Preliminary Effectiveness of a Therapist-Supported Digital Mental Health Intervention in Reducing Suicidal Ideation

Abstract

Suicidal ideation (SI) is a significant public health concern with increasing prevalence. Therapist-supported digital mental health interventions (DMHI) are an emergent modality to address common mental health problems like depression and anxiety, although less is known about SI. This study examined SI trajectories among 778 patients who participated in a therapist-supported DMHI using multilevel models during and up to 6-months post-treatment. Estimates of associated suicide attempts and deaths by suicide were calculated using published data linking PHQ-9-assessed SI to records of suicide attempts and deaths by suicide. The proportion of participants reporting no SI significantly increased between baseline and end-of-treatment (78.02% to 91.00%). Effect sizes of SI changes between baseline and end-of-treatment, 3-month, and 6-month follow-ups were 0.33 (95%CI = 0.27–0.38), 0.32 (95%CI = 0.27–0.38), and 0.32 (95%CI = 0.27–0.38), respectively. Results also indicated an estimated 30.49% reduction (95%CI = 25.15%-35.13%) in suicide attempts and death by suicide across treatment. This study provides preliminary evidence of the effectiveness of a therapist-supported DMHI in reducing SI.

Keywords:

• Epidemiology
• mental health services
• suicide
• telemedicine

Suicide continues to be a leading cause of preventable mortality in the United States (US). In 2019, it was estimated that approximately 5% of the US population reported suicidal ideation (SI) in the past year (Stone et al., 2021). Importantly, SI is a strong predictor of future suicide behaviors, such that an estimated 12.5% to 29% of those with SI end up attempting to die by suicide (Nock et al., 2008). Unfortunately, the US is currently experiencing a mental health crisis characterized by large-scale structural barriers (e.g., long waitlists, high costs, transportation issues, lack of trained providers) to delivering effective suicide prevention and treatment interventions (Moutier, 2021). As such, it is estimated that only 44% of people experiencing SI and 61% of those that have attempted suicide seek mental health treatment (Ahmedani et al., 2012; Kessler et al., 2005), pointing to a dire need for novel, scalable, and cost-effective solutions to reduce SI and associated health outcomes at the population level.

Digital mental health interventions (DMHI) offer one possible solution to address SI and associated mental health problems (Büscher et al., 2020; Torok et al., 2020). Prior research has consistently demonstrated that DMHIs incorporating cognitive-behavioral therapy (CBT) and other evidence-based modalities (e.g., mindfulness, behavioral activation, stress reduction) are more effective than treatment as usual, waitlist controls, and attention controls in reducing common mental health symptoms (Goldberg et al., 2022). DMHI interventions also provide a novel avenue for population-level scalability by lowering the barriers to treatment, improving cost-efficiency, reducing stigma, and ameliorating expected shortages of behavioral health providers (Andrilla et al., 2018; Askari et al., 2023; Kaur et al., 2023). Moreover, emergent research indicates that DMHIs may be useful in reducing SI (Büscher et al., 2020; Hetrick et al., 2017; Sander et al., 2020), yet there is comparatively less research about the association between therapist-supported DMHIs and reductions in SI. Therapist-supported DMHIs represent a unique modality for addressing SI, as they facilitate a remote continuous care approach whereby therapists first establish rapport and trust with patients that provide the foundation for a digital therapeutic alliance over the duration of the intervention (Forman-Hoffman et al., 2021; Krieger et al., 2023). As such, there remains a crucial need for real-world research that evaluates the potential effectiveness of therapist-supported DMHIs in reducing SI.

The primary goal of this study was to investigate the feasibility and preliminary effectiveness of a 12-week, therapist-supported DMHI in reducing SI among people experiencing clinically significant symptoms of depression and anxiety. We hypothesized that treatment week would be negatively associated with SI, with significant reductions in SI from baseline to end-of-treatment as well as at 3- and 6-months post-treatment. The secondary goal was to explore potential reductions in suicide attempts and mortality during and after the DMHI using estimates from a large electronic health records database (Rossom et al., 2017; Simon et al., 2013). We hypothesized that participation in the DMHI would be associated with significant reductions in estimated suicide attempts and mortality from baseline to end-of-treatment and both post-treatment follow-ups.

