How can big data analytics be used for healthcare organization management? Literary framework and future research from a systematic review | BMC Health Services Research

According to the SLR, the iterative process shown in the Fig. 1, has allowed to delete the duplicates and match the results with the RQs.

As shown in Fig. 1 the initial search on Scopus database has delivered 227 results. By limiting research to papers published between 2016 and 2021, 11% of records have been removed. At the second stage, by selecting the subject areas, the screening has allowed to exclude 131 records; thus, the 57.7% of the results initially selected. The last step of the process has conducted to exclude document types such as Review, Book, Conference Review, Letter, and Note. In other words, 37 records were excluded, representing 16.3% of the sample. At the end of the screening process, 34 articles were selected, representing about 15% of the sample.

In the descriptive analysis the time distribution of the studies from 2016 to 2021 is included. It is important to note the increasing of publication trend from 2017 to 2019. This output confirms a growing interest in the research field of BDA applied to healthcare organizations (Fig. 2).

The trend of research steams considers a sample of 34 scientific contributions as they come from the screening process above described. Although 6% of the total sample was collected in the years 2016 and 2017, it is only indicative of the growing trend of scientific studies on BDA in healthcare sector. The overall incidence in 2018 was 12% but the turning point was reached in 2019 as 32% of the studies collected in the sample were reached. This outcome could be read considering the Covid-19 pandemic outbreak which has been a representative testing ground for BDA tools by helping managers and decision-makers to plan healthcare managerial strategies.

In this context, the use of the BDA by Chinese healthcare organizations for tracking people’s flow during the lockdown, represents an important case study that has registered the peak in the time flow of research. By looking at 2020 and 2021 data, which represent respectively 24% and 21% of the total scientific contributions, the growing trend seems to be confirmed by validating the rising interest in BDA research seen as a planning tool for healthcare processes.

The pie-chart shows the scientific production by country. It is necessary to specify that Scopus database clusters the studies by home country author’s organization, therefore the same study could be referred to more than one country and thus belong to more than one cluster.

The geographical locations of the studies showed in the Fig. 3 outlining India, UK, and USA as more than one third of the total scientific producers. It is well known that IT companies as Google, Apple, Amazon, and Microsoft are investing considerable resources on BDA tools for healthcare. China and India contribute together with 22% of the scientific articles. Big data technology has played a key role in virus tracking during the pandemic crisis. The “Internet Plus Healthcare”, a big data center in Zhongwei (China), provides cloud services to both healthcare institutions and IT companies. In Yinchuan (China), an industrial park for big data acts as a catalyst for IT company involved in healthcare sector. India confirms to be one of the heavily adopter countries of artificial intelligence, big data analytics, and IoT technologies. Although India must face the challenge to provide basic healthcare services in a predominantly rural country, start-ups with BDA skills in healthcare are springing up.

Geographical locations of the studies

It is also important underlining the performance of the European countries. UK, Greece, Italy, Spain, Germany, and Portugal support the research with almost 40% of the studies published, confirming that Europe will be a driving force for the BDA research in the next future. The development of a European Health Data Space (EHDS) is an ambitious project of the European Commission. It will lead member states to share an efficient infrastructure for both exchange and management health data by providing citizens with equal treatment, free access to clinical data, and quality healthcare services.

In the area “Others” all the other countries contributing marginally to research have been included.

The next step of the study is focused on a content analysis to show the experiences of applying BDA in healthcare organizations.

Starting from the 34 articles selected for the descriptive analysis, to identify in detail the core issue of the study, a second screening was performed. 18 articles were excluded because weakly focused on the research objective which concerns specifically how BDA can be used for healthcare organization management. Thus, after an in-depth reading of abstracts and full papers, the scholars have identified 16 papers closer targeted on the mentioned research objective. The 16 studies selected through a content analysis were clustered into 4 research areas (RAs) as showed in the following table (Table 1). The clustering procedure identifies 4 relevant topics: Potentialities of BDA (RA1), Resource management (RA2), BDA and management of health surveillance system (RA3), BDA technology for healthcare organization (RA4). The proposed clustering has been though to give an easy-to-go research map and to support the healthcare managers.

