The therapeutic alliance was defined by Bordin and colleagues in 1979 as a measure of the quality of the relationship between a patient and health care provider, and its pivotal influence on clinical outcomes is well documented and supported [1-3]. The following three main components constitute the therapeutic alliance: (1) the patient-provider bond, (2) a collaborative approach to assignment of therapeutic tasks, and (3) agreement on therapeutic goals [1]. In face-to-face therapy, the quality of this working alliance predicts clinical outcomes independently of the type of implemented psychotherapy or assessed outcome measures [2-6]. Several studies suggest that there is an association between a positive therapeutic alliance and improved outcomes in people with mental illness, including schizophrenia. For example, there is evidence that maintaining a positive alliance is associated with adherence to schizophrenia medication [7,8], that a positive alliance predicts improved overall psychotic symptomatic outcomes [9,10], and that strong alliances are associated with better treatment engagement and improvements in schizophrenia symptoms [11].

Schizophrenia is a chronic debilitating psychiatric disorder that occurs in approximately 1 in every 300 people worldwide [12]. Symptoms generally emerge in late adolescence or early adulthood and are categorized as positive, negative, or cognitive [13,14]. Positive symptoms, such as delusions, hallucinations and disorganized thinking, represent an excess or distortion of normal functions, and tend to relapse and remit. Negative symptoms involve reductions or absences in normal emotional and social functioning. Cognitive symptoms affect memory, attention and executive function. Both negative and cognitive symptoms being chronic in nature contribute substantially to the significant emotional and socioeconomic burden associated with schizophrenia [15-18]. Experiential negative symptoms (ENS) are a key subset of negative symptoms characterized by deficits in pleasure (anhedonia), motivation (avolition), and social interest (asociality), and have a profound impact on functional outcomes [19,20].

The American Psychiatric Association practice guidelines for the treatment of people with schizophrenia highlight the benefits of evidence-based psychosocial treatments, social skills training and psychoeducation as adjunct therapies to medication [21]. To date there is no treatment approved by the US Food and Drug Administration (FDA) for negative symptoms. Despite promising evidence for psychological approaches including cognitive behavioral therapy and cognitive remediation in treating negative symptoms [22-24], people with schizophrenia face significant barriers to engaging in psychosocial therapy. These include symptom-related barriers, stigma and social isolation, and insufficient health care resources [25-27], all of which may contribute to a low uptake of psychosocial therapy in people with schizophrenia [27-30].

Digital therapeutics (DTx) have the potential to bridge this gap by broadening access to adjunctive psychosocial therapies and facilitating greater frequency of exposure in a way that is flexible and convenient for those receiving treatment [31,32]. DTx are defined by the International Organization for Standardization as any evidence-based “health software intended to treat or alleviate a disease, disorder, condition, or that has a demonstrable positive therapeutic impact on a patient’s health” [33]. For DTx intended as treatments for vulnerable populations such as those with serious mental illness, it is particularly important that they are regulated, rigorously tested, and found to have proven effectiveness and strong safety profiles. Recent studies evaluating novel DTx in the form of smartphone apps have shown this modality to be acceptable and feasible to implement in individuals with chronic psychiatric disorders such as schizophrenia [34-37]. While there is evidence to suggest that the therapeutic alliance can be established and maintained in remote online settings of psychotherapy [4,38,39] it is not fully understood if individuals with severe mental illness, such as schizophrenia, can establish and maintain a therapeutic alliance with a DTx [40-43]. Importantly, while recent studies in people with schizophrenia evaluated the feasibility of DTx they did not assess the establishment of a therapeutic alliance [34,35,44].

We conducted 2 independent studies to assess the establishment and maintenance of a digital working alliance (DWA) between adults with ENS of schizophrenia and a beta version of CT-155/BI 3972080 (hereafter referred to as CT-155 beta), a novel software accessible on mobile phones under development to treat ENS of schizophrenia adjunctive to standard of care. In December 2023, CT-155/BI 3972080 was granted breakthrough device designation by the FDA [45]. In the studies reported here, we assessed whether it is feasible to establish and maintain a positive DWA between people living with ENS of schizophrenia and CT-155 beta, and if the strength of the DWA correlates or differs according to severity of negative symptoms, age, or race.

