This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on https://formative.jmir.org, as well as this copyright and license information must be included.
Enhanced patient-provider engagement can improve patient health outcomes in chronic conditions, including major depressive disorder (MDD).
We evaluated the impact of a digitally enabled care mobile app, Pathway, designed to improve MDD patient-provider engagement. Patients used a mobile interface to assess treatment progress and share this information with primary care providers (PCPs).
In this 52-week, real-world effectiveness and feasibility study conducted in primary care clinics, 40 patients with MDD who were recently prescribed antidepressant monotherapy were randomized to use a mobile app with usual care (20/40, 50%) or usual care alone (20/40, 50%). Patients in the app arm engaged with the app daily for 18 weeks; a report was generated at 6-week intervals and shared with the PCPs to facilitate shared treatment decision-making discussions. The patients discontinued the app at week 18 and were followed through year 1. Coprimary outcome measures, assessed via research visits, included change from baseline in the 13-item Patient Activation Measure (PAM-13) and 7-item Patient-Provider Engagement Scale scores at week 18. Additional outcome measures included depression severity (9-item Patient Health Questionnaire [PHQ-9]) and cognitive symptoms (5-item Perceived Deficits Questionnaire–Depression).
All 37 patients (app arm: n=18, 49%; usual care arm: n=19, 51%) who completed the 18-week follow-up period (n=31, 84% female, mean age 36, SD 11.3 years) had moderate to moderately severe depression. Improvements in PAM-13 and PHQ-9 scores were observed in both arms. Increases in PAM-13 scores from baseline to 18 weeks were numerically greater in the app arm than in the usual care arm (mean 10.5, SD 13.2 vs mean 8.8, SD 9.4;
Patients with MDD will engage with a mobile app designed to track treatment and disease progression. PCPs will use the data generated as part of their assessment to inform clinical care. The study results suggest that an app-enabled clinical care pathway may enhance patient activation and benefit MDD management.
ClinicalTrials.gov NCT03242213; https://clinicaltrials.gov/ct2/show/NCT03242213
Major depressive disorder (MDD) is among the most common mental health disorders in the United States, with a lifetime prevalence of up to 20% [
Effective management of MDD in primary care requires a systematic approach to diagnosis, patient education, treatment, close follow-up, and a commitment to adjusting care or consulting with specialists when needed [
One proposed strategy to increase patient engagement, improve patient-provider communication to facilitate MDD management, and improve patient outcomes in primary care is the use of mobile health apps and digital platforms [
The development of any mobile health and information technology tool, including those supporting MDD management, may benefit from collaboration between industry, app developers, and the health care team representing large health care systems [
The primary objective of this study was to determine whether the addition of the Pathway mobile app to usual clinical care improves patient-provider engagement in the management of MDD over an 18-week period. The secondary objectives of the study were to evaluate the impact of the Pathway mobile app on changes in certain patient-reported outcomes, including self-reported clinical depression severity (via the 9-item Patient Health Questionnaire [PHQ-9]), cognitive dysfunction (via the 5-item Perceived Deficits Questionnaire–Depression [PDQ-D5]), quality of life (via the 5-item World Health Organization Well-Being Index [WHO-5]), and patient satisfaction, as well as changes in medication and medication adherence. Additional retrospective assessments were also planned to evaluate the impact of the app on measures of health care use 1 year after enrollment.