RA1: potentialities of BDA

Wang and Hajli [16] define BDA potentialities in the healthcare context as “the ability to acquire, store, process and analyze large amounts of health data in various forms, and deliver meaningful information to users, which allows them to discover business values and insights in a timely fashion”. The relationship between BDA and the benefits for the healthcare organizations it has been well expressed by the theory of the “path to value chain” [16]. This path represents an important contribution to the exploration of business value, not only for drawing the generic and well-established connection between big data capabilities [19] and the benefits, but also for empirically showing how capabilities can be developed and what benefits can be achieved in the healthcare organizations. Another study included in this area, explores the key role of BDA capabilities in developing healthcare supply chain integrations and its impact on hospital flexibility [27]. Specifically, the BDA has a fundamental role in developing healthcare integration supply chain and the operational flexibility. Considering the health and economic crises caused by the Covid-19, this dimension of BDA has been an especially important leverage for managers to improve operational flexibility of the healthcare organizations. The ability to provide predictive models and real-time insights, is a powerful prospective of the BDA for helping healthcare professionals and managers in decision-making process. In this regard, the literature presents several applications of big data in healthcare that support the data collection, management, and integration of data in healthcare organizations [37]. Moreover, BDA enables the integration of massive datasets, supporting decisions of manager and monitoring the managerial aspects of healthcare organizations. Building a decision-making process based on BDA, firstly means identifying the big data keys that can implement ad-hoc strategies to improve efficiency along the healthcare value chain. To this end, the research carried out by Sousa et al., [37] underlines the benefits that BDA can give to the decision-making process, through predictive models and real-time analytics, assisting in the collection, management, and integration of data in healthcare organizations.

To date, thanks to an integrated and interconnected ecosystem, is becoming possible to provide personalized healthcare services, collect an enormous quantity of both clinical and biometrics data and, thus, implement BDA instruments. Nevertheless, to take a real advantage from these tools and turn them into useful decision support systems (DSS), is necessary for R&D to be focused on data filtering mechanisms in order to obtain good-quality reliable information [38]. The healthcare models based on BDA and implementation of new healthcare programs, enable both medical and managerial decision support for the healthcare services provision. New types of interactions with and among users of the healthcare ecosystem will produce in the next future a wide variety of complex data, thus, the main challenges refer to information processing and analytics.

In light of the above, the RA1 includes studies for which the quality of data and the need for high performance filtering mechanisms are becoming keys factor for the success of BDA-based management systems in the healthcare organizations. For example, the study carried out by Maglaveras et al., [38], included in this area, explores new R&D pathways in biomedical information processing and management, as well as to the design of new intelligent decision support systems.

RA2: resource management

Another important research direction emerged from the literature review, concerns positive impact of the BDA on the resource management. Insufficient policy for managing medical materials waste, energy use and environmental burden, restricts the resources conservation. The BDA is extremely useful in this aspect; it could provide in the next future an important contribution to implement the circular economy processes and to support sustainable development initiatives in the healthcare organizations [39]. To this end, the study developed by Kazançoğlu et al. [39], underline the importance of circularity and sustainability concepts to mitigate the sector’s negative impacts on the environment. Furthermore, the study identifies the barriers related to circular economy in the healthcare organization and provides solutions to these barriers by implementing BDA-based management systems. Lastly, the authors, have developed a managerial, policy and theoretical framework to support healthcare managers to launch sustainable initiatives in the context of healthcare organization.

The impact on the performance has been also investigated by studies that have linked benefits of BDA and artificial intelligence with green supply chain integration process [40]. Digital learning is more becoming a “moderator” of the green supply chain process with a significant positive impact on environmental performance of the healthcare organization. BDA-AI technologies will lead to improvement of the environmental process integration and green supply chain collaboration and, consequently, will support the managers’ decisions involved in the supply processes. This study also provides an important reference framework for logistics/supply chain managers who want to implement BDA-AI technologies for supporting green supply processes and enhancing environmental performance of the healthcare organization [40].

Nowadays, many scholars are focusing on BDA-driven decision support systems to sustain the healthcare managers [41]. These types of BDA-based analytical tools will provide a useful quantitative support for managers of healthcare organizations. The authors have reported design and technical details of the system implementations using case studies. They have developed a toolkit which represents a framework reference for resources management, allowing to create strategic models and obtain analytical results for evidence-based decisions and managerial evaluations.

In this RA, two other important topics investigated by BDA are: high quality healthcare service, and healthcare costs. Optimize the supply chain activities is an imperative to keep lower the healthcare costs. The data generated by medical equipment and devices can be successfully used in forecasting, decision-making process, and to make more efficient the healthcare supply chain management [42]. The study carried out by Alotaibi et al. [42], thus, presents a review on the use of big data in healthcare organizations underling opportunities and challenges deriving from the application of BDA-based management systems within the organizations.

