This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Mental Health, is properly cited. The complete bibliographic information, a link to the original publication on http://mental.jmir.org/, as well as this copyright and license information must be included.
Adverse Childhood Experiences (ACEs), a set of negative events and processes that a person might encounter during childhood and adolescence, have been proven to be linked to increased risks of a multitude of negative health outcomes and conditions when children reach adulthood and beyond.
To better understand the relationship between ACEs and their relevant risk factors with associated health outcomes and to eventually design and implement preventive interventions, access to an integrated coherent dataset is needed. Therefore, we implemented a formal ontology as a resource to allow the mental health community to facilitate data integration and knowledge modeling and to improve ACEs’ surveillance and research.
We use advanced knowledge representation and semantic Web tools and techniques to implement the ontology. The current implementation of the ontology is expressed in the description logic ALCRIQ(D), a sublogic of Web Ontology Language (OWL 2).
The ACEs Ontology has been implemented and made available to the mental health community and the public via the BioPortal repository. Moreover, multiple use-case scenarios have been introduced to showcase and evaluate the usability of the ontology in action. The ontology was created to be used by major actors in the ACEs community with different applications, from the diagnosis of individuals and predicting potential negative outcomes that they might encounter to the prevention of ACEs in a population and designing interventions and policies.
The ACEs Ontology provides a uniform and reusable semantic network and an integrated knowledge structure for mental health practitioners and researchers to improve ACEs’ surveillance and evaluation.
The study of Adverse Childhood Experiences (ACEs) and their consequences in terms of diseases and health risks has emerged during the past 20 years [
The complexity of ACEs’ causal pathway demands comprehensive and multidimensional data coordination. To design effective preventive interventions to reduce the burdens of ACEs, researchers and clinicians need access to a consistent knowledge-driven, evidence-based comprehensive analytic framework to study and monitor the causes of ACEs and their impacts on health (ie, obesity, mental health, and substance abuse), education (ie, cognitive developments, educational attainment, and graduation rates), and social dimensions (ie, placement in foster care and involvement with the justice system). The ideal ACEs knowledge-based system should be able to capture and identify ACEs indicators, detect individuals and groups at high risk of ACEs, integrate and validate ACEs health and social determinants, as well as exposure and genetic variations at individual and population levels. In order to improve the surveillance of ACEs, it is crucial to access a standard vocabulary that facilitates data collection, analysis, interpretation, and exchange of data between different parties and disciplines. An ontology is a standardized computational artifact that is used to capture, represent, and reason about the knowledge in the field. Ontologies capture knowledge by defining concepts, instances, relationships, and axioms. They increase the interoperability between different data sources and systems and improve the dissemination of data and knowledge across different disciplines. They also allow the use of semantic technologies to reveal new associations between the datasets and, therefore, discover new knowledge through logical inference. Ontologies are widely used in health and biomedicine and have made a substantial contribution to translational and clinical research as well as public and personalized health care [
The creation of the ACEs Ontology is rooted in recent efforts to study the effects of early experiences and social-environmental factors on children’s development and life-course health. These efforts have been made by multiple actors from different disciplines at various levels, including international organizations (eg, the World Health Organization [
While several ontologies exist in the domain of mental health, the ACEs Ontology is designed to deal with a carefully restricted scope, which is the study, prevention, and treatment of ACEs. We have included in the ontology some aspects that we deem important because they are tangentially connected to ACEs, such as potential health outcomes as well as their possible causes and aggravating conditions. Many other projects and resources exist that provide more detailed views of these adjacent domains; for example, the Children’s Health Exposure Analysis Resource (CHEAR) project [
In this paper, we present the ACEs Ontology as a formal reusable resource that can be used by the mental health research community to advance the surveillance and study of ACEs. The ACEs ontological structure provides a semantic backbone supporting the entire data model and hierarchy and supports logical inference and query answering. The ontology is used to form a consensus on the kind of data that is relevant and to create a common lexicon that makes it easier to share and reuse knowledge and information in the domain. The ACEs Ontology is an integral part of the Semantic Platform for Adverse Childhood Experiences Surveillance (SPACES) [
To facilitate the integration of ACEs-related data from different data sources, the ontology reuses several existing ontological concepts and relations as much as possible, while introducing new components. In some cases, the ontology integrates multiple existing elements in order to represent new pieces of knowledge. The surveillance of ACEs is focused on monitoring, detection, and prevention of ACEs as well as studying their causality pathways and short- or long-term consequences. The knowledge required to describe these activities is linked to concepts and terminologies coming from a wide variety of domains, including medicine, legal justice, cognitive and personal behavior, and community responses. This means that building an integrated ontology and knowledge base that covers the whole range of relevant subjects is a tedious, time-consuming, and complex task.
