the degree of ease involved when using an information system.” Likewise, Soon et al. (2016) explained that ease in utilizing the new technology will pave the ways for users to accept this; using an information system free of effort will increase the organizational performance (Brock and Khan 2017; Venkatesh and Davis 2000). Shin and Bohlin (2020) detailed that BDA may produce benefits for various types of organizations, such as cutting costs, managing and controlling risk factors, and facilitating the process of informed decisions.

H1:

Perceived ease of use (PEoU) has a positive impact on perceived usefulness (PU) (Davis 1989).

H2:

There is a positive impact of Perceived Ease of Use (PEoU) on Behavioral Intention (BI).

3.2.2 Perceived Usefulness (PU)

PU is one of the outside forces affecting TAM. PU measures how strongly a person thinks that utilizing modern technology would enhance their ability to accomplish their work (Al-Gahtani 2001; Davis 1993). Wu et al. (2017) stated that PU is a basic motivator and the most frequently used variable for technology adoption. Similarly, PU can be explained as whether a person considers that by utilizing the modern system their job performance will be enhanced (Brock and Khan 2017). According to the reviewed literature, there is a positive significant impact of PU on BI, BDA skills, and current practices (Ambak et al. 2016; Soon, Lee, and Boursier 2016; Wu and Chen 2017). Similarly, the PU influences the skills and work experience in an accounting virtual learning environment (Herrador-Alcaide and Hernández-Solís 2017). Based on the reviewed literature the following hypotheses have been developed:

H3A:

There is a significant positive impact of PU on BI.

H3B:

PU has a positive impact on BDA skills.

H3C:

There is a positive relationship between PU and current practices of BDA.

3.2.3 Behavioral Intention (BI)

The amount to which a person has made conscious plans to conduct or not perform some defined future behavior is referred to as behavioral intention (BI) (Budu et al. 2018). Esteves and Curto (2013) explained that BI intends some definite future behavior. The reviewed literature has produced evidence of a positive relationship between PU and BI which has been measured in different fields (Ambak et al. 2016; Wu and Chen 2017).

H4:

There is a positive relationship between BI and BDA skills.

4 Research Method

The current study follows the Positivist research paradigm, which employs a quantitative research design. In Pakistan, there were 369 medical institutes and teaching hospitals (PMDC 2019), with data obtained from these healthcare institutes’ library staff using a data gathering tool. The questionnaire was created using current literature in the relevant fields, which was sent to two BDA specialists for validation. The questionnaire’s validity and reliability were further tested using the Smart Partial Least Square (PLS) measurement model, which demonstrated high reliability and validity by fulfilling the essential threshold. The questionnaire used five Likert scales, with the items for the measured variables adapted from the pertinent literature. In the first case, a single phase quantitative research strategy was employed to collect data from respondents, namely library professionals from medical institutes and affiliated teaching hospitals, using a questionnaire. The data gathering tool was made available to responders through social media, personal visits, posts, and Google Docs. Furthermore, the researcher expanded the facility to clear the respondents’ ambiguity via cell phones discovered when filling out the questionnaires. The country had 369 medical institutes and teaching hospitals, and the respondents provided 256 responses out of the entire population. The questionnaire survey data was analyzed using the SPSS twenty-sixth version and Partial Least Squares (PLS-4). For hypothesis validation and data analysis, PLS software was employed.

