Introduction

Schizophrenia is a complex and disabling psychiatric disorder that affects approximately 0.45% of adults worldwide1 and 0.25–64% of adults in the US2,3,4, imposing a substantial health, social, and economic burden5,6. Furthermore, schizophrenia disproportionately impacts marginalized communities, a pattern attributed not to increased biological vulnerability but to a combination of social determinants, structural inequalities, and potential diagnostic bias7,8. Symptoms of schizophrenia are described in terms of positive, negative, and cognitive symptoms, and 50–60% of people living with schizophrenia experience at least one negative symptom9,10. Although antipsychotics have been shown to reduce positive symptoms (i.e., delusions and hallucinations) and can also improve secondary negative symptoms11 (e.g., blunted affect arising from the emotional distress of delusions or hallucinations), enduring primary negative symptoms typically do not respond to currently available antipsychotics12. While several antipsychotics are approved by the Food and Drug Administration (FDA) for the treatment of schizophrenia, none are specifically approved to target cognitive and negative symptoms, which remain inadequately addressed by current pharmacological options13,14.

Negative symptoms can be subdivided into expressive negative symptoms (blunted affect and alogia)15,16,17,18, and experiential negative symptoms (ENS; avolition, anhedonia, and asociality)15,16,18,19. ENS have been associated with poorer outcomes than expressive negative symptoms, including significantly greater conceptual disorganization and psychosis, increased likelihood of hospitalization, poorer social functioning, greater social cognitive impairment, greater likelihood of unemployment20, fewer hours worked, and a reduction in wages earned20,21. Furthermore, people with schizophrenia who have elevated ENS (i.e., those with higher asociality, avolition, and/or anhedonia) report reduced quality of life compared to those with less severe ENS22.

Individuals with negative symptoms often exhibit impaired positive reward learning and heightened defeatist performance beliefs23,24. Research has shown that engaging individuals in rewarding activities and value-based goal setting has shown significant reductions in avolition-apathy negative symptoms and improvement in global functioning vs treatment as usual25,26,27. Furthermore, cognitive behavioral therapy (CBT), along with other psychosocial treatments including social skills training and cognitive remediation, are recommended in clinical guidelines as effective strategies to augment other treatments, and may all improve ENS12,28,29,30,31,32. While psychosocial approaches have proven effective in managing negative symptoms, their accessibility and uptake can be limited due to insufficient resources (e.g., a lack of trained professionals to deliver these interventions) and poor healthcare resource utilization in general33,34. Furthermore, negative symptoms often predict poor engagement with clinical services, creating an additional barrier to treatment35. These barriers to psychosocial accessibility and uptake, as well as the substantial burden of ENS on people living with schizophrenia, highlight the need for new, effective, and accessible therapies. These new treatments could significantly enhance current care options and improve the overall continuum of support for patients, ultimately leading to better health outcomes and quality of life.

Digital therapeutics (DTx), defined by the International Organisation for Standardization as “health software intended to treat or alleviate a disease, disorder, condition, or that has a demonstrable positive therapeutic impact on a patient’s health,”36 may address this unmet need. DTx can be delivered remotely to augment existing mental health services in a way that is flexible and convenient for patients, and therefore has the potential to broaden access to evidence-based psychosocial therapies37,38. Indeed, DTx has demonstrated effectiveness in difficult-to-treat populations, including people with post-traumatic stress disorder39,40, substance use disorder41,42, and attention deficit hyperactivity disorder43,44. Although DTx for people living with schizophrenia is more limited, promising results have emerged45,46,47,48,49. For example, a study demonstrated significant reductions in ENS and defeatist beliefs in a cohort of 31 patients with psychosis after 12 weeks of mobile-assisted CBT48. In another study, a DTx designed to improve motivational impairments in patients with early psychosis led to significant improvements in defeatist beliefs, motivation, and self-efficacy49.

