MedicalForms: Integrated Management of Semantics for Electronic Health Record Systems and Research Platforms

Abstract

(1) Background: Clinical information modeling tools are software instruments designed to support the definition of semantic structures able to be implemented in health information systems. Based on the analysis of existing tools, this research developed a tool that proposes new approaches to promoting clinician involvement and supporting information modeling processes through mechanisms that ensure governance, information consistency and consensus building. (2) Method: This research developed the MedicalForms system, which is based on the requirements identified in both a Delphi study about tool requirements and the ISO/TS 13972 specifications. (3) Results: This system allows the management of projects, information structures and implementable forms related to clinical documentation. Users can easily define clinical documents in collaboration with the rest of the professionals in their team by being able to reuse previously defined forms, terminologies and information structures. The system is able to export the defined forms as interoperable specifications or as several implementable form formats compatible with multiple open source EHR systems and research platforms. End user perception of this tool was evaluated through the Technology Acceptance Questionnaire with satisfactory results. Finally, the system was applied to develop 12 research registries and 2 clinical trial research forms, 3 mobile applications and 1 decision support system.

1. Introduction

Semantic interoperability in health informatics is a field where multiple technologies and specifications have been defined. Multiple standard development organizations (HL7, ISO and CEN) and other international initiatives (IHE, CDISC, openEHR) work towards defining how information should be structured to allow computer consistent processing and interpretation in well-defined scenarios. As long as Electronic Health Record (EHR) systems are becoming more mature, they are advancing towards supporting patient information flow associated with continuity of care and reusing healthcare data scenarios such as research, epidemiological analysis, business analytics and decision support. As part of this research, we use the term clinical information model (CIM) to describe any semantic structure able to represent how clinical concepts are interrelated, including terminology bindings, to be able to provide the full context to safely process this information in multiple EHR systems or research platforms.

This paper explains the development and validation of a new tool focused on promoting clinician involvement in processes associated with the defined implementable forms and semantic structures for EHR systems through mechanisms that ensure governance, information consistency and consensus building. As a result, improvements in management of semantic relationships between EHR concepts could contribute towards the reuse of clinical data for multiple use cases such as clinical decision support integrated with biomedical knowledge [1], advanced analytics and wearable technologies [2].

1.1. Quality Standards for Developing Clinical Information Models

The clinical information modeling process (CIMP) could be described as an iterative process that includes the analysis of the domain and requirements, designing, implementing, validating and maintaining CIMs. The analysis of how the multiple interoperability resources are defined shows that there are no particular differences on the processes applied for modeling CIMs without regard for the adopted specification [3]. On the other hand, the process carried out for modeling clinical information is identified as a relevant factor to determine the quality of the semantic interoperability resources [4].

In the clinical information modeling field, ISO 13972 Technical Specification describes how to implement quality processes that lead to the adoption of best practices based on the definition of CIMs [5]. Moreover, the standard details a set of testable quality attributes for these resulting models and how to implement a Quality Management System for the CIMP. The implementation of a Quality Management System allows a continuous improvement cycle to be established. The definition of a “Plan, Do, Check, Act” (PDCA) cycle [6] promotes the continuous adaptation of processes and measurements within each of the steps to obtain improved quality in the final product. As a result, a software-based methodology based on the ISO 13972 specification could guide the participation of multiple domain experts in tasks associated with definition, validation and maintenance of CIMs in compliance with defined best practice and support the adoption of a continuous improvement cycle associated with information management.

1.2. Interoperability Specifications and International Initiatives

Some of the most relevant international initiatives are openEHR [7] and the HL7 CIMI working groups [8], which are focused on defining CIMs. These initiatives are focused on obtaining a consistent semantic definition of CIMs able to be implemented in multiple scenarios. They are working on defining CIMs compatible with Archetype Description Language, each of them on their own specific reference model. CIMI is working on the definition of CIMs able to be the CIMI Reference Model Specification and define HL7 FHIR profiles. On the other hand, the openEHR CIMs are known as archetypes based on the openEHR reference models and they are able to be transferred as EHR extracts [7]. In addition, they are working on supporting mappings to HL7 FHIR resources given the increased acceptance of EHR vendors.