In this randomized controlled pilot study (ClinicalTrials.gov NCT03242213), we enrolled patients diagnosed with MDD who were receiving primary care services at Advocate Health Care, which is now part of AAH. The study took place between July 2017 and January 2019 and involved 4 study sites in suburban and urban settings within metropolitan Chicago, Illinois, including Advocate Medical Group–Huntley, Advocate Medical Group–Hometown Family Medicine, Family Medicine Center in Ravenswood, and Adult Medicine Center in Oak Lawn. On identifying a patient with MDD who met the criteria for study participation, the physician providing care for that patient introduced the study; a designated research study coordinator then explained the study and obtained informed consent. Patients were then randomized to receive either the Pathway mobile app along with usual care or usual care alone. Participating patients were randomized based on the results from a randomized study list created using serially generated random numbers obtained by the study staff using a random number generator. Patients in the mobile app arm were encouraged to engage with the mobile app daily for 18 weeks (
The mobile app was specifically designed to enhance patient-provider engagement, promote shared decision-making, and support measurement-based care in the management of clinical depression. It also provided patients with a way to track changes in clinical depression severity, cognitive symptoms, quality of life, emotional well-being, and adherence to medications; set up medication reminders; and record side-effect experiences. App functions included PHQ-9 and PDQ-D5 assessments conducted every 2 weeks, daily assessments of depression using 2 questions from the PHQ-9 and 1 question from the PDQ-D5, daily assessments of emotional well-being using a visual analog measurement of global well-being on a scale of 0 to 100, and daily cognitive symptoms assessed using the Cognition Kit 2-back test, in which patients indicate whether the symbol (usually an abstract shape) matches the symbol 2 items back [
The single-phase study included 2 follow-up periods: the primary follow-up period and the long-term follow-up period. The primary follow-up period began with randomization and continued through week 18. The long-term follow-up period began after the final visit at week 18 and continued through the 1-year follow-up phone interview (34 weeks after the use of the mobile app was discontinued). At year 1, the results from the follow-up phone interview and extraction of data from patients’ electronic medical records (EMRs) were analyzed to evaluate the residual impact of the app on patient-reported outcomes and health care use.
Study design. Long-term follow-up period indicates that no visits, calls, or use of the mobile app occurred during this phase in either treatment arm. The app arm also included usual care. ED: emergency department; LOS: length of stay; PAM-13: 13-item Patient Activation Measure; PDQ-D5: 5-item Perceived Deficits Questionnaire–Depression; PHQ-9: 9-item Patient Health Questionnaire; PPES-7: 7-item Patient-Provider Engagement Scale; WHO-5: 5-item World Health Organization Well-Being Index.
Screenshots of the app and a sample report.
Patients were eligible to participate in the study if they were aged 18 to 70 years, were receiving primary care for MDD at AAH, used an iPhone version 5 or later or a smartphone with an Android operating system, and had an active cellular data plan or regular Wi-Fi access. Patients were also required to have a PHQ-9 score of >5 at baseline, indicating at least mild depression severity [
Patients were excluded from the study if they had a diagnosis of a major psychiatric disorder other than MDD, were considered to be at imminent risk of hospitalization due to MDD, had been hospitalized due to MDD within 3 months, had a significant risk of suicide or had a previous suicide attempt within 6 months, or had a history of responding only to combination or augmentation therapy in their current depressive episode. Patients for whom the use of antidepressants was contraindicated were not eligible for the study.
Eligible patients were randomized to usual care (20/40, 50%) or usual care plus the mobile app (20/40, 50%) for 18 weeks (unblinded for both participants and researchers). An in-person introduction to the app and instructional handouts were provided to the patients at the time of enrollment. Although the use of the mobile app was encouraged, it was not required. Patients assigned to usual care received regular care as needed from their PCP; no specific interventions were mandated. At the end of the long-term follow-up period (1 year), patient-reported outcome measures were collected via follow-up phone calls. Data on health care resource use and medication changes during the long-term follow-up period were collected from medical chart reviews. A quality control committee reviewed the data for adequate completion and integrity.
Study coprimary, secondary, and exploratory outcomes were assessed at 18 weeks for each arm via in-person research visits and phone interviews. Coprimary outcomes included changes in the 13-item Patient Activation Measure (PAM-13) scale and 7-item Patient Provider Engagement Scale (PPES-7) between baseline and 18 weeks. The PAM-13 scale was developed and validated to assess patient engagement and confidence in self-management of the disease [
The secondary outcomes included changes (between baseline and 18 weeks) in depression severity evaluated using the PHQ-9 measure [
Exploratory outcomes included medication and dose switches, defined as medication switches, dosage changes, medication add-ons, or discontinuations (assessed via retrospective chart review), and patient and provider satisfaction with the care and use of the app. Additional assessments were collected throughout the study in the app arm via app functions such as the PHQ-9, PDQ-D5, emotional well-being, and cognitive symptom assessments. In addition, information on patient satisfaction with the app and app use data were collected from the patients in the app arm.
After the 18-week end-of-study visit, patients and providers were invited to participate in a remote, qualitative, semistructured interview using a digital tool to discuss sentiments on app features and future features. This qualitative tool allowed the interviewer to observe and talk to the patients as they looked at the app’s features on their own devices, record the interview, and capture time-stamped notes.