CT-155/BI 3972080 (CT-155), is being developed as a smartphone-based prescription DTx that provides psychosocial intervention techniques adjunctive to standard of care treatments for patients 18 years of age and older with ENS of schizophrenia. The CT-155 program has been granted Breakthrough Device designation by the FDA50,51. CT-155 was designed in line with FDA recommendations that emphasize the importance of incorporating patient experience data into the development process52, by taking into consideration the unique characteristics and needs of the schizophrenia population and integrating feedback from individuals with lived experience of schizophrenia53. In this exploratory study, we examined the engagement, adherence, within-group effectiveness, acceptability, user experience, and safety of a beta version of CT-155 (CT-155 beta; an abbreviated version of CT-155/BI 3972080).

Results

Demographics and baseline clinical characteristics

Seventy-four individuals were screened, of whom 50 were enrolled (32% screen fail rate) and 43 (86%) completed the study (Fig. 1). The median age of the cohort was 53.5 years (range: 23–64), and the majority (n = 40, 80%) were male. Twenty-nine (58%) participants were Black or African American, 32 (64%) participants did not have a college-level education, and 47 (94%) participants had a self-reported annual income of <$25,000. Complete sociodemographic characteristics are listed in Table 1, and clinical characteristics are listed in Table 2. The mean (standard deviation [SD]) Motivation and Pleasure scale – Self Report (MAP-SR) score at screening was 14.9 (8.9) points, and the mean Clinical Assessment Interview for Negative Symptoms Motivation and Pleasure scale (CAINS-MAP) total score at baseline was 20.5 (8.3) points. Most participants demonstrated proficiency in the use of mobile devices prior to accessing CT-155 beta, as indicated by the median (interquartile range [IQR]) total mobile device proficiency questionnaire (MDPQ) score at baseline of 30.1 (21.4, 35.5) points. Concomitant use of second-generation antipsychotics was reported by all 50 participants (n = 44 receiving 1 antipsychotic; n = 6 receiving 2 antipsychotics) (Table 2 and Supplementary Table 1).

Fig. 1: Participant disposition.
figure 1

PTSD post-traumatic stress disorder, MAP-SR motivation and pleasure scale—self report. Figure was adapted from Snipes et al JMIR Mental Health 2025 licensed under Creative Commons Attribution CC-BY 4.053.

Full size image
Table 1 Baseline sociodemographic characteristics.
Full size table
Table 2 Clinical characteristics.
Full size table

Participant engagement and adherence with CT-155 beta

Participants’ engagement with CT-155 beta was measured passively throughout the intervention period by the app. Engagement metrics included daily app opens, daily check-ins completed, and app session length. In addition, adherence metrics included the proportion of therapeutic lessons and the number of therapeutic goals completed.

Kaplan–Meier retention analysis shows that the majority of the participants (84%; N = 42/50) engaged with CT-155 beta (i.e., last app open) until the end of the study period (Day 49; Fig. 2a). The median (IQR) daily app check-ins were completed on 43.0 (19–47) days of the 49 possible days (88%) (Fig. 2b). In terms of time spent in the app, on average participants engaged with CT-155 beta for 11.6 (8.1–16.1) min per session (Fig. 2b).

Fig. 2: Participant engagement and adherence with CT-155 beta.
figure 2

a Kaplan–Meier curve showing participant retention (last app open) (b) summary of participant engagement metrics and c summary of adherence metrics. 1Forty-nine total possible days; 2Twenty-one total therapeutic lessons (18 required core and 3 bonus lessons); and 3defined as completing an activity or reporting the completion of an activity in CT-155 beta. The time (days) to when participants stopped engaging with CT-155 beta during the engagement period (n = 50) was analyzedusing a Kaplan–Meier survival plot. If a participant skipped some days, but returned to continue with activities, then the last day of their engagement was considered the day that they stopped engaging with CT-155 beta. Box plots depict the median and IQR of participant engagement metrics; the median is represented by a horizontal line in each box. The whiskers depict the range within 1.5 times the IQR, values outside this range are considered outliers. The red dashed line indicates the maximal attainable number where applicable.

Full size image

Overall, there was no significant correlation between baseline CAINS-MAP score and engagement with CT-155 beta (p > 0.05) (Supplementary Table 2). In addition, daily app engagement with sessions of 60 s or longer was also analyzed to further characterize meaningful user engagement with the study app (Supplementary Fig. 1).