METHODS

Participants

The present study included patients treated with the Meru Health Program (MHP), an evidence-based therapist-supported DMHI (Goldin et al., 2019; Peiper et al., 2023). All data came from 778 adult participants in the United States who started the program on or after January 1, 2020 and had an end date on or before December 1, 2020. Participants entered the MHP via referrals from employee assistance programs or via their healthcare providers. Inclusion criteria of the MHP required patients to have at least mild levels of depression, anxiety, or burnout; own a smartphone; and not have an active substance use disorder, severe active suicidal ideation with a specific plan, severe active self-harm, or a history of psychosis or mania.

All enrolled patients provided informed consent to participate and have their de-identified data used for research purposes. Data collected as part of care were stored in Health Insurance Portability and Accountability Act-compliant electronic medical records that included protected health information. All data were encrypted in transit, and at rest. Institutional review board exemption for this analysis was obtained from the Pearl Institutional Review Board (#21-Meru-113) for analyses of previously collected and de-identified data.

Intervention Description

After the informed consent procedure, patients were enrolled in the intervention. The MHP is composed of several different evidence-based components or processes of evidence-based therapies delivered via a smartphone app. The program is self-guided, but incorporates a stepped and continuous care model that includes daily interaction with a dedicated, licensed clinical therapist and a medical doctor and psychiatrist available for consultation as needed. The program lasts 12 weeks and contains components of cognitive-behavioral therapy (CBT), behavioral activation, mindfulness-based stress reduction, sleep therapy, nutritional psychiatry, and heartrate variability biofeedback (Carney et al., 2017; Lehrer et al., 2013; Morgan, 2003; Sarris et al., 2015). The MHP sequentially delivers self-guided modules covering specific topics over a 12-week period: 1) mindfulness; 2) negative cognition; 3) mood and motivation; 4) rumination; 5) emotional regulation; 6) self-compassion; 7) introspection; 8) social functioning; 9) sleep habits; 10) healthy diet; 11) nutrition and mood; and 12) resilience.

Prior to the start of treatment, participants are trained on how to use the app, including how to participate in the anonymous group interaction and how to communicate with their assigned therapist via chat or phone/video calls. Each week of the program begins with an introductory video that gives the patient an overview of the topics covered that week. Patients are prompted to complete various practices, including CBT exercises and prompts from the app to record progress in a journal, as well as short audio-guided mindfulness meditation exercises on a daily or almost daily basis. Daily content and practices range from 5–15 minutes, except for the first day of each week, in which the weekly psychoeducation video lessons extend the content to a maximum of 25 minutes.

A licensed therapist provides support to participants via messaging (and less frequently, phone or video calls) as needed, and reviews practice logs using a provider dashboard and electronic medical records that detail individual progress (including participant engagement and patient-reported outcomes to date). In total, therapists allocate approximately 10–20 minutes (on average) per week per participant. Interaction between therapist and participant can be initiated via either party.

As a safety measure, therapists conduct a phone-based protocol assessment for any participants that show signs of mental deterioration during the intervention. In case of an emergency, such as having SI with intent to act or onset of psychotic symptoms, the intervention includes a written action plan for declining mental health, which all participants are required to review with their therapist prior to engaging with the intervention. In these situations, care coordinators and therapists will help connect the patient to immediate and local care outside of the program, however, a psychiatrist employed by Meru Health is also available for consultation in these situations.

Patients complete the intervention as a cohort of typically 8–15 patients. As such, cohorts can interact anonymously when sharing their practice experiences, such as providing and receiving support and feedback on their experiences as they navigate the intervention. Direct communication between participants, however, is not allowed. Instead, participants can post anonymous reflections on practices and lessons to the chat discussion board, to which their therapist can respond freely, and to which other group members can respond with pre-written empathy statements and/or emoticons.