5 Results

6 Discussions

The importance of Big Data analytics implementation in healthcare units and medical institutes’ libraries are increasing day by day (Ambigavathi and Sridharan 2018). Compared to the role of the past librarians, current library professionals have many challenges to face. Now they are working in the digital era, they are supposed to act like data scientists, data curators, and digital service providers (Ahmad et al. 2019). The results showed excellent validity and reliability of the instrument; both the convergent and discriminant validity were tested and proved excellent. Tables 2, 3, and 4 provide the values of the validity and reliability of the tool, which meet the required criterion in all aspects (Fornell and Larcker 1981). The results showed that most of the medical libraries and librarians performed BDA practices but in a disorganized form. The core reasons for these jumbled practices included lack of knowledge and skills, non-availability of modern equipment, and administrative issues. There were total of six main latent variables, namely, PEoU, PU, BI, Skills, Current Practices, and Performance/Actual Use. Table 2 shows all the measured/latent variables of the study in hand, with TAM used as the base for the current study (Venkatesh and Davis 2000). The same was modified and some new variables were added to the model with the help of literature in the field of BDA as given in Figure 2. The researchers also went through the structure model of the SEM to examine the impact and relationship of a number of variables, details of which are given in Table 5. For said purpose, eight hypotheses were developed and tested, with the results revealing a positive impact of PEoU on PU, which has a similar result of the study conducted by Nnaji et al. (2020) on adoption of modern technology. This indicates that medical librarians perceived the use of BDA to be an easy task and helpful in enhancing the standard of work at their libraries. The policymakers of health departments and administration of these medical institutes must encourage these medical librarians and extend all facilities in this regard. PEoU has less impact on the BI as the P-value of said hypothesis does not meet the required criterion (Anna and Mannan 2020) as shown in Table 5. It indicates that the perception of medical library practitioners about the ease of use does not affect their behavioral intention towards the implementation of BDA in these medical libraries. The result of the study conducted on the adoption of Wearable Sensing Devices (WSDs) is adverse to the current study where the PEoU has a significantly positive effect on BI (Nnaji et al. 2020). Furthermore, the structure model also indicated that PU has a positive impact on BI as its P and T values meet the threshold values. In similar studies conducted by Davis (1989), Soon et al. (2016), and Cheng (2019), the same results were observed in the adoption and implementation of modern technology. These medical institutes and teaching hospitals should sensitize their medical librarians regarding the benefits of these modern technologies to ensure that they will adopt this modern information system. Similarly, PU influenced the skills of these medical library professionals as proved by the structural model of SEM by meeting the required criterion. This demonstrated that PU and BDA skills are correlated and the skills of medical library practitioners are a strong predictor for the adoption of BDA in these medical institutes. The same relationship was reported in yet another study where the PU has influenced both the skills and work experience (Herrador-Alcaide and Hernández-Solís 2017). PU has a significantly positive relationship with CP, with the P and T values (P = 0.000, T = 3.345) of the structural model showing the result which met the required criterion. Another study, conducted on PU, PEoU, and User Acceptance of IT, had the same result (Davis 1989). Moreover, this indicates that current practices are performing the vital role to compel these medical librarians to adopt this new information technology. In the same vein, the BI has a strong positive impact on skills (P = 000.0, T = 5.833). The result indicated that the BI is a strong predictor of BDA skills as compared to PU. It is obvious from the analysis that the intention to adopt BDA in these institutes has a strong correlation with the BDA skills of these medical library practitioners. The same impact can be found in yet another study where a strong positive impact of BI on skills was found (Adetoro et al. 2017). Furthermore, the analysis also showed that BDA Skills have a positive impact on current practices (PU) (P = 0.000, T = 4.405). It can be elaborated that the BDA skills of these medical librarians is a strong predictor for the implementation and practices of BDA in these information centers. The policymakers and high level individuals of these institutions should arrange training workshops and seminars for their library professionals to enhance their BDA skills which will ultimately lead to the smooth implementation of BDA in these medical institutes. The findings of the current study are in line with a study conducted by Richards (2017) where a positive impact of skills on current practices was found. The study in hand also illustrated that there is a strong positive impact of current practices on performance/actual use. According to the coefficient analysis with P = 0.000 and T = 18.343 as given in Table 5, the current practice (CP) is a strong predictor for the actual use of a modern information system. The administration of these medical schools should focus on present BDA practices and provide facilities to increase the standard of these practices, as these will help prepare the way for subsequent BDA technology implementation in these institutes’ and teaching hospitals’ libraries (Figures 3 and 4).

Figure 2:

Proposed model.

Figure 3:

Measurement model.

Figure 4:

Structural model.

7 Conclusion

BDA’s importance has been recognized in many areas of life, including commerce, companies, business, government organizations, healthcare units, libraries, and information centers. Based on the literature review and data analysis, it is clear that BDA implementation and practices in medical and teaching hospital libraries are urgently needed. The health department and administration of medical institutes and teaching hospitals should play a critical role in providing all types of infrastructure and BDA technology training, as well as removing all barriers to the smooth implementation of this new information technology. The current study deals with an emerging issue and covers a research gap in the field of modern technology implementation and practices in the form of BDA at the country’s medical institutes and teaching hospitals. This research will help to raise awareness and create a roadmap for medical librarians and policymakers to utilize this technology in their respective medical libraries. Furthermore, the current study has offered a full authentic model for examining people’s intentions towards the implementation of BDA technologies in many domains. Overall, this study adds significantly to existing research on BDA implementation and practices. It relates to the current condition of BDA implementation and practice in medical libraries, as well as the loopholes and obstacles to the smooth application of BDA. BDA practices in these medical facilities can increase the utility of library services, resulting in the best services being provided to lifesavers and other library users. The use of SEM-PLS and hypothesis testing showed that the libraries of medical institutes and affiliated teaching hospitals in Pakistan provide a conducive environment for the adoption and use of BDA technologies. Additionally, the library staff at these libraries have highly good behavioral intentions with regard to the deployment of this information system. In a nutshell, the execution of this new information system in these medical libraries requires the help and cooperation of top management and high-level individuals of these institutions.