There are additional specifications that aim to define how clinical information is structured in order to be transferred between EHR systems. Some of the most relevant specifications are the Clinical Information Model [5], HL7 CDA templates [9] and ISO 13606 [10]. There are national and regional Ministries of Health working on each of these specifications. The CDISC consortium developed several specifications to define how information associated with clinical trials should be structured. They have several standards which are able to define how to satisfy regulatory submission to the Food and Drug Administration (FDA) such as the Operational Data Model (ODM) [11]. Some of their standards are harmonizing element names, definitions and metadata to establish a standardized data collection baseline across all submissions through Clinical Data Acquisition Standards Harmonization (CDASH) [12].

1.3. Clinical Information Modeling Process and Tools

Clinical information modeling tools are software platforms designed to support the processes associated with the definition of CIMs, as well as establish governance for the multiple CIMs applicable within an infrastructure or domain. Currently there are multiple tools such as the openEHR suite, DCM suite and LinkEHR that provide the mechanisms for defining and managing CIMs. A previous evaluation of existing modeling tools showed that they generally have a good adoption of functionalities related to the management EHR specifications, data types, terminology binding and CIM metadata. However, the need to improve tool support for information modeling and software development processes has been identified, especially in those areas related to governance, clinician involvement and optimizing the technical validation of testing processes [13]. In addition, several experts pointed out the need to improve modeling tools to provide better support for the modeling process in an international survey about CIMP [14].

1.4. Implementation as Medical Form for Clinical Care or Research

Instead of focusing only on producing semantic structures addressing the needs defined by existing health informatics specifications, our approach aims to also produce semantic structures able to be widely implemented for use in systems. There are multiple open source initiatives for EHR systems and research platforms that provide mechanisms to define implementable forms. They are able to import and process the defined forms from an external file that specifies how to display multiple items on a screen, including presentation and user interaction functionalities. This research study has identified the following systems as relevant initiatives able to test the implementation of semantic structures for data collection or analytics tasks:

• Diraya is an open source EHR system developed by the Andalusian Health Service able to be shared with any public administrations in all the primary care centers of the region. This system has been deployed to provide care for 8 million people through the full network of 29 hospitals and 1500 Primary care centers distributed throughout the region [15,16].
• RedCap and OpenClinica are the most widespread open source Clinical Trial Management Systems (CMTS). OpenClinica is an open source software that has been used in 3 million clinical trials in 100 different countries [17]. REDCap is an open source software that has been used by 2800 institutions in 120 countries in 500,000 projects [18]. These systems allow the documentation of patients’ clinical information in accordance with FDA regulations.
• TranSmart and i2b2 research platforms are patient cohort analysis systems designed to analyze data obtained through conventional care while ensuring compliance with patient privacy requirements. This system supports the anonymized analysis of specific patient cohorts. As a consequence, it facilitates the generation of new research hypotheses based on the analysis of the patient population treated at their centers [19,20].
• The CroniCare system is a mobile health platform that is focused on establishing mechanisms for an agile and customized development of interventions based on the use of mobile technologies. The system allows users to configure the information, interactions, questionnaires and alarms that each patient will access through the mobile app in order to improve the capacity for self-management, empowerment and control of chronic and multimorbidity patients [21,22].

2. Materials and Methods

2.1. Project Objective

This project aimed to develop and validate a system for information management in the healthcare and the research field. This system should enable the participation of clinicians in CIMP to produce standards-based specifications that support the interoperability of health information systems, as well as implementable forms and semantic structures for EHR systems and research platforms.

We aimed to advance knowledge in areas related to promoting clinician involvement and support CIMP through mechanisms that ensure governance, information consistency and consensus building.