Measures

SI and depression symptoms were collected during the intervention on a biweekly basis within the MHP app. For the 3- and 6-month post-treatment follow-ups, patients received emails with instructions on how to report their mental health symptoms in a secure web-based platform. Participants completed the PHQ-9 during and after the intervention. The PHQ-9 is a clinically validated and widely used instrument used to screen for depression (Kroenke et al., 2001). We used the first 8 items of the PHQ-9 to calculate the PHQ-8, which assessed purely depressive symptoms without the last item that assesses SI, the primary outcome of interest (Kroenke et al., 2009). We coded treatment baseline PHQ-8 symptoms using categories established for the PHQ-9 with the 9th item removed as None/Minimal (0–4), Mild (5–9), Moderate (10–14), Moderately Severe (15–19), and Severe (20–24) to assess baseline depression severity.

Self-reported measures collected during clinical intake included age, gender, number of lifetime major depressive episodes (none, first episode, or recurrent), lifetime suicide attempts (yes/no), lifetime prior psychiatric hospitalization (yes/no), lifetime exposure to a major traumatic event (yes/no), and current psychotropic medication use (yes/no). Intervention engagement was defined as the number of days active in the app each week and captured as a core app metric.

We estimated reductions in suicide attempts and death by suicide using published risk data aggregated from 297,290 adult outpatients who received mental healthcare or had a mental health diagnosis documented in primary care (Rossom et al., 2017). Based on these findings, there was a 0.40% risk of suicide attempt and 0.03% risk of suicide death for 0 (not at all), 1.60% and 0.12% risk for 1 (several days), 2.80% and 0.21% risk for 2 (more than half the days), and 4.00% and 0.30% risk for 3 (nearly every day).

Statistical Analysis

All statistical analyses were conducted with RStudio, Version 1.3.959. The study was preregistered on Open Science Framework (https://osf.io/eug8w/). Statistical significance was defined as a p-value of 0.05. Exploratory analyses including histograms as well as skew and kurtosis statistics were run for each variable to check for normality. Any variable that had a skew of +/- 2 was log transformed, except for the SI item (skew = 3.81). Descriptive statistics were calculated for each patient demographic and clinical variable. Outcome measures were analyzed using an intention-to-treat (ITT) analysis in which all participants with outcome measures at baseline were included, regardless of intervention engagement or attrition. Approximately 12.7% of data were missing from the final sample (note this study used ITT, so we included all treatment weeks 0–12 and 3- and 6-month follow-up data as missing for participants that dropped out). To assess whether data were missing completely at random (MCAR) we performed a parametric test using the naniar package (Tierney & Cook, 2018). Although the data were not MCAR (p < 0.001), we performed multiple imputation (10 imputations) using the mice package to account for missing data based on recent recommendations (van Buuren & Groothuis-Oudshoorn, 2011). Additional details about missing data procedures may be found in the supplemental materials.

First, we examined the impact of treatment week on SI by conducting a multilevel model (MLM) that nested treatment week (Level-1) within individuals (Level-2) across linear, Poisson, and negative binomial models to account for the skewed distribution of the SI item (skew = 3.81). We used the lme4 package for linear and negative binomial models and the glmmTMB package for the Poisson model (Bates et al., 2014; Brooks et al., 2017). Fixed effects including covariates were tested at the level of participants (Level-2). This statistical approach accounts for dependency within participants and introduces less bias due to missing data compared to traditional statistical analyses, such as repeated measures of analysis of variance (Hoffman & Stawski, 2009). We tested concurrent associations between treatment week and SI during the same week (Level-1).

Second, we used ggstatsplot package to conduct a nonparametric Friedman Test, an alternative to a repeated measures ANOVA for nonnormal data, to determine whether there were significant reductions in SI from baseline to end-of-treatment as well as at the 3- and 6-month post-treatment follow-ups (Patil, 2021). We then used the rstatix package to conduct nonparametric paired samples Wilcoxon Test contrasts that were corrected for multiple comparisons using the Holm Method to determine whether there was a significant reduction in SI from baseline to end-of-treatment as well as at the 3- and 6-month post-treatment follow-ups (Holm, 1979; Kassambara, 2020).