In terms of adherence to therapeutic lessons in CT-155 beta, participants completed a median (IQR) of 18 (13–20) therapeutic lessons out of the total of 21 available. Furthermore, during the active phase, the median (IQR) number of goal-centered activities completed was 15 (4–23; Fig. 2c) out of 24 available, and a median (IQR) of 4 (0–5) goals was completed per participant. This indicates on average, 1 goal per week was completed.

Negative symptoms

Severity of participants’ ENS of schizophrenia was evaluated by the CAINS-MAP assessment measure15. For those participants with baseline and Week 7 assessments (n = 44) baseline mean CAINS-MAP total score was 20.4 (8.6) points. At the end of the intervention period (Week 7), the mean CAINS-MAP total score was 16.8 (7.7) points (Fig. 3). This reflects an average within subject reduction of 17.6% in the mean (95% confidence interval [CI]) CAINS-MAP total score of 3.6 (1.3, 5.8) points from baseline (t[43] = 3.2; p = 0.0026; n = 44).

Fig. 3: Individual and mean (violin plot) CAINS-MAP scores at baseline and Week 71.
figure 3

1Twenty-nine of the 44 participants included in the CAINS-MAP analysis were from a single study site. Within-subject changes in CAINS-MAP scores at baseline were compared with those at Week 7 using a Student’s paired t-test. CAINS-MAP clinical assessment interview for negative symptoms, motivation, and pleasure subscale; ENS experiential negative symptoms.

Full size image

Notably, the improvement in CAINS-MAP total at Week 7 was driven by a change in two sub-domains: (i) social subscale score (the average Likert score out of 4) (median [IQR] −0.38 [–0.88, 0.00]; p = 0.0020) and (ii) recreational subscale score (−0.67 [−1.33, 0.00]; p = 0.0001). No significant change from baseline was observed in the Work & School subscale score (0.00 [−0.50, 0.50]; p = 0.5873) at Week 7 (Table 3).

Table 3 Change in CAINS-MAP subscale scores from baseline visit to Week 7.
Full size table

Acceptability and user experience of CT-155 beta

Participants’ rating of study app quality and satisfaction was evaluated at the end of the study using a validated Mobile App Rating Scale (MARS) assessment. Participants rated the quality of CT-155 beta as either acceptable or good across the four key domains of the assessment—engagement, functionality, esthetics, and information quality (Fig. 4 and Supplementary Table 3). Across these four domains, the mean MARS scores ranged from 3.8 (0.8) to 4.2 (0.8) points out of 5 (“excellent”) on the Likert scale for the engagement and functionality domains, respectively. The subjective quality domain, asking about recommendation of the app, expected frequency of usage over the next 12 months, willingness to pay for the app, and star rating, was scored in the average range (mean score: 3.7 [0.8]; Fig. 4 and Supplementary Table 3). Of the 43 participants who responded, the majority of them scored the app-specific items as either 4 or 5 on the Likert scale indicating that they slightly or strongly agreed; knowledge (32/43), attitudes (29/43), awareness (29/43), intention to change (26/43), help seeking (27/43), and behavior change (26/43) (Supplementary Fig. 2 and Supplementary Table 3).

Fig. 4: Modified* MARS domain scores at Week 71.
figure 4

1n = 43 participants with observed data. *Two items were omitted from the information domain due to a lack of relevance. Individual MARS subscale scores are shown, and ‘X’ indicates the mean. Box plots depict the median and IQR of MARS subscale scores; the median is represented by a horizontal line in each box. The whiskers depict the range within 1.5 times the IQR, values outside this range are considered outliers. MARS includes several domains: Engagement (items 1–5), functionality (items 6–9), esthetics (items 10–12), and information quality (items 14–18). Items were rated as either 1 (inadequate), 2 (poor), 3 (acceptable), 4 (good), or 5 (excellent), and statistically summarized per domain. The MARS also includes a subjective quality domain (items 20–23), which is scored separately using a 5-point Likert scale, with higher scores indicating better quality. The MARS app-specific items were scored as either 1 (strongly disagree), 2 (slightly disagree), 3 (neutral), 4 (slightly agree), or 5 (strongly agree). MARS mobile app rating scale.