At the end of the long-term follow-up phase of the study (at year 1), a phone interview was conducted (34 weeks after the use of the mobile app was discontinued) to assess patient-reported outcomes, including patient and provider engagement (using PAM-13 and PPES-7), quality of life (using WHO-5), and depressive symptoms (using PHQ-9). At this time, a retrospective analysis that compared health care use between the app arm and the usual care arm was also conducted. Data were collected on inpatient visits, including depression-related hospitalizations; emergency department (ED) visits; outpatient visits, including visits to PCPs, psychiatrists, behavioral therapy specialists, and other health care providers; and medication and dose switches. Spontaneously reported serious adverse events were also recorded during the study period.
This was a pilot study, and thus no sample size estimation was conducted. A sample size of 20 patients per group was expected to be sufficient to provide initial information about the potential effects and benefits of the app and the feasibility of its real-world use to inform future larger-scale studies. Patients were included in the analysis based on treatment allocation, and an intent-to-treat analysis was used. As 18-week follow-up data were not available for 3 randomized patients, these patients were dropped from the analysis, and an intent-to-treat analysis (with the exclusion of missing data) was conducted on the remaining population. For the primary and long-term follow-up periods, between-group differences in changes in continuous variables were evaluated using the 2-tailed Student
A retrospective comparison of health care use was conducted between patients in the usual care arm and those in the mobile app arm using medical record extraction. This comparison included data from the time of consent to 1 year after enrollment for each patient. Health care use was compared between groups for long-term differences at 1 year after each patient’s study enrollment, overall, by cause (depression-related or not) and by category (inpatient via ED, outpatient, or specialty). Sensitivity analysis was used to assess attrition bias among patients lost to follow-up in the long-term follow-up period; these outcomes were used to assess the generalizability of the 52-week results across the original study group. Statistical analysis was performed using SAS software (version 9.4; SAS Institute).
This study (#Y5000249) was approved by the Advocate Health Care Institutional Review Board.
A total of 40 patients were enrolled, and of them, 37 (93%) completed the 18-week primary follow-up period (
CONSORT (Consolidated Standards of Reporting Trials) diagram.
The baseline characteristics of patients who completed 18 weeks of the study are shown in
Demographics and baseline characteristics for patients completing 18 weeks.
|
App (n=18) | Usual care (n=19) | Total (N=37) | |
Sex (female), n (%) | 14 (78) | 17 (90) | 31 (84) | |
Age (years), mean (SD) | 33.8 (11.4) | 38.9 (11.0) | 36.4 (11.3) | |
Aged ≥45 years, n (%) | 2 (11) | 5 (26) | 7 (19) | |
|
||||
|
Hispanic | 8 (44) | 6 (32) | 14 (38) |
|
Non-Hispanic Black | 2 (11) | 5 (26) | 7 (19) |
|
Non-Hispanic White | 7 (39) | 8 (42) | 15 (41) |
|
Non-Hispanic multiracial | 1 (6) | 0 (0) | 1 (3) |
|
||||
|
Employed full-time | 6 (33) | 10 (53) | 16 (43) |
|
Employed part-time | 5 (28) | 3 (16) | 8 (22) |
|
Self-employed | 1 (6) | 2 (11) | 3 (8) |
|
Not employed | 3 (17) | 2 (11) | 5 (14) |
|
Student | 3 (17) | 2 (11) | 5 (14) |
|
Nonworking spouse, retired, or other | 2 (11) | 2 (11) | 4 (11) |
Annual income <US $40,000, n (%) | 13 (72) | 8 (42) | 21 (57) | |
Education (associate’s degree or higher), n (%) | 3 (17) | 12 (63) | 15 (41) | |
|
||||
|
0 | 6 (33) | 7 (37) | 13 (35) |
|
1-3 | 8 (44) | 9 (47) | 17 (46) |
|
≥4 | 4 (22) | 3 (16) | 7 (19) |
PHQ-9b, mean (SD) | 15.3 (5.1) | 14.1 (5.0) | 14.7 (5.0) | |
|
||||
|
SSRIsc | 13 (72) | 15 (79) | 28 (76) |
|
Bupropion | 2 (11) | 3 (16) | 5 (14) |
|
SNRIsd | 2 (11) | 1 (5) | 3 (8) |
|
TCAse, MAOIsf, SMSsg, or other | 1 (6) | 0 (0) | 1 (3) |
|
||||
|
None | 10 (56) | 11 (58) | 21 (57) |
|
<1 | 2 (11) | 2 (11) | 4 (11) |
|
≥1 | 6 (33) | 4 (21) | 10 (27) |
|
Unknown | 0 (0) | 2 (11) | 2 (5) |
aMDD: major depressive disorder.
