Keywords

1 Introduction

Based on data from the University of South California, by 2025, minimum 465 data points per person will be created daily. The total capacity of world data currently is about 45 zettabytes (ZB), 41 times more than crystals of salt in the oceans. Furthermore, while not all newly generted data will necessitate long-term storage, IT specialists forecast that by 2025, about 7.6 ZB will be required, compared to about 1.2 ZB in 2020. The progress of patient and healthy populations medical and health data is linear, and data like imaging and electronic medical records (EMR) have a specific growth pattern, generating challenges and opportunities for healthcare productivity. This appraisal of almost 80 megabytes per year per patient mirrors the increasing volume and complexity of the long-term storage of health information (Bakos et al. 2018; Chodacki et al. 2016; Morgan and Janke 2017). Several vital mechanisms are obvious when inspecting the wide-ranging structure for the long-term digital storage and consumption of medical research data. These important elements comprise the Data Management Plan (DMP), the Standards for Digital Imaging in Medicine, SNOMED CT, ICD-11, Health Level Seven (HL7), and usage of Artificial Intelligence (AI) technologies (Simms and Jones 2017). At the same time, professionals face a considerable volume of patient data demanding trustworthy storage infrastructure. Thus, healthcare organizations worldwide must invest in scalable and safe storage solutions to adapt to the growing data load, and archiving of outdated electronic health records (HER). Compliance with data storage and regulatory requirements is essential (Simms et al. 2017; Williams et al. 2017). For example, integrating and understanding data from outdated EHR systems can take time and effort. If funding is secured for data storage infrastructure in many countries, the issues of effectively extracting relevant information from large data sets still requires attention.

Data protection issues are at the center of all the aspects mentioned above. With the growth of patients’ data volume, maintaining confidentiality and the safety of medical information becomes increasingly important. Health organizations should introduce reliable cybersecurity measures to protect sensitive data on patients from unauthorized access and violations (Stephens et al. 2015). This policy is facilitated by the commitment to data protection rules (for example, HIPAA in the United States). Suppliers of medical services should constantly update their security protocols and comply with developing standards.

Given the epidemiological challenges during the COVID-19 pandemic, actively archiving data solutions such as tracking disease outbreaks, getting up-to-date information on treatments and vaccines, tracking patient diagnoses, and supporting the growth of telemedicine, the role of health data has never been more significant. The active archiving solution aligns with the evolving expectations in healthcare IT by ensuring secure, accessible, and compliant storage of historical patient records. This archiving method supports patient care continuity and enhances data management efficiency within healthcare organizations (Fig. 1).

Fig. 1
A flow diagram. Producer leads to ingest via S I P, leading to data management and archival storage via A I P. This setup integrates with consumer queries, responses, orders, and D I P. Producer leads to preservation planning and administration management, which also integrates with the consumer.

A model for data management effectiveness in healthcare organizations

Full size image

Long-term digital storage and usage of research data, especially in data pooling, involves several considerations to ensure the integrity, accessibility, and usability of the data over an extended period (Michener 2015; Schnell 2015; National Library of Medicine 2013). The key aspect is scalability and performance. Databanks should be customized to handle large volumes of data and provide high performance. Scalability allows it to grow effectively with increasing data volume and number of users. Implementing a scalable platform for processing and managing biomedical data will help make research more efficient while ensuring data security and accessibility for many users.

One example is a real-time medical data collection system for medical laboratories (using a laboratory information management system), which reduces errors, minimizes extra work, and ensures data and metadata integrity. Another notable example would be Submission Information Packages (SIPs) for clinical trials, patient registries, and other human subjects’ studies, where using Common Data Elements (CDEs) is a strategic approach. Developing SIP for clinical research and patient registers, including CDE, can significantly improve the standardization and harmonization of data, which, in the future, will enhance the quality of data and promote cooperation and joint use of research initiatives (Williams et al. 2017; O’Reilly 2018; Kazic 2015).

CDEs are standardized terms, definitions, and guidelines for data collection in clinical trials. They aim to ensure uniformity and consistency in data collection and presentation, facilitating comparison and sharing of data between different studies and organizations. Examples of CDEs used in the clinical and biomedical fields include Big Data biobanks and repositories. For instance, CDEs can define standards for describing genetic variants, sequencing methods, and other parameters, as well as multicentred studies, which help ensure consistency of data collected across clinical sites, making it easier to analyze and compare results available through the National Institutes of Health (NIH) CDE resource portal (Rubinstein and McInnes 2015).