2.2. Requirement Definition

The definition of requirements was based on: (i) CIMP identified as part of the systematic literature review about papers talking about semantic interoperability in EHR systems [3]; (ii) metrics defined as part of the ISO 13972 standard for implementing a Quality Management System [6]; and (iii) essential requirements for modeling tools [23].

2.3. Use Cases and Validation

The MedicalForm system has designed a set of use cases to validate its proper management of semantics for:

• Representing international datasets, structures and forms: This use case will be validated through the definition of semantic structures and forms specified as part of the European Heart Failure Summary defined by SemanticHealthNet project [24].
• Representing datasets, structures and forms for large healthcare providers: This use case will be validated through the definition of semantic structures and forms according to the Andalusian Stroke Integrated Care process EHR [25].
• Export semantic structures as compliance instances with multiple relevant health informatics specifications. This use case will be validated by exporting CIMs as an implementable interoperability specification such as ISO 13606 and CDISC.
• Export semantic structures with W3C semantic web technology stack. This use case will be validated exporting CIMs according to the Web Ontology Language (OWL).
• Define implementable forms for research platforms, EHR systems and mHealth solutions: This use case will be validated by implementing a questionnaires in platforms (ITCBio, openClinica and Redcap), Diraya EHR system and CroniCare mobile app for patient follow-up from hospital [21,22].
• Define a mechanism for reusing collected data in data analytics tools: This use case will be validated with the database configuration in TranSmart and i2b2 analytics tool. As a result, it will be required to be translated to define a semantic structure in the analytics tools in the form of a tree of concepts consistent with the MedicalForm definition in order to ensure the appropriate management of semantics in the performed analysis.

2.4. End User Evaluation

There was a defined questionnaire based on the Technology Acceptance Model [26] addressing user perception through 14 questions about perceived usefulness (U), ease of use, attitude toward using (AU) and behavioral intention to use (IU). The answers for each dimension were collected through a 5-point Likert Scale (from 1—high level of disagreement to 5—high level of agreement). This questionnaire was applied to collect the perceptions from members of the hospital committee for clinical documentation and healthcare professionals who required the definition of the EHR model or research registries in our hospital.

• Group of Members of the Clinical Documentation Committee: The system was presented to a group of end users in a 45-min session in order to obtain a functional evaluation of the system by healthcare professionals, experts in clinical documentation and EHR systems. At the end of the session, the professionals filled in the TAM questionnaire on a voluntary basis.
• Group of Principal Investigators: These users received a 45-min training session and used the tool to define the forms to create a research record on the patients they cared for during their daily practice. As described in the piloting section for the definition, several revisions were made to the forms until reaching the consensus version that would finally be implemented in the registry. After finalizing the definition of the forms, users completed the TAM questionnaire on a voluntary basis.

2.5. Implementation in Real Use

The system was required to be applied in multiple clinical domains to support the specification of how information is structured for research and healthcare scenarios at the Virgen Macarena University Hospital.

3. Results

3.1. Technological Infrastructure

The MedicalForms system was developed on a Linux Centos server with AngularJS framework (frontend), NodeJS (BackEnd) and MongoDB. The system includes an Enterprise Service Bus (MirthConnect) to support model and form exportation to external systems. Figure 1 details the system architecture and exporting formats. The whole system was included as a docker container; this makes it easily deployed. The MedicalForms website provides additional information about how to access and test it [27].

3.2. Requirements Definition

The current version of the MedicalForms system was evaluated against the requirements defined as part of the Delphi study about tool requirements [23] and compared with obtained results from 11 CIMT identified in a previous study [13]. The evaluation presented in Figure 2 shows that the MedicalForms system has:

• An excellent level of fulfillment: in areas related to clinician involvement, collaboration, searching capabilities and metadata of the CIMs.
• Above average level of compliance: in areas related to communication with terminology servers.
• Average level of compliance: for supporting the management of semantics, as well as the ontology and terminology binding process.
• Below average level of compliance: in areas related to supporting CIM evolution and specialization and support of the testing and validation processes.