Full size image

In addition, participants’ subjective experiences of using CT-155 beta were gathered through a brief user experience survey administered at Weeks 3, 5, and 7. Most participants reported using CT-155 beta either once a day or multiple times a day consistently over the 7-week intervention period (Week 3: 91.1% [41/45], Week 5: 91.1% [41/45]; Week 7: 84.1% [37/44]) and 95.5% would recommend the app to friends by the end of the study (Table 4). For free-text response questions, participants reported that CT-155 beta helped to stabilize their mood, improve their confidence, and helped them to change their perspective/outlook on life (Table 5). Overall, only a small proportion of participants provided suggestions for improving CT-155 beta, such as adding more content (~11–14%), increasing interactivity (~7%), and modifying videos (~2–7%; Table 5). Finally, feedback from the end-of-study user experience survey indicated that 89% (39/44) of participants agreed or strongly agreed that the study app was an effective treatment for their negative symptoms.

Table 4 User experience survey.
Full size table
Table 5 Participants’ subjective experiences of using CT-155 beta (free text responses).
Full size table

Safety

Three adverse events (AEs) were reported by 2 participants during the study, none of which were considered severe or related to CT-155 beta, nor led to discontinuation of the study. AEs reported were a sinus infection (n = 1), arthralgia (n = 1), and a rash (n = 1).

Discussion

This exploratory study aimed to evaluate the feasibility of an investigational digital therapeutic app (CT-155 beta) as a potential treatment of ENS of schizophrenia, in terms of engagement, adherence, effectiveness, acceptability, user experience, and safety. The study shows that participants living with schizophrenia with clinically assessed negative symptoms showed high levels of engagement, adherence, and acceptance over the 7-week intervention period with minimal risk to users. Moreover, results showed a within-group improvement in negative symptoms at the end of the study (Week 7) compared with baseline. Notably, participant engagement with the study app was not associated with negative symptom severity at baseline, suggesting that individuals would use CT-155 beta regardless of deficits in motivation and goal-directed behavior.

Participants demonstrated a high level of engagement, with 84% continuing to engage with CT-155 beta until the last day of the study. While engagement with DTx varies substantially across disease areas, our findings show that engagement with CT-155 beta was high in relation to other DTx studies, which typically range from 21% to 88%54. Furthermore, participant engagement and study completion rates reported in this study were higher than previous studies examining DTx for people with schizophrenia46. For example, 1 study examining smartphone delivery of CBT-based coaching for negative symptoms in people with schizophrenia initially reported acceptable DTx engagement and completion rates (e.g., above 80%), but then dropped by almost half after the first 4 weeks of the study49. Another DTx delivering CBT for negative symptoms found engagement rates as low as 19% by the end of the intervention period (24 weeks)48. Furthermore, no correlation was observed between multiple measures of engagement and the severity of ENS. This demonstrates that even at severe levels of ENS, where deficits in motivation and goal-directed behavior pose especially large barriers for accessing care, DTx, may be meaningful additions to treatment plans for those with schizophrenia. This high level of participant engagement with the study app reflects the end-of-study feedback in which participants found CT-155 beta engaging, visually appealing, and rich in high-quality content. Additionally, user experience surveys showed positive feedback, with 84% of participants stating they used CT-155 beta once or multiple times a day by Week 7. Furthermore, by the end of the 7-week intervention period, 95.5% of participants indicated they would recommend CT-155 beta to a friend. It is possible that aspects of the therapuetic played a role in driving engagement. Specifically, the CT-155 beta intervention was pointedly developed to establish and maintain a digital working alliance, akin to the therapeutic alliance in face-to-face treatment, which is linked with improved clinical outcomes in mental health55,56,57. The development of CT-155 beta was informed by an iterative patient-centered design process, which consisted of 2 phases during which feedback was captured from 4 peer support specialists with lived experiences of schizophrenia, and 15 patient panel participants with a diagnosis of schizophrenia53. Additionally, through an empathic style of communication, CT-155 beta provides core features of in-person therapeutic alliance to encourage daily adherence to lessons and activities by providing schizophrenia-specific psychoeducation and introducing core therapeutic skills. Like therapists getting to know their patient, CT-155 beta focuses on digital working alliance establishment through asking questions and rapport building, followed by a collaborative treatment approach and goal definition. These patient-centric features that support digital working alliance are subsequently integrated into all therapeutic components to optimize the experience by infusing an interactive, empathic, knowledgeable, and personally meaningful nature to the in-app experience that builds confidence for achieving treatment goals and is critical to treatment outcome. Findings from two single-arm, multicenter, exploratory studies show that a digital working alliance can be established and maintained between people with schizophrenia and CT-155 beta in independent cohorts53. Importantly, the formation of a positive digital working alliance was not limited by severe ENS, which supports the potential benefits of digital therapy for people experiencing severe symptoms. Furthermore, CT-155 beta was intended to be engaged with in small daily increments, as compared to the longer sessions (e.g., 20–50 min, once per week or less) that typically characterize face-to-face treatment. It may be that this type of daily engagement was particularly acceptable to this patient population. Furthermore, daily tasks were personalized based on participants’ feedback on how they were feeling on a particular day. For example, if a participant’s mood was low on a particular day, simpler tasks were suggested. Additionally, participants were permitted to skip a day during periods of low mood. The CT-155 beta app design may have contributed to engagement, as each goal was divided into four sequential steps, with a minimum of one day allocated for the completion of each step. As a result, regular engagement with the app was essential for progressing through the treatment. Finally, participants also received daily SMS from CT-155 beta as reminders to engage with the treatment during the intervention period. Overall, the high level of engagement observed in this study provides further evidence to suggest DTx can be feasible for individuals with mental illnesses, including schizophrenia45,58,59,60. However, our engagement data should be interpreted in the context of a clinical study environment where high engagement could be linked to the motivation of willing participants61,62. Further investigation of engagement with DTx in naturalistic conditions for the target population could be assessed through evidence collected through real-world studies.