Third, we calculated frequency statistics on the response options of the SI item at baseline and end-of-treatment. We then extrapolated suicide attempt and mortality data from Rossom et al. (2017) to calculate estimated suicide attempts and mortality at baseline, end-of-treatment, 3-months post-treatment, and 6-months post-treatment. The percent reduction in suicide attempts and mortality at post-treatment timepoints was calculated with the following equation: (baseline percent - post-treatment percent)/baseline percent*100. We used the equation 1/(baseline percent-post-treatment percent) to calculate the number needed to treat (NNT) for both suicide attempts and deaths by suicide outcomes.

RESULTS

Demographic and Clinical Characteristics

Table 1 presents demographic and clinical characteristics for the 778 patients. Most patients were female (77.12%) with an average age of 40. The mean baseline PHQ-8 was 11.1, which falls into the moderate range of symptom severity. Just under half of the sample (44.22%) reported taking psychiatric medication, while 42.03% reported a past traumatic event. Approximately 4.5% of participants reported a lifetime suicide attempt and 5.01% reported a lifetime psychiatric hospitalization.

TABLE 1. Patient demographic and clinical characteristics (N = 778).

Variable	N	Percentage	Mean (SD), Range
Age	778		40 (11.2), 19–82
Gender
Female	600	77.12%
Male	177	22.75%
Other	1	0.13%
Baseline PHQ-8 Severity			11.1 (5.9), 0-24
Less than Mild	117	15.04%
Mild	222	28.53%
Moderate	187	24.04%
Moderately Severe	182	23.39%
Severe	70	9.00%
MDEs
None	376	48.33%
First	114	14.65%
Recurrent	288	37.02%
Taking Psychiatric Medication
Yes	334	44.22%
No	434	55.78%
Lifetime Suicide Attempt
Yes	35	4.50%
No	743	95.50%
Lifetime Psychiatric Hospitalization
Yes	39	5.01%
No	739	94.99%
Lifetime History of Major Traumatic Event
Yes	327	42.03%
No	435	55.91%
Unknown	16	2.06%

Notes: MDE = Major Depressive Episode; PHQ = Patient Health Questionnaire

Symptom severity is measured by baseline PHQ-8 score as follows: less than mild = 0–4, mild = 5–9, moderate = 10–14, moderately severe = 15–19, severe = 20+.

Changes in Suicidal Ideation

The proportion of participants reporting SI “not at all” significantly increased between baseline and end-of-treatment (78.02% to 91.00%, see Table 2). Conversely, there was a significant decline in reported SI during treatment that was maintained at 3- and 6-month follow-ups (Figure 1). Effect sizes of SI changes between baseline and end-of-treatment, baseline and 3-month, and baseline and 6-month follow-ups were Hedge’s g = 0.33 (95%CI = 0.27–0.38), Hedge’s g = 0.32 (95%CI = 0.27–0.38), and Hedge’s g = 0.32 (95%CI = 0.27–0.38), respectively, indicating moderate effect sizes. The significant declines in SI remained after adjusting for demographic and clinical characteristics across linear (β = −0.110, p < 0.001), Poisson (log mean = −0.158, p < 0.001), and negative binomial (log-mean = −0.158, p < 0.001) models (Table 3). Higher engagement was also significantly associated with declines in SI across all three models (p’s < 0.001). The trajectories of SI by depression symptom severity appeared to be steepest in the first few weeks of treatment and then tapered off toward the end-of-treatment and continued to be stable over the 3-and 6-month follow-up assessments for each of the severity groups (see Figure 2 for individual and pretreatment symptom severity group trajectories). Individual participant depression symptom trajectories are displayed in Supplementary Figure S1.

FIGURE 1. Severity of suicidal ideation during and after treatment (N = 778).

Notes: Suicidal ideation over the last 2 weeks is measured by the ninth item of the Patient Health Questionnaire-9 item (PHQ-9), with 0 = not at all, 1 = several days, 2 = more than half the days, and 3 = nearly every day. Non-significant Wilcoxon Test comparisons for changes in suicidal ideation are not shown. That is, there were no significant differences between end-of-treatment and 3-month follow-up, between end-of-treatment and 6-month follow-up, or between 3-month and 6-month follow-up.

FIGURE 2. Suicidal ideation during and after treatment by baseline depressive symptom severity.