3.3. Clinical Information Process Support

The MedicalForms system was designed as a software instrument to support the coordination of those healthcare professionals involved in the modeling process. Figure 2 details how multiple professionals involved in multiple tasks associated with CIMP could benefit from system functionalities. This system includes roles to classify users up to four levels to support the establishment of governance in large healthcare organizations or international modeling initiatives.

The first level includes the coordinators that ensure the establishment of the information governance process. They coordinate and manage the definition of resources applicable for the multiple healthcare domains in the generic CIM form. The second level includes a project leader that will define the project scope and guide clinicians participating in the CIMP. The third level includes a core team of multidisciplinary experts who work in depth on the detailed clinical and technical needs that the system and CIMs will need to satisfy. The fourth level comprises a larger group of domain experts responsible for validating the proposed clinical document or EHR form. Figure 3 details how multiple roles interact as part of this process.

3.4. System Description

The system was specifically designed to be understood by clinicians and end users without knowledge about technical specifications. They can work on one or multiple projects, including the following areas:

• Terminology: Allows to define atomic semantic structures corresponding to basic data types specifying bindings to terminology concepts. This section allows basic semantic structures to be generated that will be equivalent for the most commonly used data types included in ISO 21090 for dates, number, text, sections, cluster, physical magnitude and autocalculated results. Although it was expected to include additional data types as long as they were required, this subset was enough to address multiple use cases described in future sections.
• Structures: Allows to assemble and organize previously defined atomic structures establishing hierarchical and complex structures. This section includes an additional type of item that allows for defining indexes that provide associated value to a specific answer in a dropdown menu. They are relevant for medical scores where multiple options are mapped to a value rather than a terminology concept.
• Forms: Allows to define how structures are presented on screen detailing user interaction and supporting specialization through combination and post-coordination capabilities. This section adds new elements focused on specifying presentation capabilities for those structures previously defined, including table, radiobutton, checkbox, file and differs between text area and text field.

The system includes a repository where users can navigate between the defined semantic structures and forms. Figure 4 details how the system allows the creation and reuse of semantic structures and forms, and the multiple options for exportation.

3.5. Semantic Infrastructure

In order to support the definition of terminology subsets, structures and forms, the MedicalForms system defined a set of basic elements. The following table (

Table 1. Description of the most relevant classes of the MedicalForms JSON information structure.

MedicalForm Classes	Description	Options
Type	Details the type of element	section
		tab
		table
		header
		number
		magnitude
		text
		textArea
		select
		checkbox
		radioGroup
		autocalculated
TypeStructure	Details if the structure is composed by multiple items	complex
		simple
TypeShow	Details if the structure is able to contain other elements	container
		item

) describes the most relevant classes of the MedicalForms JSON information structure. This structure has been defined according to the needs for specifying multiple elements able to be included in MedicalForms according to the type of interaction with the end user (type class), how many items can be interrelated (typeStructure class) and its representation on the screen (typeShow class). Appendix B. Semantic Infrastructure description describes the multiple elements included as part of the MedicalForms system specifying its associated attributes and how they are mapped to ISO 21090 Datatypes, ISO 13606 and openClinica CTMS.

3.6. Semantics Capabilities for Information Consistency

The system allows the establishment of a highly controlled CIMP where the project leaders and editorial team are allowed to generate any semantic structure or form based on the existing terminologies already incorporated in the system. Once the need for a concept that is not included in the existing terminologies is identified, the coordinator is required to be contacted for additions. This will ensure that all data items collected will be consistent with the terminologies approved and well known in an organization.