Our findings from 2 independent single-arm multicenter exploratory studies show that a DWA can be established and maintained between people with schizophrenia and a smartphone-based digital therapeutic, namely the abbreviated beta version of CT-155/BI 3972080, a prescription digital therapeutic to target ENS adjunctive to standard of care treatments for schizophrenia. During the orientation phase in study 1, participants established a positive DWA with CT-155 beta. This was replicated in a separate and independent cohort of participants at the end of an equivalent orientation phase in study 2. Furthermore, the positive DWA was maintained during the additional 4-week active phase in study 2, as demonstrated by an equivalent positive mARM score at week 7.

While a strong therapeutic alliance is linked to successful clinical outcomes for in-person therapy, the relationship between DWA, engagement, and outcomes requires further investigation [2-6]. The concept of a therapeutic alliance with DTx is gaining traction in line with the rapid expansion of digital mental health research and implementation; however, the role and relevance of the bond, a core feature of the in-person therapeutic alliance, is somewhat unclear for DTx [40] and not investigated in sufficient detail. Our results add to a growing body of evidence demonstrating that a therapeutic alliance can be established between people with serious mental illnesses and DTx. For example, in recent studies evaluating the benefits of conversational agents (or chatbots) that deliver cognitive behavioral therapy for symptoms of depression, anxiety, and substance use disorder, participants reported high acceptability and affective bond formation [64-67]. Another investigation of a digital health intervention for people with early psychosis, based on cognitive behavioral therapy principles and delivered through a smartphone app, demonstrated that engagement was significantly and positively associated with therapeutic alliance, as assessed by mARM [68]. However, measures to assess bond formation varied with a short-revised version of the Working Alliance Inventory and a Working Alliance Questionnaire (based on the Working Alliance Inventory) being used, making comparisons across studies difficult.

Although there are several lines of evidence demonstrating that a therapeutic alliance is possible for people with schizophrenia and is associated with improved outcomes, there are some factors related to the disorder itself, such as negative symptoms, lack of insight and self-stigma, that can adversely impact the therapeutic alliance between patient and therapist [69-71]. One could therefore speculate that patients who formed and valued a therapeutic alliance with a DTx may accept this additional therapeutic modality for the treatment of schizophrenia. Notably, in our studies, there was no correlation between the severity of negative symptoms at baseline and the establishment or maintenance of a DWA, indicating that even the participants with the most severe symptoms could form and maintain an alliance with CT-155 beta. In line with previous reports examining associations between age and therapeutic alliance we also did not observe any correlation between age and the DWA strength [11,39,71,72]. Similarly, we did not observe any correlation between race and strength of DWA. While some reports emphasize the importance of ethnic matching between patients and clinicians in improving therapeutic alliance [73,74], the lack of correlation with race here highlights the adaptable nature of a DTx modality to fit the needs of different communities.

Although not specifically assessed, the user-centered design process that guided the development of CT-155/BI 3972080 could have improved overall participant engagement and effectiveness. A recent systematic review showed that making end-users central to the product development process leads to enhanced cultural sensitivity, increased acceptance and engagement [75-77]. The review also provided some key recommendations for coproduction, including stakeholder involvement at all stages of development, which was implemented into the development of CT-155/BI 3972080. Considering the 3 pillars of the therapeutic alliance involve enhancing the patient-provider bond, shared goals and adopting a collaborative approach in the therapeutic approach, the patient-centered development process may have contributed to the establishment of a positive DWA achieved here.