Applying CDE in such contexts provides a standardized and unified approach to data collection, improving research findings’ accuracy, consistency, and transferability. It also enhances the efficiency of research processes and facilitates data sharing between research groups and organizations. Furthermore, using online collaboration tools to map reported terms to a preferred ontology is common in biomedical research and other fields. Such tools enable effective collaboration between researchers and help ensure consistency and uniformity in the use of terminology (Williams et al. 2017; Simms and Jones 2017). Ontologies establish standard terminology and definitions for use in a given field (International Classification of Diseases), which ensures uniformity in understanding and communication. Ontologies are widely used in AI systems in healthcare to train machines, support decision-making in developing new drugs, and search for target cells.

Implementing the biomedical dataset model provides new data integration, analysis, and discovery capabilities, which help advance fundamental, translational, and clinical research. Assessment of data resources needs and ensuring their protection is essential in defining data characteristics, storage requirements, and security measures to protect confidentiality and data integrity. Effective data management includes understanding the data lifecycle, from collection and processing to storage and long-term access. The Open Information Archive model provides a structured approach to data management and long-term preservation. Application of this model promotes the conservation of the authenticity, accuracy, and reliability of biomedical data, providing standards for long-term access (Simms et al. 2017; Ravagli et al. 2017; Leonelli 2017; Navale et al. 2018).

In summary, effective data management is becoming a key element in contemporary science, where large volumes of data require careful planning and management. Below we discuss the key parameters that determine the effective operation and development of long-term digital storage and research data usage.

2 Data Management Plan (DMP)

The first step is a formal document, the data management plan (DMP), which outlines the strategies and procedures for managing research data throughout its lifecycle. It serves as a roadmap for researchers and research teams, providing a structured framework for organizing, storing, documenting, and sharing data. The primary goals of a DMP include ensuring the integrity, accessibility, and long-term preservation of research data, as well as facilitating compliance with ethical, legal, and institutional requirements (Kirlew 2017).

Key components typically included in a DMP are data architecture, data models, data generated, data quality checks, data governance, project overview, data types and formats, data collection and processing methods, ethical and legal considerations, data ownership and responsibility, data storage and backup, data security, data sharing, and access, data preservation.

Leading funding and research agencies, including the National Institutes of Health (NIH), the National Science Foundation (NSF), the Centres for Disease Control and Prevention (CDC), and the Agency for Toxic Substances and Disease Registry. (ATSDR) require medical researchers to submit a DMP for funding decisions. The DMP assures sponsors that data loss prevention strategies, regular backups, protection against losses due to hardware failures, and other precautions are in place (Data Storage Best Practices 2018). The DMP facilitates the planning and standardization of metadata, improving data quality, making it easier to interpret and compare, and ensuring better reproducibility. DMP development includes consideration of long-term data storage and access issues, which is essential to ensure data safety and future use.

A typical example of a DMP is a text document written as a detailed narrative. These documents include various guidance documents and templates for DMP production (NNLM National Network of Libraries of Medicine), and the first generation of tools to facilitate DMP production have been created and are widely available (Goodman et al. 2014; The NNLM website 2020; Navale and Bourne 2018). Examples of the first generation of DMP tools could be simple online forms offered by universities or funding agencies to help researchers develop data management plans for their projects. Later generations of tools have become more complex, integrated, and focused on advanced data management capabilities.

The second generation of DMP tools aimed to make them “machine-accessible” or “machine-readable.” These tools focused on automating data management processes and integrating with information systems, contributing to more efficient data processing and use. The tools used semantic technologies and data formats such as RDF (Resource Description Framework) to present information in machine-readable form. Integration with other information systems, such as data repositories, library catalogs, or project management systems, was done (Ohno-Machado et al. 2017). Standardized protocols like API (Application Programming Interface) enabled interoperability with other applications and systems. As data management requirements and standards have evolved, new DMP tools have begun to include more detailed and in-depth information.