The system includes simplified functionality for post-coordinating multiple terminology concepts. This allows us to define semantic structures based on pre-coordinated terminology concepts that are post-coordinated when they are implemented as screen forms. This is a very relevant point as semantic structures based on pre-coordinated concepts are able to support a larger number of clinical scenarios with easier management of semantics to support aggregation and compatibility for data exploitation and interoperability. On the other hand, clinicians prefer to use post-coordinated terms when they are working on a specialized domain since it reduces the number of items that need to be recorded in the form. Current initiatives such as CIMI WG are working with a similar approach defining iso-semantic transformation from the precoordinated CIM to each specific scenario implemented.

A library was developed containing generic CIMs complemented with educational material about how to manage with consistency negation, thresholds, indexes and calculation questionnaires, as well as options for others in multiple option questions.

3.7. Promoting the Participation of Clinicians in CIMP

The MedicalForms system was designed with special focus on supporting the participation of clinicians as part of the CIMP. Rather than applying technical concepts from EHR specifications, we chose the words “terminologies”, “structures” and “forms” to explain the modeling process emphasizing the benefits from reusing the library of available structures able to be specialized when they are adapted in forms. User interface was simplified to allow clinicians to drag and drop any structure within a CIM in order to support freedom in the definition. Lastly, functionalities designed to support consensus building were included. There is a forum associated with each CIM and a survey tool that helps to obtain feedback from multiple clinical experts working in a project.

3.8. Validating the Management of Semantics for International and Large Healthcare Provider Use Cases

The semantic validation was performed based on the Heart Failure Summary developed within SemanticHealthNet and represented as openEHR archetypes and templates. As part of this project, 18 openehr archetypes were applied to develop a template focused on representing a clinically-orientated summary of a first visit to a consultant-led heart failure clinic (e.g., as a clinic visit record or letter to the GP) but meeting the needs of research/audit and care pathways where possible. Figure 5 shows how the Heart Failure Summary has been implemented, including multiple tabs for each of the main sections of the form. In addition, there were two structures generated and one form according to according to the defined Andalusian Stroke Integrated Care process to validate the system capabilities to address large healthcare provider needs

3.9. End User Perception

Next,

Table 2. Evaluation of end user perception through Technology Acceptance Model. Each domain provides the mean ± the standard deviation.

Group	n	Ease to Use	Usefulness	Attitude toward Using	Intention to Use
Clinical Documentation committee	9	3.1 ± 1.1	3.8 ± 0.8	3.8 ± 0.8	3.8 ± 0.8
Principal Investigators	6	4.3 ± 0.7	4.7 ± 0.5	4.8 ± 0.4	4.6 ± 0.5
Total	15	3.6 ± 1.1	4.2 ± 0.8	4.2 ± 0.8	4.1 ± 0.8

presents the perception from a group of nine members of the Clinical Documentation Committee (group 1) and a group of six principal investigators (group 2) collected through the Technology Acceptance Questionnaire. Global results show that both groups agree with the usefulness (4.2), attitude toward using (4.2) and intention to use (4.1) the tool. Although the overall result in the easy to use dimension was (3.6), there were differences found in the perception of how easy to use the tool was. On the one hand, group 1 only received 45 min of training about the system and resulted in a neutral perception (3.1). On the other hand, group 2 received 45 min of training and later agreed (4.3) that the system was easy to use.

3.10. Implementation

The system was applied to the development of 10 research registries, 1 decision support system and 3 mobile apps where our team participated. The included 60 users acted in the following roles: 1 regional coordinator, 6 project leaders, 9 editorial team and 44 validators.

Table 3. Details of the projects where the MedicalForms system was applied.