The participants in both studies demonstrated a reasonable level of mobile device proficiency. This finding is aligned with recent data from the United States showing that 70%-80% among people with severe mental illness own a smartphone [78-80]. However, smartphone ownership does not necessarily equate with digital literacy [81]. In this study we confirmed that participating people with schizophrenia owning a smartphone showed digital literacy similar to that from the general population [82], but greater than that reported for people with serious mental illness [83]. Indeed, a study in the United Kingdom showed that 42.2% of people with serious mental illness had no foundation digital skills [83], while electronic health literacy has been described as either moderate or low in 2 independent cohorts of adults with schizophrenia spectrum disorders [84]. However, one should consider that the younger generation diagnosed with severe mental illness and schizophrenia are now digital natives and past research may not reflect current trends.

Limitations of these studies include their small sample size, single-arm design, and exploratory nature. While the sample size was limited, the demographics of participants were reflective of the general schizophrenia population including a higher incidence in males [85,86], the majority being Black or African American [87] and a level of education no greater than high school [88]. In addition, almost all (94%) participants in this study reported an annual income under US $25,000, which likely reflects low employment rates in people with schizophrenia [16]. Furthermore, while all participants met inclusion criteria, there was a difference in median MAP-SR score at screening between study 1 and 2. However, as no correlation between ENS severity and DWA was observed in both studies, the difference in baseline MAP-SR scores could be linked to broader population compared in 2 studies. While only a few out of protocol window deviations were recorded, they may impact the DWA assessments over time through asking questions and rapport-building. Preliminary data from these studies support the further evaluation of CT-155 in an ongoing large phase III trial (CONVOKE; NCT05838625). Although the mARM assessment possesses good face and content validity [59], further validation of this assessment for the measurement of DWA in severe chronic mental illness is essential to inform further optimization of this assessment for use in clinical trials [89]. In addition, establishment and assessment of therapeutic alliance between participants and digital devices or apps is relatively new. There is a need to use standardized and validated scales with psychometric validation to assess DWA in future studies which will also make it feasible to compare data and outcomes. Finally, it should also be noted that the eligibility criteria for the studies mandated that participants had a smartphone, an email address, and regular access to the internet through mobile data plan, Wi-Fi or both. Accordingly, the association of digital literacy measured by MDPQ with DWA and engagement should be interpreted in the context of this study population. The study inclusion criteria for participants to own their own mobile device may have biased toward a more digital proficient study population and therefore not reflective of patients who do not own a mobile device and who potentially may have a lower device proficiency. As such, future studies are warranted to understand the impact of lower digital proficiency on adherence to a digital therapeutic and on establishing DWA. While recent evidence suggests that rates of smartphone ownership among adults with severe mental illness (including schizophrenia) are comparable to the general adult US population [78,80], it is encouraging to see high levels of mobile device proficiency in people living with schizophrenia and owning smartphones.

The results from these 2 studies confirm that a positive DWA can be formed within a 3-week period and this DWA was maintained for the entire 7-week duration of study 2, demonstrating that a therapeutic alliance can be formed with a DTx in people living with schizophrenia. Importantly, the formation of a positive DWA was not limited by severe ENS, supporting its potential benefits for those experiencing severe symptoms. Furthermore, people with schizophrenia who participated in these studies demonstrate good digital literacy and acceptance of CT-155 beta for the treatment of ENS. The development of a DTx with proven acceptability, safety, and effectiveness profiles as an adjunctive treatment for ENS of schizophrenia would represent a promising, accessible, scalable, and affordable treatment option for clinical and patient communities for whom the lack of effective treatments for ENS has presented significant challenges. These preliminary data support further evaluation of the potential importance of the patient-CT-155 DWA to therapeutic effectiveness in an ongoing large phase III trial (CONVOKE; NCT05838625). With the expansion of mobile device ownership, DTx offers a potential patient-centered, evidence-based therapeutic support 24 hours a day, 7 days a week. In addition, incorporated into standard of care, evidence based DTx may act as clinician extenders, increasing the breadth and depth of care provided. Understanding how best to engage and optimally implement DTx with people living with schizophrenia to have the greatest reach and impact is of great importance and may shape future clinical research. In addition, understanding who may benefit from a DTx, how to best introduce these tools to patients as part of the clinical workflow, and how to support continued engagement, are important future steps.