Notes: Suicidal ideation over the last 2 weeks is measured by the ninth item of the Patient Health Questionnaire-9 item (PHQ-9), with 0 = not at all, 1 = several days, 2 = more than half the days, and 3 = nearly every day. Treatment week 0 is the pretreatment (baseline) assessment. Treatment week 12 is the end-of-treatment assessment. Treatment week 25 is the 3-month follow-up assessment. Treatment week 38 is the 6-month follow-up assessment. Symptom severity is measured by baseline PHQ-8 score as follows: less than mild = 0–4, mild = 5–9, moderate = 10–14, moderately severe = 15–19, severe = 20+. Light lines are individual trajectories and dark lines are pretreatment symptom severity group trajectories.

TABLE 2. Suicidal ideation during and after treatment.

	Baseline	End-of-Treatment	3 Month Follow-Up	6 Month Follow-Up
Suicidal Ideation	n (%)	n (%)	n (%)	n (%)
0 (not at all)	607 (78.02%)	708 (91.00%)	717 (92.16%)	724 (93.06%)
1 (several days)	127 (16.32%)	63 (8.10%)	58 (7.46%)	53 (6.81%)
2 (more than half the days)	33 (4.24%)	6 (0.77%)	2 (0.26%)	1 (0.13%)
3 (nearly every day)	11 (1.41%)	1 (0.13%)	1 (0.13%)	0 (0.00%)
Mean (SD)	0.29 (0.61)	0.10 (0.34)	0.08 (0.30)	0.07 (0.26)
Effect Size		0.33 (0.27, 0.38)	0.32 (0.27, 0.38)	0.32 (0.27, 0.38)

Notes: SD = standard deviation. Suicidal ideation over the last 2 weeks is measured by the ninth item of the Patient Health Questionnaire-9 item (PHQ-9), with 0 = not at all, 1 = several days, 2 = more than half the days, and 3 = nearly every day.

TABLE 3. Adjusted associations of changes in suicidal ideation across meru health program treatment.

	Linear Model			Poisson Model				Negative Binomial Model
Predictors	Std. Beta	Std. SE	P-Value	Log-Mean		SE	P-Value	Log-Mean		SE	P-Value
(Intercept)	−0.321	0.062	<0.001	−4.197		0.481	<0.001	−4.197		0.560	<0.001
Treatment Week	−0.110	0.013	<0.001	−0.158		0.022	<0.001	−0.158		0.013	<0.001
Gender- Male	0.087	0.055	0.111	0.259		0.171	0.130	0.259		0.177	0.144
Gender- Other	−0.620	0.631	0.326	−15.901		2063.644	0.994	−22.229		65819.200	1.000
Age	−0.045	0.023	0.054	−0.016		0.006	0.014	−0.016		0.007	0.017
Average Days Active Weekly	−0.098	0.012	<0.001	−0.047		0.010	<0.001	−0.047		0.010	<0.001
Single MDE	0.121	0.069	0.078	0.472		0.233	0.043	0.472		0.249	0.058
Recurrent MDE	0.205	0.055	<0.001	0.871		0.176	<0.001	0.871		0.185	<0.001
Lifetime Psychiatric Hospitalization-Yes	0.433	0.116	<0.001	0.563		0.290	0.053	0.563		0.290	0.052
Lifetime Suicide Attempt-Yes	−0.005	0.123	0.967	−0.090		0.314	0.774	−0.090		0.316	0.775
Lifetime Major Traumatic Event-Unknown	−0.018	0.162	0.910	0.151		0.427	0.724	0.151		0.434	0.728
Lifetime Major Traumatic Event-Yes	−0.015	0.048	0.753	0.152		0.154	0.322	0.152		0.160	0.342
Baseline PHQ-8 Symptom Severity- Mild	0.012	0.068	0.864	1.049		0.437	0.016	1.049		0.533	0.049
Baseline PHQ-8 Symptom Severity- Moderate	0.131	0.072	0.068	2.010		0.424	<0.001	2.010		0.515	<0.001
Baseline PHQ-8 Symptom Severity- Moderately Severe	0.420	0.075	<0.001	2.797		0.420	<0.001	2.797		0.509	<0.001
Baseline PHQ-8 Symptom Severity- Severe	0.579	0.099	<0.001	3.210		0.441	<0.001	3.210		0.526	<0.001
Taking Psychotropic Medication	0.022	0.034	0.519	0.005		0.114	0.967	0.005		0.118	0.968
Marginal R^2 / Conditional R^2 = 0.289 / 0.540					0.289 / 0.540				0.313 / 0.556