bPHQ-9: 9-item Patient Health Questionnaire.
cSSRI: selective serotonin reuptake inhibitor.
dSNRI: serotonin and norepinephrine reuptake inhibitor.
eTCA: tricyclic antidepressant.
fMAOI: monoamine oxidase inhibitor.
gSMS: serotonin modulator and stimulator.
At week 18, both arms exhibited an increase in patient activation based on the PAM-13 scores (Table S2 in
Depression severity (as measured by the PHQ-9 score) decreased in both arms (Table S2 in
A total of 11% (2/18) of patients in the app arm and 0% (0/18) of patients in the usual care arm switched medications during the study. One serious adverse event (inpatient hospitalization related to depression) was reported in the app arm.
All patients randomized to the app arm (20/20, 100%) completed at least one app assessment during the study period. A majority of patients (12/20, 60%) completed the PHQ-9 and PDQ-D5 assessments biweekly for at least 12 weeks. A total of 70% (14/20) of the app users completed the self-report of medication assessment daily for >100 days.
Patient satisfaction in the app arm was high, as shown in the results of the patient satisfaction survey administered at 18 weeks (Table S3 in
Positive and negative patient and PCP sentiments on Pathway app features (at 18 weeks). Responses were classified as follows: Gray=positive responses: app features were well received; black=negative responses: app features need further work. Data from remote interviews with providers (n=15) and patients (n=20). PCP: primary care provider.
At year 1, a significant increase in patient activation (PAM-13) was observed for patients in the app arm (Table S2 in
Among the 17 patients assessed during the 1-year follow-up period, there were no inpatient hospitalizations. One patient in the usual care arm visited the ED twice; however, neither visit was considered related to depression. Patients in the usual care arm (n=9) had more outpatient clinic visits to any provider than those in the app arm (n=8; 88 visits vs 49 visits), including visits to PCPs (59 visits vs 38 visits). At 1-year follow-up, 3 patients in each group had a medication change: 1 patient in the mobile app arm and 2 in the usual care arm switched medications; 1 patient in each group had a medication dose change; and 1 patient in the mobile app arm added a new medication to their regimen.
The results of this pilot study suggest that the Pathway mobile app may facilitate the systematic use of measurement-based care in MDD management, which can enhance shared decision-making and patient-provider communication, with improved medication adherence and treatment outcomes [
In this study, the mean PAM-13 scores suggested that patients in the app arm, on average, moved from PAM-13 level 3 (58.2) to level 4 (74.2) at the end of the primary follow-up period, whereas patients in the usual care group, on average, started and remained at level 3. A cross-sectional study of patients visiting a primary care clinic reported that every 10-point increase in the PAM-13 score was associated with a 1% reduction in the predicted probability of having an ED visit or being obese and a 1% increase in the predicted probability of having clinical indicators (eg, hemoglobin A1c) in the normal range [
The severity of depression, quality of life, and subjective cognitive symptoms improved in both arms, with no statistically significant differences between groups from baseline to week 18. Although not statistically significant, a trend toward improvement in depressive symptoms (PHQ-9) was observed among the patients in the app arm. Patient-provider engagement also showed small improvements in both arms. These results suggest that a larger study is warranted to determine whether the use of the app is associated with a clinically meaningful improvement in the symptoms of depression or patient-provider engagement.
At year 1, greater improvement in patient activation (PAM-13;
Moreover, although the overall number of medication changes was similar in both groups at 52 weeks, 2 switches occurred in the app group before the week 18 assessment, with no observed switches in the usual care group in that time frame. In addition, the number of outpatient visits (overall and PCP visits) was greater for patients in the usual care arm than for those in the app arm. These examples may suggest that the app, through improved patient-provider communication, allowed for a more rapid response to changes in patient status while reducing the burden of in-person office visits.