Moreover, the DMP plays a vital role in the grant application evaluation process and post-award evaluations, although this role may need to be better defined and understood. However, not all guidelines for scoring Research Proposals (RPPs) explicitly mention DMPs, although data-sharing plans can be a required element of proposals. The reasons may be varied, and this may depend on the country, organization, or specific area of study. Understanding the importance of DMPs and their inclusion in the NCD assessment process is just beginning to gain traction, and in the future, such documents may more clearly highlight the role of DMPs (Corpas et al. 2018; Jagodnik et al. 2017).

In terms of specific examples, the Interdisciplinary Earth Data Alliance (IEDA) is an organization that provides infrastructure for storing, processing, and sharing geoscience data. They support tools and resources for researchers to ensure accessibility and management of data in the field. IEDA provides researchers with the ability to generate a data compliance report based on an NSF (National Science Foundation) award number; this is likely because NSF often sets data management standards and requirements for projects they fund, as well as part of the “Results” preliminary support for NSF” in subsequent proposals, but, again, it is unclear how much weight they are given in the evaluation process.

In another example, Canada’s Tri-Agency Statement of Principles for Digital Data Governance emphasizes researchers’ obligations in developing and adhering to DMPs. However, the Canadian Institutes of Health Research needs to include DMPs in its evaluation criteria. However, given the growing significance of effective data administration in scientific research, organizations may want to review their policies and criteria in line with evolving standards and practices (Rubinstein and McInnes 2015).

The development of second generation DMP tools in response to the changing requirements of funding agencies and the generalized learnings from the first generation of tools represents a logical development in the field of data management in scientific research. Creating a “meta-DMP,” or tool that provides consistent guidance irrespective of an agency’s specific reporting requirements, has several potential benefits: universality, flexibility, automation, compliance with updates, training and support, integration, and performance tracking.

For example, maDMP, according to Simms and Jones 2017, can help predict data storage costs (Williams et al. 2017; Simms and Jones 2017). The proposed formal machine-readable document allows data exchange between different objects through the entire data life cycle. maDMP’s emphasis on metadata, such as quantity and type of data, regardless of storage location, allows for evaluating the time-varying cost of storing such data. A standard has yet to emerge, although several use cases exist.

Creating such a tool could address the challenges posed by the diversity of requirements from different agencies and make it easier for scientists to develop and manage data management plans.

3 Data Standards and Documentation

Data standards and documentation are crucial for ensuring accurate, consistent, and interoperable healthcare information in the medical and hospital sector. Health Level Seven (HL7): HL7 is a widely used international standard for exchanging, integrating, sharing, and retrieving electronic health information. It defines a framework and common standards for messaging, clinical documents, and interoperability. HL7 develops standards for exchanging information in various areas of health care, including clinical, administrative, and financial aspects. These standards are essential in supporting interoperability between different information systems, ensuring a normalized and structured exchange of information.

Examples of standards developed by HL7 are:

  1. 1.

    HL7 v2 (Health Level Seven Version 2) is a standard for healthcare messaging. It transfers data between different systems, such as electronic medical records (EMR) systems and patient management systems.

  2. 2.

    HL7 CDA (Clinical Document Architecture). A standard for structuring clinical documents such as case reports and patient histories to ensure standardized exchange.

  3. 3.

    HL7 FHIR еhe standard is focused on providing a faster and more flexible exchange of information in healthcare, especially in web and mobile applications.

  4. 4.

    HL7 v3 (Health Level Seven Version 3) is designed to solve data interoperability problems and define a standard healthcare model.

International Classification of Diseases (ICD): The ICD is a standard system for classifying diseases, conditions, and health-related problems. It provides a common language for global reporting and monitoring health conditions. It is important to consider the International Classification of Diseases (ICD) as an example of standardization of data and documents as part of the long-term digital storage and usage of research data systems. It is a standard developed by the WHO to classify and code different diseases and health conditions. It is used worldwide for uniform documentation of diseases, health statistics, and medical and health information exchange. The latest version of ICD-11 was adopted by the 72nd World Health Assembly in 2019 and entered into force on January 1, 2022 (Annex 3.8 of the Reference 2019).

The International Classification of Diseases, Eleventh Revision (ICD-11), is an updated classification system covering various aspects of diseases, including their diagnosis, treatment, research, and statistics. The ICD-11 classification deals with various aspects, such as using research data systems for long-term digital storage. Such elements may include coding diseases in research data to uniquely identify diseases in long-term data storage and use systems, providing a uniform and standardized way to represent medical concepts.