Notes: CI = confidence interval, MDE = Major Depressive Episode, PHQ = Patient Health Questionnaire, SE = standard error. Symptom severity is measured by baseline PHQ-8 score as follows: less than mild = 0–4, mild = 5–9, moderate = 10–14, moderately severe = 15–19, severe = 20+.

Correlates of Changes in Suicidal Ideation

Significant correlates of change in SI across all models included average days of weekly activity (-), having a history of recurrent MDEs versus no lifetime MDEs (+), and having either moderately severe or severe baseline depression symptoms versus less than mild depression symptoms (+). Other significant correlates included age (- in Poisson and negative binomial models only), having a single MDE versus no lifetime MDEs (+ in the Poisson model only), having a history of a psychiatric hospitalization (+ in the linear model only), and mild or moderate depression symptom severity versus less than mild depression symptoms (+ in Poisson and negative binomial models only).

Estimates of Change in Suicide Attempts and Deaths

Suicide attempt estimates fell from 0.75% to 0.52% from baseline to end-of-treatment and further declined to 0.50% and 0.49% at 3- and 6-month follow-up. For suicide mortality, estimates reduced from 0.033% to 0.003% between baseline and end-of-treatment that were maintained at the 3- and 6-month follow-ups (0.003% and 0.0003%, respectively). These represent a 30.49% reduction (95%CI = 25.15%–35.13%) in estimates of both suicide attempts and death by suicide between baseline and end-of-treatment. Extrapolating findings to number needed to treat (NNT) estimates for end-of-treatment, analyses revealed that one suicide attempt would be prevented for every 438 participants enrolled in the intervention and one death by suicide would be prevented for every 5841 participants enrolled.

DISCUSSION

The study investigated whether SI decreased during and following participation in a 12-week, therapist-supported DMHI for depression and anxiety symptoms. In addition, the study aimed to estimate reductions in suicide attempts and deaths by suicide among this sample. Overall, this study shows moderate evidence of preliminary feasibility and effectiveness of improving SI among patients who participated in a therapist-supported DMHI.

Our hypotheses were supported by our findings. First, treatment week was associated with significant decreases in SI across the MLM models independent of pretreatment depression symptom severity, demographic factors, and clinical characteristics. In addition to treatment week effects, these models revealed that a history of MDE, particularly recurrent history of MDE, was associated with higher SI across treatment and that higher pretreatment depression symptom severity was associated with greater SI as would be expected. Older age was also associated with lower SI, which coincides with findings that SI tends to have a negative association with age (Dennis et al., 2007). Moreover, the number of days patients engaged with therapy each week was associated with lower rates of SI, which is consistent with recent research indicating that twice vs once weekly therapy sessions is more effective at reducing depression symptoms and attrition (Bruijniks et al., 2020). This finding suggests DMHIs that prioritize frequent practice of psychotherapy skills may be particularly useful in lowering SI and other symptoms of depression. A history of suicide attempt and psychiatric hospitalization were not significantly associated with SI in the past two weeks, except for the latter in linear models, which did not fit the data as well as Poisson and negative binomial models (see Table S1). As previous research indicates that risk factors of SI differ from those for the transition from SI to attempt, additional studies will be necessary to better understand how DMHIs may impact different suicide pathways (e.g., ideation, planning, attempts) and complex etiological mechanisms (Nock et al., 2016)

Results also indicated that there was a significant overall reduction in SI from baseline to post-treatment, which was maintained at both 3- and 6-month follow-ups, indicating a robust and long-lasting association between treatment on SI trajectories across time. Importantly, while SI continued to decrease from post-treatment to 3- and 6-month follow-ups, there was no significant decrease, possibly due to a floor effect. Finally, there was a greater than 30% decrease in both estimated suicide attempts and deaths by suicide during the 12-week DMHI. The estimated NNT of 438 and 5,841 for attempts and deaths, respectively, suggests that therapist-supported DMHIs focusing on depression and anxiety symptoms may be a viable suicide prevention strategy through remote continuous monitoring and post-treatment follow-ups (Forman-Hoffman et al., 2021; Krieger et al., 2023).