Domain	Units	Implementation	Participants	Results
Library of Questionnaires	Generic	Designed for reusability	1 regional coordinator 1 project leader 1 validator	12 structures
Library of Lab tests	Generic	Designed for reusability	1 regional coordinator 1 project leader 1 validator	30 structures
Library of Patient Constants	Generic	Designed for reusability	1 regional coordinator 1 project leader 1 validator	39 structures
Fracture Prevention	Traumatology	Research registry	1 project leader 2 validators	3 structures 1 form
Stroke Prevention	Neurology	Research registry	1 project leader 2 validators	1 structure 3 forms
Breast Cancer	Oncology	Research registry	1 project leader 3 validators	4 structures 1 form
National Registry of Recurrent Laryngeal Papillomatosis in Pediatric Patients	Pediatrics	Research registry	1 project leader 1 validator	4 forms
Anticoagulant decision support	Internal Medicine	Decision support system	1 project leader 2 editorial team members 1 validator	12 structures 2 forms
Stroke Unit	Neurology	Research registry	1 project leader 7 validators	2 forms
Headache	Neurology	Research registry	1 project leader 1 editorial team member 3 validators	3 forms
Cardiovascular Risk	Cardiology	Research registry	1 coordinator 8 validators	1 form
Ankle Fracture,	Traumatology	Research registry	1 project leader 1 editorial team member 1 validator	7 forms
Clozaid	Psychiatry	Clinical trial CRF	1 project leader 1 validator	11 forms
Heart Failure	Internal Medicine	Research registry	1 project leader 1 editorial team member 3 validators	1 form
PC COVID-19	Infectious diseases	Clinical trial CRF	1 project leader 1 validator	10 forms
Rhinosinusitis	Otorhinolaryngology	Mobile app Research registry	1 project leader 1 validator	10 forms
Cardiovascular rehabilitation	Cardiology	Mobile app Research registry	1 project leader 1 editorial team member 1 validator	1 form
Multimorbidity	4 departments	Mobile app Research registry	1 project leader 3 editorial team members 12 validators	1 form
TOTAL	10 departments	3 Mobile app 1 Decision support 12 research registries 2 Clinical trials	1 regional coordinator 6 project leaders 9 editorial team members 44 validators	101 structures 58 forms

details the different domains, medical units, participants and resultant forms and semantic structures for each of the projects carried out with the MedicalForms system.

4. Discussion

The MedicalForms system successfully managed and ensured semantic consistency in defined EHR specifications and forms for clinical practice and research. It was able to establish a coordinated management of information between EHR, CTMS, analytics platform and mobile apps. In addition, the system demonstrated the potential for reusing standardized approaches across multiple clinical domains. There were no particularities found in any of the clinical domains where the system was applied.

Clinician involvement: This research expects to contribute to guide future evolution of existing CIMTs towards obtaining a greater involvement of clinicians. Our survey based on the TAM questionnaire shows a good level of acceptance. The proposed simplified language that does not require learning new concepts reduces the learning curve associated with other approaches. Our approach confirmed results from a previous international survey [14] pointing out that applying screens and forms as part of the CIP is beneficial for clinicians in understanding system requirements.

Obtaining consensus in large healthcare providers is a time-consuming task. The proposed functionalities for supporting coordination between multiple clinical experts are expected to contribute towards systemizing this process resulting in a reduction of time in the long term.

Combined information modeling process with form design and data analytics: It is possible to find CIMTs that define how the information is to be structured in the screen form (for example, those based on the openEHR template), but these are not capable of producing implementable forms because this specification does not determine the options for representing a cluster such as dropdown, checkbox or radiobutton. Our research aims to provide a complementary approach that is not only focused on the definition of interoperability specifications for EHR communication, but on producing implementable forms for patient care and research with integrated data analytics.

Screen presentation harmonization: The necessity for finding a balance between standardization of common data elements and allowing for innovation, in order to preserve the capability for different interfaces to be created by vendors, has previously been reported [14]. It has been identified that several systems allow the definition of screen forms by importing a configuration file. It is of no surprise that more systems adopt this approach and possible advances in the harmonization of definitions could be implemented.