Systems using ICD-11 can monitor morbidity, mortality, and other aspects of population health over the long term. Classification allows the creation of standardized reports and analysis of trends. The use of ICD-11 can help with this by unifying the way diseases are classified. In systems for long-term storage and use of research data related to medical research, ICD-11 can serve as a basis for structuring and analyzing data related to various diseases (ICD-11 2022).

The Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) is a comprehensive, multilingual clinical terminology used in healthcare and clinical research. It is a standardized coding system for representing and exchanging clinical information globally. SNOMED CT facilitates precise and standardized health information exchange across healthcare settings and systems (Vuokko et al. 2023; Cangioli et al. 2023).

SNOMED CT and ICD are two different standards addressing different aspects of healthcare. Still, they can interact and complement each other across multiple health information systems. SNOMED CT is more detailed and designed to describe clinical concepts, while the ICD is more often used for statistics, classification, and coding underlying diseases.

The primary goal of SNOMED CT is to provide standardized codes to describe diseases, procedures, symptoms, and other clinical concepts. SNOMED CT incorporates the Fully Qualified Ingredient (FQI) concept to characterize drug products with great detail. SNOMED CT provides standardized terms for clinical decision support, semantic search, and data analytics in healthcare. Concepts are organized into hierarchies, allowing for a more granular representation of clinical information. The hierarchy enables the classification of concepts based on broader or narrower relationships.

Each concept in SNOMED CT is associated with one or more human-readable terms or clinical descriptions. These descriptions represent the concept in an understandable way to healthcare professionals. SNOMED CT allows for creating post-coordinated expressions, combining multiple concepts to represent complex clinical situations. This feature enhances the specificity of clinical coding. SNOMED CT is designed to support multiple languages, making it a versatile terminology for international use. SNOMED CT covers various clinical domains, including anatomy, clinical findings, and procedures. This comprehensive coverage makes it suitable for representing diverse aspects of healthcare. SNOMED International maintains SNOMED CT, a not-for-profit organization that oversees its development, distribution, and ongoing updates. It is widely used in electronic health records (EHRs), health information exchange, clinical research, and other healthcare-related applications to ensure standardized and interoperable representation of clinical information. SNOMED CT promotes unambiguity and standardization in medical terminology, enabling more efficient data exchange, system interoperability, and semantic accuracy in healthcare (ICD-11 2022; Vuokko et al. 2023; Cangioli et al. 2023; HIMSS Adoption Model for Analytics Maturity (AMAM) 2023; Cheemalapati et al. 2016).

Additionally, using standardized methods and formats reduces the risk of data loss due to technology obsolescence. Long-term storage of data by standards ensures data stability and reproducibility. Standardized systems are more flexible and can be easily updated or expanded, allowing to adapt to new requirements, technologies, and standards without significant costs. The overall standardization of data and documents creates a sustainable basis for the long-term storage and use of research data, ensuring its integrity, availability, and relevance over an extended period.

4 Clinical Document Architecture (CDA)

Developed by the Health Level Seven (HL7) initiative, Clinical Document Architecture (CDA) provides a standardized format for storing and sharing clinical information. Over 20 long-term periods of storage and use of research data, the CDA architecture offers a structured form for presenting medical information. CDA allows for storing data in organized ways that influence subsequent data access, retrieval, and analysis.

  • Every CDA document begins with a Clinical Document, the root element of the entire document.

  • The document header contains metadata such as document ID, document type, creation date, patient ID, and other attributes.

  • The CDA is divided into sections, each containing specific clinical information. For example, a section with medical history and laboratory results may exist.

  • There is structured clinical information in the document’s central part. This section contains structured data, such as tables, lists, and other elements in a specific format.

  • A record is a specific piece of information within a section. It contains primary data such as test results, diagnoses, procedures, and other elements.

  • The document’s body may also contain textual information and unstructured content necessary for additional comments or descriptions.

The CDA architecture includes standardized semantic elements such as LOINC, SNOMED CT, and others, which provide a more precise and unambiguous understanding of document content. Semantic elements are essential for the accuracy and consistency of data during long-term storage (Hart et al. 2016; Ghatnekar et al. 2021; Blackley et al. 2019). CDA can embed contextual information such as patient IDs, healthcare facility information, timestamps, and other details. Contextual information ensures that data is complete and correct for future compliance. CDA is often used in patients’ electronic records to present clinical information, ensuring medical research continuity in digital medical institutions.