As already asserted, a good implementation of BDA in the healthcare organization will play a fundamental role in improving the clinical outcomes management, giving helpful insights for decision makers and managers, in order to avoiding diseases, reducing healthcare expenses, and improving the performance of the healthcare organization [43]. However, to achieve these ambitious outcomes the research will face a crucial challenge: how to rationalize, make easily usable, and at affordable costs, heterogeneous data coming from diverse sources. The research developed by Kundella and Gobinath [43] represents an important contribute to explore key challenges, techniques, technologies, privacy issues, security algorithms and future directions of the use of BDA in the healthcare organization.

RA3: BDA and management of health surveillance system

The rise of BDA promises to solve many healthcare challenges in the developing countries. The BDA applied to healthcare organization help managers to rationalize the resources, and health system to better delivery treatments to the patients [44]. In this regard, the government of Zambia is thinking to implement BDA solutions to provide more effective and efficient healthcare services. A well-managed health surveillance system represents an important driver to improve the quality of life and reduce the medical waste, especially in developing countries where the lack of resources is severe and limits economic development. For all these reasons, Europe is investing on BDA initiatives in public health and in the oncology sectors, to generate new knowledge, improve clinical care and make more efficient the management of the public health surveillance system [45]. The BDA capability for identifying specific population pattern, managing high volume of data and turn it into real (or near real) time insights, contributes to identify it as a powerful tool to support the managers for the decision-making processes. Despite this, implementing a BDA-based management systems within the healthcare organizations requires investment in the human capital, strong collaboration with stakeholders, and data integration with and among the healthcare units. To this end, Gunapal et al., [46] has highlighted that Singapore has setup a Regional Health System (RHS) database to facilitate BDA for proactive population health management (PHM) and health services research [46]. The structure of the healthcare database has been built collecting data from four database coming from three RHSs: National Healthcare Group (NHG), Tan Tock Seng Hospital (TTSH), National University Hospital (NUH) and Alexandra Hospital (AH). The result has been a database including information useful for the healthcare managers which incorporates data on patient demographics, chronic disease, and healthcare utilization information. These characteristics facilitate the identification of specific patients’ paths linked by past healthcare utilization and chronic disease information. Converging information into a single database helps to understand the cross-utilization of healthcare services across the three RHSs. A such approach allows to setup the RHSs structure for initiative-taking population health management (PHM) and to improve the performance of healthcare organizations [46].

RA 4: BDA technology for healthcare organization

The wearable devices and different kind of sensors, able to collect clinical data, in combination with BDA, will constitute the basis of personalized medicine and will be crucial tools to improve the performance of healthcare organizations [47]. The scientific research has to face the important challenge to adapt data acquisition, storage, transmission and analytics to healthcare demand. Thus, the healthcare data should be categorized, homogenized, and implemented into specific models by adapting machine-learning techniques to the nature of the healthcare organization.

A fruitful field of interest for the application of BDA in healthcare organization is the diagnostic imaging. To take out maximum benefits from it and to be useful for managers of healthcare organizations, it is necessary to implement digital platforms and applications [48]. Indeed, the simple production of a large amount of data does not automatically translate to an advantage for the healthcare performance. Specific applications are required to favor the correct and advantageous management of diagnostic images [48]. The link between BDA and IoT technologies, as instrument to incorporate the accessibility, capacity to customize, and practical conveyance of clinical data, emerged as another research direction investigated by the papers included in this RA. These tools allow: (1) the healthcare organizations to decrease expenses; (2) the people to self regulates treatments; (3) practitioners to take as quickly as possible decisions in remote way and keep constant contact with patients [49].

In light of these results, it is possible to state that IoT, big data, and artificial intelligence as machine-learning algorithms, are three of the most significative innovations in the healthcare organization. These types of organizations are implementing home-centric data collection networks and intelligent BDA systems based on machine learning technologies. For example, a high-level implementation of these systems has been efficiently implemented in Cartagena, Colombia, for hypertensive patients by using an e-Health sensor and Amazon Web Services components [50]. The authors stress the importance of using the combination of IoT, big data, and artificial intelligence as tools to obtain better health outcomes for the communities and improved performance for healthcare organization. The new generation of machine-learning algorithms can use standardized data sets generated by these sources to improve the effectiveness of public health interventions [50]. To this end, as pointed out by numerous studies in the field of BDA applied on healthcare organizations, it becomes crucial for the next future research to concentrate R&D efforts towards full standardized dataset protocols.