In this study population, comparing ENS severity using CAINS-MAP from baseline to end of study (Week 7) across subjects showed notable improvement (17.6%). Previous research has reported that a CAINS-MAP total score of 17 or greater classifies negative symptoms as severe, persistent, and treatment resistant63, suggesting the improvement in negative symptoms to <17 in our study may represent a meaningful clinical improvement. A similar magnitude of improvement (3.5 point reduction in CAINS-MAP) was described in a previous study using mobile-assisted CBT intervention48. However, the within-subject change in both studies should be interpreted with caution, given the single-arm study design, but warrants further investigation. Furthermore, in the present study, the overall reduction in the CAINS-MAP total score was primarily driven by a change in 2 of the 3 sub-domains: social and recreational. There was no significant change observed in the work and school sub-domain, which is not surprising as changes in work and school situations are unlikely to happen in the short timeframe of the study. Additionally, the psychosocial interventions in CT-155 beta are designed to enhance goal-directed behavior, which is known to improve social and recreational skills64,65.

In this study, 24 out of 74 (32%) participants did not meet the strict eligibility criteria, particularly the moderate-to-severe ENS criterion (score of ≤30 on the MAP-SR), accounting for 12 of the 24 (50%) of screen failures. Screen fail rates vary across DTx studies in schizophrenia, with rates typically ranging from 14% to 66%47,48,49,58,66,67. It is important to note that study eligibility criteria mitigated potential confounding variables that may affect study assessments. For instance, individuals were excluded from participation if they were treated with more than 2 antipsychotic medications, had changed dose(s) of antipsychotic medication(s) within the last 3 months, were currently receiving inpatient treatment or psychotherapy, or had received inpatient treatment or psychotherapy within the last 3 months. Additionally, any beneficial effect of concomitant antipsychotics on ENS would presumably already have been observed at baseline. Therefore, any observed improvement in the severity of negative symptoms throughout the 7-week study is likely indicative of treatment benefit from CT‑155 beta. As clinical trials of DTx targeting ENS in people living with schizophrenia are uncommon, this study makes an important contribution by demonstrating improvements in ENS as a primary target rather than secondary improvements related to changes in positive symptoms.