The structured CDA format can store additional metadata and tags that periodically identify research data, which is essential for subsequent analysis, meta-analysis, and re-use of data. CDA can be included in a strategy for secure data storage, ensuring the integrity and confidentiality of health information throughout its lifespan. CDA enables interoperability between different areas of health and healthcare, which is essential for the exchange and sharing of research data on a large scale.

The application of CDA in long-term storage and use of research data provides semantic standards of clarity and capability with existing progressive information trends. Being a document markup standard and defining the structure and semantics of “clinical documents,” it has defining characteristics: persistence, control, auto-authentication, and integrity. The clinical document continues to exist in an unmodified state for a period determined by local and regulatory requirements. The internal rules of the organization regulate the storage of the document.

The CDA standard’s requirements and uses focus on creating standardized, structured, and semantically interoperable clinical documents.

Requirements:

  1. 1.

    The first requirement for a CDA is that its format is structured, and the document must contain specific sections and elements to ensure consistency and understanding of the data.

  2. 2.

    CDA includes using standardized terminologies and encoding to ensure semantic interaction, including codes for procedures, diagnoses, drugs, and other items.

  3. 3.

    An essential factor in creating context and identifying healthcare stakeholders is that the CDA contains information about the document’s source (e.g., healthcare organization) and patient.

  4. 4.

    CDA provides a means to embed clinical context into the document, including information about the physician, dates of procedures, and laboratory results.

  5. 5.

    CDA can support various documents such as examination reports, case histories, treatment plans, and others, making it versatile for multiple clinical scenarios.

The use of ready-made solutions and platforms that support the CDA standard, such as EHR and other tools, already have been implemented byhealthcare organizations, ensuring integration of the CDA standard with existing healthcare systems and information technologies, such as electronic medical records, hospital management systems, etc. (HIMSS Adoption Model for Analytics Maturity (AMAM) 2023; Ghatnekar et al. 2021; Blackley et al. 2020).

Automated tools (template auto-completion and electronic forms integration) are used to create, transmit, and store CDA documents. Creating simple and intuitive user interfaces for CDA is the basis for users to help users quickly adapt to new technologies. Providing feedback and the availability of a technical support mechanism for users will help solve potential technical problems and provide assistance if necessary. Attraction to decision-making for all interested parties, including medical and technical personnel and the administration, will help to ensure broad support and understanding of the importance of implementing the standard.

The CDA may include links to external resources such as images, files, or other documents. Despite its many advantages, clinical document architecture has its limitations, making it difficult to use in long-term storage and use of research data.

5 Digital Imaging and Communications in Medicine (DICOM)

The Digital Imaging and Communications in Medicine (DICOM) standard provides a critical framework for systems for the long-term storage and use of research data in the medical field. DICOM provides a standardized format for storing and exchanging medical images, including data such as X-rays, magnetic resonance imaging (MRI), computed tomography (CT), and other methods. A single format makes processing, analyzing, and visualizing these images more accessible (Ghatnekar et al. 2021; Blackley et al. 2019; Blackley et al. 2020; Gottlieb et al. 2021). DICOM includes metadata standards, including patients about patients, information about the device, and date and time that provide context and semantic information for medical images, which is essential for correctly interpreting and using data. DICOM provides interoperability with health information systems, EHRs, and other systems, making integrating medical images into the overall information flow easy (Torab-Miandoab et al. 2023).

DICOM provides frameworks and guidelines for long-term storage and archiving of medical images and includes standards for data backup, recovery, and long-term availability. DICOM provides mechanisms to ensure the confidentiality and security of health data, including encryption and authentication mechanisms, which are essential to comply with laws and protect sensitive health information. The DICOM standard is regularly updated and expanded to meet needs and emerging technologies in the field of medical imaging, ensuring long-term adaptation to new requirements and technological innovations.

DICOM defines a messaging protocol. The DICOM protocol is compatible with the control of the transmission and the Internet protocol, which allows the objects of the DICOM application to communicate via the Internet. Examples of DIMSE are N-SET (Attribute Setting), C-STORE (Store), C-FIND (Query/Retrieve), C-MOVE (Query/Retrieve), C-GET (Query/Retrieve), C-ECHO (Verification), N-GET (Attribute Retrieval). These service primitives allow devices to communicate with each other, send and receive requests, transfer data, and manage various aspects of health information on a DICOM network. Each service primitive has its format and structure, defined in the DICOM standard.