Semantic infrastructure definition: Our initiative is based on the CIMs since it can be identified as the more generic mechanism to adopt the two-level modeling approach. Based on CIM, we were able to produce forms and semantic structures compatible with existing tools such as openClinica, RedCap, Diraya and TranSMART with a reduced level of complexity compared to existing EHR specifications. In addition, we took care to satisfy requirements included as part of the ISO 13606, openEHR and CDISC reference models. Last, we analyzed existing HL7 specifications (v2, CD and FHIR) and planned a journey to support future compliance as long as we are involved in projects related to these specifications.

Semantic Web technologies: Associated with the capabilities for exporting CIMs in OWL format, it is expected that the MedicalForms system contributes towards the adoption of ontology-based tools. These technologies apply defined relationships between concepts from existing medical ontologies to provide advanced capabilities for clinical decision support, healthcare data analysis, retrieving synthesis information from the entered data and visualization capabilities.

Issues exporting semantic structures as implementable forms: The MedicalForm system was designed with the goal of supporting the definition of CIMs according to multiple EHR interoperability specifications. As a consequence, the system was required to be flexible enough to define semantic structures with no limitation on the number of depth levels that sections can have in compliance with the ISO 13606 and openEHR specifications. This causes certain limitations when we want to export implementable forms in systems such as OpenClinica or Diraya due to the maximum depth of sections in these systems being just one level. Although systems with increased flexibility of depth can be found, they always have a maximum level of depth. This is reasonable since the screen presentation capabilities will not be able to display an unlimited number of subsections. Two approaches were tested to overcome this problem: providing guidance to modelers about the limitation of depth of subsections and providing automatic mapping that reduces the number of sublevels included in the structure.

Issues designing implementable forms for data analytics: Given that the system has been designed to support the establishment of how data will be managed in data collection tools and analytics tools, it was recommended to define from the beginning those items that will be applied in the data analysis. This is especially relevant when data analysis is performed from variables that are derived from transformations and calculations from the original collected data. There were identified benefits from specifying these transformations at the moment of designing data collection since they were allowed to define automatic transformations for data analysis of the collected data.

Limitations and future work: This research was focused initially on the definition of a semantic consistent transformation between multiple EHR specifications with a generic reference model such as openEHR, ISO 13606 and CDISC. Additional work will be required to define a library of resources correctly mapped with other specifications whose reference model is dependent on the information carried out such as HL7 FHIR and HL7 CDA. It is expected that it will be carried out as long as future use cases based on these specifications will emerge. They will be able to benefit from the ongoing cross-mapping work carried out by openEHR and HL7.

5. Conclusions

The MedicalForms system proposes a consistent software-based methodology to support clinical information modeling providing support for terminology subset, information models and forms definition and validation. The system proposes innovative approaches for incorporating clinicians as part of the CIMP supporting the automatic mapping of defined semantic structures to some of the most relevant specifications (openEHR, ISO 13606 and CDISC) and widely deployed tools (OpenClinica, RedCap and Diraya).

System capabilities for semantic management have been validated through the definition of semantic structures and forms in complex international and large healthcare provider scenarios such as those addressed by the European Heart Failure Summary, the defined SemanticHealthNet project and the Andalusian Stroke integrated care process. The evaluation of the MedicalForms system against the requirements defined as part of the Delphi study about tool requirements [23] shows that this tool is above average in areas related to supporting clinician involvement, collaboration, terminology server communication and searching capabilities and metadata of the CIMs.

It has been implemented in a reference hospital that provides hospital and community care services to 480,000 people. The system provides support for developing more than 100 structures and 50 forms in 15 projects related to healthcare and research data collection. As a result, we expect that our system-based methodology could contribute to improve the semantic consistency in information between multiple research projects and healthcare scenarios in large healthcare providers and international scenarios. It is expected that our scalable approach could gradually incorporate new specifications and systems in the future. In addition, our initiative could be identified as a relevant example for other tools focused on clinical information modeling.

6. Patents

The Andalusian Technology Transfer Office registered MedicalForms: Herramienta de Modelado de Información Clínica (1909023179153) through the Save Creative digital copyright registration service.