It is important that the findings from this study should be interpreted in the context of its potential limitations. This was an exploratory single-arm study to test the feasibility of a beta version of the CT-155 app in a target population limited to 50 participants and not powered to detect improvement in ENS severity. Given the single-arm study design, within-subject change in ENS severity should be interpreted with caution and requires validation in a larger controlled study. The standard rigor of Phase 3 clinical study design includes but is not limited to (i) use of an appropriate control arm such as treatment-as-usual, waitlist control, or a digital control app68, (ii) appropriate blinding of the involved study and clinical site staff, and (iii) safety endpoints. These study design elements will allow DTx trials to be held to the same evidentiary standards as pharmacotherapy and strengthen study rigor. Importantly, in this exploratory single-arm study, a placebo effect due to spontaneous changes of the condition, expectancy beliefs of the patient, or the psychological and supportive benefits experienced during the intervention, cannot be excluded as possible reasons for observing potential effectiveness. The CONVOKE registrational study (NCT05838625) will provide a robust assessment to validate these preliminary effectiveness data. While the high patient engagement and adherence observed in this study is promising, it remains to be seen whether engagement (measured as time spent in the app or frequency of opening the app) and adherence (measured as proportion of therapeutic lessons and number of therapeutic goals completed) translate into clinical effectiveness in larger, statistically powered randomized controlled trials. Finally, the study size was limited, and findings may not generalize to a broader population of people living with schizophrenia. However, the demographics of study participants were generally consistent with those reported in schizophrenia population studies, including a higher incidence in males69,70, a predominance of participants identifying as Black or African American—reflecting the higher reported diagnostic rates in this group7, and a majority having an education level of high school or below71. 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 schizophrenia72. Furthermore, the inclusion criteria for participants to own their own mobile device may have biased the sample toward a more digitally proficient population, and may therefore not be reflective of patients who do not own a mobile device and who potentially may have a lower device proficiency. However, this inclusion criterion was critical to ensure adequate fit between participant abilities (e.g., digital literacy) and engagement demands of CT-155 beta. Doing so ensured that findings were due to the intervention rather than user error, thereby strengthening the internal validity of the findings.

Despite these limitations, this study demonstrated the feasibility of CT-155 beta, highlighting the ability of patients with schizophrenia to engage with the digital intervention and an associated reduction in ENS severity from baseline. Findings from this exploratory study show the potential of CT-155 beta to be a viable adjunct to current treatment modalities in an area with a pressing need for new and effective therapies. The improvement in ENS observed after 7 weeks of CT-155 beta suggests that digital treatments may help alleviate the significant disease burden associated with ENS in schizophrenia. Although these findings are encouraging, they are exploratory, and a large Phase III study will validate these initial insights (CONVOKE). Nevertheless, the results contribute to a growing body of evidence that shows that people living with schizophrenia are not only capable and willing to use technology to support their treatment but also benefit from these opportunities. Given the urgent need for innovative, effective, and accessible treatments to enhance standard care and mental health services for schizophrenia, particularly for negative symptoms, evidence-based prescription DTx could provide a promising solution. By leveraging the widespread use of smartphones, DTx has the potential to deliver high-quality, standardized, and scalable interventions to individuals with schizophrenia.

Methods

Study design

This multicenter, single-arm, exploratory study (clinicaltrial.gov: NCT05486312) was conducted across 9 sites in the United States between July 2022 and November 2022. The study was approved by the Western Copernicus Group IRB (approval number: 20220609), with written informed consent obtained prior to participation. During an in-person screening visit (Visit 1), site personnel assisted with the download and set-up of CT-155 beta onto participants’ personal smartphone devices. CT-155 beta was activated following confirmation of eligibility during baseline visit (Fig. 5A). During the 7-week intervention period, participants attended in-person clinic visits at Week 3 (Visit 3), Week 5 (Visit 4), and Week 7 (Visit 5). Further details on data privacy can be found in the Supplementary Materials. Participants were offered up to US$735 for partaking in the study. Incentives were offered for attending study visits, but not for daily app usage.