DICOM servers and archives are a third option. DICOM servers provide the infrastructure for storing, retrieving, and transmitting medical images in DICOM format. These servers can be deployed as local systems in medical institutions or as part of cloud-based DICOM archives, providing centralized and convenient access to medical data.

Each of these implementations of the DICOM protocol has its characteristics and applications. Software packages typically provide rich medical image processing and analysis capabilities, while built-in support in medical equipment provides transparency to end users. DICOM servers and archives offer the infrastructure for efficient storage and exchange of large volumes of medical data. DICOM’s internal mechanisms are continually updated to support new types of medical devices. As technology advances, new types of equipment may appear, such as more advanced scanners, tomographs, and MRI machines, and internal mechanisms must be adapted to communicate with these devices.

With improvements in power and image processing algorithms, DICOM’s internal mechanisms were implemented to support more complex applications critical for diagnostics and analysis. With the development of 3D visualization and virtual reality technology, the internal mechanisms of DICOM have expanded to algorithms for data exchange and processing, which is especially important for surgery and diagnostics. Due to increasing cybersecurity threats, DICOM’s internal mechanisms have sometimes worked out to ensure the confidentiality, integrity, and availability of health data. In some cases, DICOM’s internal mechanisms can support integration with artificial intelligence systems, allowing machine learning and AI algorithms to analyze medical data. Furthermore, DICOM provides standardized formats for telemedicine consultations and remote exchange of medical images, which are increasingly important in modern healthcare. Overall, DICOM plays a crucial role in ensuring standardized, efficient, and secure storage and use of medical images in the long term.

6 Documentation in Healthcare. Electronic Health Records (EHR) Documentation

Healthcare documentation, especially EHRs, plays a crucial role in long-term digital storage and usage of research data systems. EHRs comprise various data on patient patient information, including medical history, laboratory results, diagnoses, treatments, and other data. This comprehensive information is a valuable source for long-term research and analysis.

EHRs can provide long-term data storage by ensuring information security for many years, which is essential to assure data availability for subsequent research and monitoring. EHRs often includes standardized formats and terminology, such as HL7 and LOINC, which facilitates collecting and integrating data from various sources (Corpas et al. 2018; Vuokko et al. 2023; Cangioli et al. 2023; HIMSS Adoption Model for Analytics Maturity (AMAM) 2023; Cheemalapati et al. 2016; Tohmasi et al. 2021).

With aggregated data from EHRs, it is possible to monitor and analyze the health of a population over time. It helps develop public health plans and make strategic decisions. EHRs may include mechanisms to maintain security and privacy standards, essential from an ethical and legal compliance perspective when dealing with health information. Clinicians often encounter several challenges related to the usability of existing EHRs, as developers and vendors not always actively engage clinicians to understand their needs to create more user-friendly, efficient, and intuitive solutions. Updates and modifications to EHR systems must consider end-user feedback to ensure optimal usability and improve the efficiency of healthcare staff (Cangioli et al. 2023; HIMSS Adoption Model for Analytics Maturity (AMAM) 2023; van Buchem et al. 2021; Wang et al. 2021).

One possible solution is to create personalized interactions between humans and computers with an automated AI system. This interaction can optimize EHR functionality for a context-sensitive adaptive documentation process, which only requires human influence to assist the machine when necessary. Implementing AI in EHRs can manifest in various ways, for example in predicting complications and risks.

A relevant example is the development of AI-based recommendation systems that provide doctors with personalized recommendations on treatment selection, drug dosage, and other medical issues. The most common use of AI in EHRs is analytics to monitor patient wait times and optimize appointment scheduling to reduce delays and improve service.

These examples demonstrate how implementing AI in EHRs can potentially improve the quality of patient care in the future by enabling more accurate diagnoses, personalized treatment, and streamlined healthcare workflows.

The above components’ combined implementation and effective interaction create an integrated system that facilitates long-term health data storage, management, and use. The integrative system advances modern research methods, improves patient care, and builds the basis for future innovations in healthcare.