To build the foundation for the ontology, we identified five different key pieces of knowledge in our scope of study. As shown in
The top conceptual model for the Adverse Childhood Experiences (ACEs) Ontology, demonstrating the interactions between five major concepts: Person, ACEs, Social Determinants of Health, Interventions, and Negative Health Outcomes.
An abstract representation of some of the ontological concepts along with their relationships with the other thesauri in the field. SNOMED CT: Systematized Nomenclature of Medicine—Clinical Terms; ACEs: Adverse Childhood Experiences; NCIT: National Cancer Institute Thesaurus; MedDRA: Medical Dictionary for Regulatory Activities.
An important decision when building the ontology was determining its overall expressivity. The more expressive the ontology, the higher the complexity of its associated reasoning tasks. In a similar way, increasing the complexity of an ontology makes it harder to maintain, as it becomes more difficult to make sure that no two inconsistent statements cohabit within. For this reason, and to encourage reusability of the ontologies, many biomedical ontologies focus almost exclusively on the taxonomical part of the ontology, which is defining inclusion relationships between various concepts (eg,
The current implementation of the ontology [
A fragment of the Adverse Childhood Experiences (ACEs) Ontology representing partial class and property hierarchies. OWL: Web Ontology Language.
The Adverse Childhood Experiences (ACEs) Ontology on the National Center for Biomedical Ontology's (NCBO) BioPortal [
Some general axioms defined in the Adverse Childhood Experiences (ACEs) Ontology. OWL: Web Ontology Language.
In addition to the actual ontology, we also use semantic rules. In our ontology, some rules are used to enable data access. While the ontology provides a language to express the knowledge, the actual data itself is usually stored in data files or repositories that can use a different lexicon. Rules are employed to map the actual data to the ontology to enable knowledge-based querying and inference. Languages, such as Positional-Slotted, Object-Applicative (PSOA) Rule Markup Language (RuleML) [
It is not possible to express exactly the existence of such a homomorphism in OWL. Several solutions can be considered. For instance, it is not possible to express that “i_p_h_t_r_i_i_t (inflicted_physical_harm_that_resulted_in_injuries_to) ≅ i_p_h_t_r_i_i (inflicted_physical_harm_that _resulted_in_injuries) o targets” but only that “i_p_h_t_r_i_i o targets ⊆ i_p_h_t_r_i_i_t.” As this representation does not allow us to fully express the knowledge and check its consistency, we can use an alternative solution by employing rules external to the ontology, such as “Physically Abused(x) → ∃ y, z has parent(x, y) ∧i_p_h_t_r_i_i(y, z) ∧ targets(z, x),” where x and y would span
The third application of rules is to build a recommendation system. In that case, the left-hand side of a rule is a recommendation made to the user. Depending on the use case, the rule and the recommendation can take many different forms. For instance, if the ontology is used during an interview about ACEs with a child patient, the rule could be “If the patient's parents are separated or divorced, ask if they are feeling loved.“ These rules are particularly important in the context of ACEs detection due to the sensitive information (eg, mental illness, handicap, and criminal activity) that needs to be approached carefully.
Another key contribution of this kind of rule is to manage and allocate resources. For example, “If the patient suffers from emotional neglect, schedule an appointment with a child psychologist” (see
By definition, rules are the result of causal reasoning (ie, if there is a match for the left-hand side, effect the right-hand side) and it is thus natural to connect the rules to causal diagrams. In the current state of this project, the rules are generated from common knowledge and currently established standard procedures and diagnostic processes. We employ causal diagrams [
A piece of data representing someone (rectangular node) who was physically injured by a parent (round node). i_p_h_t_r_i_i_t: inflicted_physical_harm_that_resulted_in_injuries_to.
A SPARQL Protocol and Resource Description Framework (RDF) Query Language (SPARQL) query that could be the left-hand side of a rule instantiating the data with Physically Abused(x). i_p_h_t_r_i_i: inflicted_physical_harm_that _resulted_in_injuries.
A recommendation rule that schedules an appointment with a psychologist if a patient suffers from emotional abuse (Em_Ab).
We present three different use-case scenarios for the ACEs Ontology to showcase the role and significance of the ontology to improving ACEs’ monitoring, detection, prevention, and management.