Fig. 5: Study design and therapeutic journey.
figure 5

a Overview of study design (CT-155-C-002; NCT05486312) and b therapeutic journey of participants engaging with CT-155 beta. 1Participant engagement and adherence with the study app (CT‑155 beta) were measured passively through the app. Engagement was defined as the number of app opens and sessions of 60 s or greater. Adherence was defined as the number of lessons and goals completed. An independent single-arm, multicenter, exploratory study (CT-155-C-002/NCT05486312) was conducted across ten sites in the United States. The study comprised a screening period, an intervention period, and a follow-up period. The intervention period comprised a 3-week orientation phase with an additional 4-week active phase. The orientation phase included experiences to provide intervention orientation, digital working alliance establishment, and build core therapeutic skills necessary for later stages of intervention. Goals were comprised of 4 steps, where each step is allotted a day to complete. Therefore, 1 goal required a minimum of 4 days to complete. Goal-centered tasks were defined as completing an in-app activity or reporting the completion of the activity during the Activity check-in. CAINS-MAP clinical assessment interview for negative symptoms motivation and pleasure scale, CT-155 beta a beta version of CT-155/BI 3972080, DWA digital working alliance, ENS experiential negative symptoms, MAP-SR motivation and pleasure scale—self report, MARS mobile app rating scale, MDPQ mobile device proficiency questionnaire.

Full size image

Participants

Eligible participants were ≥18 years of age with a primary diagnosis of schizophrenia for at least 1 year (as per the Diagnostic and Statistical Manual of Mental Disorders, fifth edition [DSM-5]), and moderate-to-severe ENS (defined at screening as a score of ≤30 on the MAP-SR73). The MAP-SR is a 15-item self-report scale where participants are asked to rate their deficits in motivation and pleasure across social, recreational, or work domains. Items are rated on a 5-point Likert scale with lower scores indicating greater pathology. Additionally, participants must have been receiving a stable dose of a maximum of two different antipsychotic medications for ≥12 weeks prior to screening. Exclusion criteria included a DSM‑5 or International Classification of Diseases 10th edition diagnosis other than schizophrenia, and/or a substance or alcohol use disorder (excluding caffeine and nicotine). Participants receiving clozapine, haloperidol, or who had received psychotherapy within 12 weeks of screening were excluded. Finally, participants with active prominent positive symptoms that, in the opinion of the investigator, would preclude effective engagement in treatment for negative symptoms were also excluded. The complete list of eligibility criteria is included in the Supplementary Materials. A list of assessments conducted at screening and baseline are described in Supplementary Table 4.

Intervention

CT-155 is currently considered an investigational medical device per FDA regulations, and more specifically, a software-as-a-medical device to treat ENS in people living with schizophrenia50,74. Individual components of CT-155 were designed to incorporate principles of evidence-based, in-person psychosocial therapy to target ENS75,76,77,78. CT-155 beta was an abbreviated 7-week version of a full intervention experience with CT‑155. The design and development of CT-155 beta were informed by participant feedback gathered during interviews with people living with schizophrenia. Further details on the participant-centered design process have been described further in Snipes et al, JMIR Mental Health 202553.

The overall intervention period (7 weeks) comprised a 3-week orientation phase, a 4-week active phase, and concluded with a 1-week follow-up period during which a remote teleconference occurred (Fig. 5A). The orientation phase of the intervention included lessons and activities designed to establish a therapeutic alliance between participants and CT-155 beta (i.e., a digital working alliance53) and build core therapeutic skills (including distress tolerance) necessary for later stages of the intervention (Fig. 5B). Participants had access to a total of 21 therapeutic lessons (18 required core lessons and 3 optional lessons) in the orientation phase. The active phase included continuation of the orientation phase goals, with incorporated adaptive goal-setting activities, and additional therapeutic skills training (Fig. 5B). Participants were encouraged to complete lessons and activities in CT-155 beta daily and were asked to complete a daily check‑in. In this phase, therapeutic techniques are delivered daily to help participants set goals (adaptive goal setting) that encourage real-world engagement (behavioral activation) while removing barriers (cognitive restructuring) and providing skills to facilitate goal attainment (positive affect training). The adaptive goal-setting modules focused on delivering interventions known to target negative symptoms in important functional areas (i.e., social items, work and school items, recreation items). As part of adaptive goal setting, activities containing behavioral activation elements were adaptively set and gradually increased in difficulty as participants acquired skills. Participants received normalizing messages and cognitive or behavioral strategies to help manage negative experiences related to cognition, beliefs, emotions, and associated behaviors. Furthermore, the positive affect training component of CT-155 beta taught participants skills derived from positive psychology and affective science to target ENS by increasing positive affect.