While this study had a number of significant strengths including a large sample size of patients with a diverse history of psychiatric and clinical characteristics with an ITT study design, the study findings should be interpreted in light of several limitations. Because this study utilized a single arm intervention design with observational data, pragmatic randomized controlled trials that include relevant control conditions (e.g., treatment as usual, waitlist) are necessary to produce stronger evidence of the effectiveness of therapist-supported DMHIs in reducing SI. Nevertheless, the ITT design and use of multiple imputation procedures increased the study’s rigor compared to other studies that commonly employ completer-only analyses. Another limitation is the reliance on the 9th item of the PHQ-9 as our SI measure. Although the 9th item of the PHQ-9 has been shown to be a robust predictor of suicidal behavior in study populations (Louzon et al., 2016; McCusker, 2016; Rossom et al., 2017; Simon et al., 2016; Yarborough et al., 2021), some research indicates that this item may not be a sufficient screen for SI (Razykov et al., 2012). Furthermore, the predictive ability of the SI item may vary by demographic and clinical characteristics (Na et al., 2018), suggesting it may be limited in distinguishing individual-level suicide. Future studies should incorporate gold-standard assessments of SI (O’Connor et al., 2013; Posner et al., 2011; Ringer et al., 2018). Similarly, the reductions in suicide attempts and mortality found in this study were based upon estimates using recently published associations found in large electronic health record databases as opposed to actual data on attempts and mortality linked to the DMHI participants (Rossom et al., 2017). Studies that collect longitudinal follow-up data and leverage administrative data linkages are needed to better assess actual rates of suicide attempts and mortality (Kessler & Luedtke, 2021). This study used a single measure of intervention engagement to ensure statistical parsimony. Follow-up studies will more thoroughly examine multiple dimensions of engagement (e.g., patient-therapist interactions, cognitive-affective responses) and differential engagement patterns to understand potential effects on SI (Chien et al., 2020; Perski et al., 2017; Wu et al., 2021). Lastly, the DMHI in this study excludes patients with a current suicide plan, limiting generalizability to other DMHIs that may target people actively transitioning from ideation to action (Nock et al., 2016).

Suicide is a significant global health concern associated not only with the human cost, but also significant secondary societal and economic impacts. There is a pressing need for novel interventions that reduce barriers to treatment, decrease costs, and increase scalability in order to reduce these suicide outcomes. Thus, this study provides real-world evidence of the potential impact of therapist-supported DMHIs on reducing SI, depression, and anxiety, which has major implications for addressing the mental health crisis in the US.

AUTHOR NOTES

Benjamin W. Nelson, Verily Life Sciences, San Francisco, California, USA. Valerie L. Forman-Hoffman, Woebot Health, San Francisco, California, USA. Nicholas C. Peiper, Department of Epidemiology and Population Health, University of Louisville, Louisville, Kentucky, USA.

DISCLOSURE STATEMENT

Dr. Peiper reports research support from the National Institute of Mental Health (R44MH126836; PI: Peiper), Lilly Oncology, Gilead Sciences, and Norton Healthcare Foundation. Drs. Peiper and Forman-Hoffman report stock options from Meru Health. The authors were affiliated with Meru Health during the conduct of this study. Drs. Nelson, Forman-Hoffman, and Peiper are now affiliated with Verily Life Sciences, Woebot Health, and University of Louisville, respectively.

Additional information

Funding

Dr. Peiper reports research support from the National Institute of Mental Health (R44MH126836), Lilly Oncology, Gilead Sciences, and Norton Healthcare Foundation. Drs. Peiper and Forman-Hoffman report stock options from Meru Health. Drs. Forman-Hoffman and Nelson were employed by Meru Health during the conduct of this research. They are now with Woebot Health and Verily Life Sciences, respectively.