The first use case takes place inside a clinic where a nurse practitioner is interviewing a child patient and her parent. The parent gives personal identification data (ie, name, age, sex, and address) as well as a description of any potential symptoms. This information will be linked to other information that already exists about this patient in the electronic medical record. From the address, using a different data source that gathers data about social determinants of health, the ACEs Ontology helps to infer the patient's social economic status (eg, patient's access to public transportation, the neighborhood's safety and poverty rate, and residential proximity to schools and daycares). The medical practitioner can then use the collected information, as well as the inferred knowledge, to diagnose the illnesses or make a new hypothesis and continue to further investigate the case. Additionally, during the interview, the health practitioner will be able to ask questions relating to ACEs, for example, by slowly introducing household challenges (eg, “Are the parents divorced or separated?” or “Is a household member incarcerated?”).
Depending on the answers, using the semantic links in the ontology, the medical practitioner will be able to ask more questions, recommended by the knowledge-based system, to reveal important clues and signs for detection of ACEs. The ontology is used both to allow access to different data sources that share important information and to foster reasoning that will facilitate the knowledge exchange between actors. In this scenario, the goal is mostly to increase the information and knowledge about the patient. Once the data is collected, however, it can be used to formulate a diagnosis. For instance, the ontology can be used to answer the question “Given that the patient has symptoms S0,..., Sn and an ACEs score of 4, which are the likely negative health outcomes to screen them for?”). An equivalent query in SPARQL Protocol and Resource Description Framework (RDF) Query Language (SPARQL) [
Policy makers and public health organizations can use the SPACES framework for knowledge-based population health surveillance [
A SPARQL Protocol and Resource Description Framework (RDF) Query Language (SPARQL) query used to discover which negative health outcomes to screen a patient for. nho: negative health outcome; ACEs: Adverse Childhood Experiences.
A SPARQL Protocol and Resource Description Framework (RDF) Query Language (SPARQL) query used to select an appropriate intervention.
Another key use of the SPACES framework is in detecting potential problematic areas in order to intervene. As an example, because inflicting abuse is a criminal activity, collecting information on ACEs from the parents can yield less-reliable data than data about neighborhood conditions. It is thus easier to query the data about SDH that are prevalent in the neighborhood in which the patient is living and, thus, which ACEs are the most likely to be a risk. A typical semantic query for such cases can be formulated as “Given that the social determinants of health S0,..., Sn have been observed in the area, what are the ACEs to screen for?” Similarly, negative health outcomes are easier to track because they often result in hospital visits or medical interventions. The question “Given that the social determinants of health S0,..., Sn have been observed, and that the negative health outcomes O0,..., Om have been frequent in the area, what are the ACEs to screen for?”
In this paper, we introduced an ontology for improving the surveillance of ACEs. The goal of the ontology is to provide a uniform structure to represent current and future studies on the causes and effects of, and ways to prevent and mitigate, ACEs. The benefits of using ontologies and semantic technologies have already been shown in several biomedical domains ranging from clinical surveillance [
The ontology is constantly growing with the advances in knowledge by adding new relevant concepts and relationships as well as new axioms. It is possible for some applications to import only the taxonomy to represent data in a standardized way, while others might use constructs such as transitivity for more complex reasoning. However, the benefits of the ontology are limited by the quality of the data available. Currently, most existing studies deal with adults for whom the negative outcomes are apparent, while more transversal studies of the impact of ACEs on the development of children are lacking [
Adverse Childhood Experience
Adverse Childhood Experiences Ontology
US Centers for Disease Control and Prevention
Children’s Health Exposure Analysis Resource
emotional abuse
inflicted_physical_harm_that _resulted_in_injuries
inflicted_physical_harm_that_resulted_in_injuries_to
Medical Dictionary for Regulatory Activities
National Center for Biomedical Ontology
National Cancer Institute Thesaurus
negative health outcome
Web Ontology Language
Positional-Slotted, Object-Applicative
Resource Description Framework
Rule Markup Language
social determinants of health
Systematized Nomenclature of Medicine—Clinical Terms
Semantic Platform for Adverse Childhood Experiences Surveillance
SPARQL Protocol and RDF Query Language
We would like to thank Dr Robert L Davis, Dr Jonathan A McCullers, Dr Sandra R Arnold, and the entire team at the Family Resilience Initiative at Le Bonheur Children's Hospital, Memphis, Tennessee, for their support and insights. This research was supported by the Memphis Research Consortium.
None declared.