Study assessments

Participant engagement and adherence with CT-155 beta

Participant engagement with CT‑155 beta was measured passively by the smartphone as a result of participants' day-to-day app usage and completion of in-app activities. These included the number of days that CT-155 beta was opened (any usage event recorded during a given day), the number of daily check-ins completed, and the duration of each CT-155 beta use session. Study retention was assessed using Kaplan–Meier retention analysis of last app open data until the end of the study period. Adherence was determined by assessing the number of available therapeutic lessons completed during the orientation phase and the number of goal-centered tasks completed and goals successfully accomplished during the active phase of the study.

Negative symptoms

The potential of CT-155 beta to reduce the ENS of schizophrenia was assessed by the change in the CAINS-MAP total score from baseline to Week 7. CAINS is a second-generation negative symptom scale15 that has demonstrated reliability, convergent and discriminant validity, and has a short test-retest administration time of 15–30 min15,79,80. CAINS was originally developed on the recommendations of an expert group convened by the National Institute of Mental Health and the Measurement and Treatment Research to Improve Cognition in Schizophrenia initiative to reconceptualize negative symptoms to include 5 domains: avolition, asociality, alogia, anhedonia, and blunted affect81. The CAINS consists of 2 subscales covering the 5 domains: Motivation and Pleasure (MAP; i.e., avolition, anhedonia, and asociality), and Expressivity (EXP; i.e., blunted affect, alogia) (Supplementary Table 5)15. CAINS offers a more detailed assessment of negative symptoms of schizophrenia compared with older scales such as the Positive and Negative Syndrome Scale (PANSS), which treats these symptoms as a more generalized construct16. It focuses on experiential deficits by using detailed probes to assess internal experiences such as joy, pleasure, interest, or intimacy, instead of simply relying on behavior. CAINS underwent multiple development and psychometric testing phases, where it has been validated against patient-reported experiences, in-the-moment experiences, and behavioral observations. CAINS-MAP total scores were grouped according to baseline ENS symptom severity. In the present study, CAINS-MAP was administered in-clinic by trained raters. In line with FDA's current best practices and terminology in medical device clinical research, improvement in clinical outcomes is herein referred to as effectiveness (rather than efficacy)82.

Acceptability and user experience of CT-155 beta

Acceptability and user experience of CT-155 beta were assessed using the validated MARS83 and a brief non-validated user experience survey, respectively. At the end of the study (Week 7), participants rated the quality of CT-155 beta through MARS (Supplementary Table 6) and shared their expectation of the therapeutic benefit of CT-155 beta via a brief 3-question survey. MARS items were rated from 1 (inadequate) to 5 (excellent), and a mean score per domain calculated83. In the user experience survey, participants were asked at Weeks 3, 5, and 7 how often they used CT-155 beta and whether they would recommend CT-155 beta to a friend. Three free-response questions covering the benefits of, frustrations with, and suggested improvements to CT-155 beta were also asked. A modified inductive approach was used to assess free-text responses with themes emerging directly from the data without predetermined categories.

Safety

Safety endpoints included the frequency and severity of all AEs, serious AEs, and discontinuations from the study due to AEs. These were collected from when informed consent was provided until the end of the follow-up period.

Statistical analysis

This was an exploratory study to assess participant engagement and adherence with CT-155 beta, along with potential effectiveness and user feedback to determine the acceptability of CT-155 beta. As hypothesis testing was not the objective of this study, no formal a priori power calculations were performed. All analyses were performed on the observed data, and no imputations were performed for missing data. Descriptive statistics were evaluated for demographic data and study variables in all participants who enrolled in the study. To examine overall retention in the study, we used the univariate Kaplan–Meier survival curve84. Within-subject changes in CAINS-MAP scores at baseline were compared with those at Week 7 using a Student’s paired t-test. For correlation analyses, a Spearman rank-order correlation coefficient was calculated together with the 95% two-sided CIs.