Testing an individualized digital decision assist system for the diagnosis and management of mental and behavior disorders in children and adolescents

Mental health disorders today account for more years lived with disability (YLDs) than any other health condition, with close to 30% of all people worldwide having a mental health disorder at one point during their lifetimes [1, 2]. Children and adolescents are particularly vulnerable. Nearly 20% of all children and adolescents at any point in time have a diagnosable mental health disorder [3, 4]. Despite this, no more than one-third of all children and adolescents with mental health disorders worldwide receive the care they require [5].

In Norway, approximately 20% of all children and adolescents are experiencing mental health problems at any point in time [6, 7]. Both, academic pressure at school and social pressure among peers, contribute to children and adolescents reporting more stress now than ever before [8]. While primary care providers (PCPs) often serve as the first point of contact for patients and their families, still most PCPs do not hold adequate training to address complex neuropsychiatric disorders, such as autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD) [9‐12].

Inability to receive a timely diagnosis and appropriate early intervention for neuropsychiatric disorders in childhood and adolescence leads to higher risk of suffering from comorbid mental health problems, including alcohol and substance abuse, behavioral problems [9, 13]. Insufficient access to required child and adolescent mental health services (CAMHS), not only directly impacts the health of the patients and their families, but also can be costly for the society as a whole [14‐16]. Misdiagnosed or untreated mental health problems in childhood and adolescence can develop into lifelong mental health disorders requiring long-term care services. Across the European Union, the annual accumulated cost of providing mental health care for children and adolescents is estimated to be nearly €21 billion [16]. Undiagnosed or untreated disorders are also detrimental to the society’s future job market and economic stability, with more frequent work absences and lower productivity among individuals struggling with long-term mental illness [17].

Every year, roughly 5% of the child and adolescent population in Norway require specialist care from CAMHS [18, 19]. Among Norwegian twelve-year-olds alone, an estimated 3.4% have ADHD at any point in time; it is one of the most common reasons for referral to CAMHS [18]. If left untreated, ADHD can predispose children and adolescents to psychiatric and social pathology later in life [20, 21]. While Norway ensures every citizen to timely care by law [22], it is essential to optimize the quality of diagnostics and early intervention care for all children and adolescents displaying signs of mental illness. One modern approach to CAMHS quality assurance, that has the potential to be cost efficient and sustainable for long-term needs, is the incorporation of information and communication technologies (ICT), including clinical decision support systems (CDSSs) [23, 24].

A CDSS is a computerized system that offers guidance to clinicians in real-time while they provide clinical care by using existing, evidence based resources [25‐27]. A CDSS uses compiled evidence-based resources to provide clinicians with clinically relevant guidance and recommendations to help best meet the needs of each individual patient. CDSSs are not designed to replace clinicians. The architecture of a CDSS typically has three distinct components– a Clinical Knowledge Base (domain of expertise based on clinical guidelines and/or health datasets), a Reasoning Engine (cloud based organization of evidence relative to the patient’s electronic health record (EHR)), and a User-Interface (electronic display that provides alerts and recommendations in response to clinician’s queries) [13, 27]. CDSS model designs however can differ; utilizing standardized clinical guidelines (i.e., DSM-V or ICD-11), and/or aggregated health datasets [27]. A CDSS that uses guidelines translates the guidelines into web ontology language (OWL) modeled computerized-guidelines as the base for clinical guidance. CDSSs that incorporate health datasets, also known as “data-driven” CDSSs, use algorithms and data analytics to identify patterns of care trajectories that are relevant to the patient being seen [28‐30]. While both approaches have relative strengths, a CDSS model that is both guideline based and data-driven also allows the clinician to receive evidence-based support that is customized to each individual patient, based on extensive evidence and previous patient cases [27‐30].

While CDSS implementation in general medicine has been found to be advantageous for the quality of services [30, 31], there is limited evidence on the use of CDSSs for CAMHS [32]. One reason for such limited research is the lack of accessibility to psychiatric data, especially those pertaining to children and adolescents. From the established studies on CDSSs in CAMHS however, most have identified various limitations, including a lack of integration with the EHR, limited predictive models for diagnosis, and inadequate inclusion of multiple disorders [32]. Norway is in a unique position to develop and implement a CDSS for CAMHS because it can potentially avoid previously identified limitations. Over the past decade, the Norwegian National Association of Child and Adolescent Mental Health (N-BUP), has collected extensive data on CAMHS and compiled unique and rich datasets, based on advanced, multidisciplinary EHRs. However, these datasets have yet to be optimally utilized for subsequent care, as maintaining access requires ethical clearance and completion of a local data protection impact assessment (DPIA) and risk assessment to ensure privacy and confidentiality will not be breached.

This paper presents the study design and protocol for the evaluation of the first CDSS for CAMHS in Norway, the Individualized Digital Decision Assist System (IDDEAS). IDDEAS is an innovative research project, contributing to the HELSEVEL Programme (Programme on Health, Care and Welfare Services Research) and funded by the Norwegian Research Council (NFR). The overall aim of IDDEAS is to improve patient care by providing practitioners with real-time, data-driven and evidence-based guidance that ensures earlier and more precise decision-making, avoids misdiagnosis and reduces inefficiency in patient care practices. The principal benefit of the study is to examine the provision of decision support for clinicians in real-time during clinical sessions and improvement of clinical care.

The objectives of the IDDEAS project include:

IDDEAS will utilize qualitative and quantitative methods to determine whether the implementation of a clinical guideline and data-driven CDSS model in CAMHS will improve the quality of diagnoses and treatment of child and adolescent mental health disorders. The inclusion of two phases of evaluation within the IDDEAS trial will allow for a thorough assessment of the specific aspects of the CDSS that are suitable and those that need to be addressed for improvement of clinical care based on clinician’s subjective perception, as well as objective statistical analysis.

IDDEAS’ expected impacts include, but are not limited to, improved use of available health data, integration of predictive interventions in care, better management of complex clinical situations in child and adolescent mental health, enhanced collaboration among stakeholders involved in patient care, and amplified inter-service coordination in management of patient’s health. Furthermore, IDDEAS will help to improve patients’ understanding of the clinical decision- making involved in their treatment, as well as help ensure greater participation and improved transparency for the patients’ families. Overall IDDEAS will improve the quality of mental health care for children and adolescents by using innovative data-analytics within the CDSS model to provide support to clinicians that is evidence-based and customized to meet the needs of each individual patient. With quality improvement of clinical standards and enhanced precision of care, IDDEAS will also improve the efficiency of practices, allowing for more patients to be seen and receive a diagnosis in a timely-manner.

Our research will provide a comprehensive understanding of how to approach the use of a CDSS as a novel innovation for CAMHS. All limitations and challenges faced through the processes of designing, modelling, evaluating and testing the implemented IDDEAS CDSS model, can serve as valuable learning opportunities for future CDSS implementations in CAMHS.