Several systematic reviews and meta-analyses have shown that the use of digital mental health interventions such as apps can aid in the reduction of depressive symptoms, with larger effects seen in patients with more severe symptoms [
Although the app users used the app for only 18 weeks, the 34-week follow-up period enabled us to determine whether the benefits of the app were sustainable. The length of the follow-up period is unique in the field of mobile health research on MDD. In fact, a recent review of the effectiveness of apps targeting patients with MDD identified 18 studies evaluating their impact on depression [
Two important strengths of our pilot study were its randomized controlled study design and the long-term 1-year follow-up period. Another strength was that the app was developed and piloted in collaboration with end users in the health care team and cocreated with patient end users. Research has demonstrated that the effectiveness of digital technology resources can depend on the extent to which end users are included as active participants in their design [
Potential limitations of the study included its small sample size, which limited our ability to identify statistically significant differences between groups, and the relatively short duration of app use although we were able to maintain follow-up with nearly half of the study’s participants after they discontinued its use. In addition, although patients and providers generally expressed high satisfaction with the app and interest in its features, both groups received limited education on how to use the app or its associated reports. The provision of additional education about the functionality and reporting features of the app might help increase patient-provider engagement and lead to improvements in the overall management of MDD. Embedding the reporting feature into the EMR rather than it being a stand-alone report might also improve the ability of PCPs to make real-time decisions about treatment. Additional work on the Pathway platform informed by the results of this study will help integrate the digitally enabled MDD care pathway into the current AAH system by assessing process and workflow improvements, clinician-patient experiences, collaborative care model enhancements, EMR integration, and efficiencies with other platforms.
Our study data and qualitative insights informed the design of a real-world, prospective, interventional study of the app currently underway at AAH, designed to test the scaling and integration of the Pathway platform, along with educational interventions, at multiple primary care sites (Clinical Trials.gov NCT04891224). The goal of this study was to determine whether the use of the app can improve adherence to measurement-based care practices in primary care to help improve outcomes for patients with MDD.
This pilot study demonstrated that patients with MDD will engage with a mobile app designed to track treatment and disease progression and that health care providers will use the data generated as part of their assessment to inform care. The study results demonstrate that it is feasible to conduct an innovative app intervention in this diverse patient population with moderate to moderately severe depression. Introducing a customized, cocreated patient app into the care pathway can provide both patients and clinicians with greater details and trend data related to the disease state outside the traditional in-person visit. Enhanced use of patient-reported data within real-world health care settings can help support measurement-based care practices by making patient self-reported data and summaries of the data easy to interpret and easy to access within existing EMR instances.
Although the sample size was small for the long-term follow-up phase of the study, the results of this feasibility study suggest that this digitally enabled MDD clinical care pathway approach may support shared decision-making and help provide sustainable benefits over at least 1 year. The impact of the app on patient activation and MDD management will be further explored in a larger prospective study of its real-world use in patients with MDD.
Key study demographics, patient-reported outcomes, and longitudinal results.
CONSORT-eHEALTH (Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth) checklist.
Advocate Aurora Health
Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth
emergency department
electronic medical record
major depressive disorder
13-item Patient Activation Measure
primary care provider
5-item Perceived Deficits Questionnaire–Depression
9-item Patient Health Questionnaire
7-item Patient-Provider Engagement Scale
5-item World Health Organization Well-Being Index
The authors thank all patients for their participation in this study. The authors also gratefully acknowledge the contributions of Ellen Rhodes, Lambros Chrones, and Anit Roy, who were involved in the project management and medical review aspects of the study. The authors also thank the study site subinvestigators (Ramon A Gonzalez, Kevin T Loughry, Jacqueline Ivey-Brown, Roxanne Smith, Robert Zitter, and Gina Schueneman) for their contributions to the study and manuscript. This study was funded by Takeda Pharmaceuticals USA, Inc and Lundbeck LLC. Medical writing assistance was provided by Nicole Cooper on behalf of Syneos Health and supported by Takeda Pharmaceuticals USA, Inc and Lundbeck LLC.
The data supporting the findings of this study are included in the published paper (and its supplementary information files). The authors may be contacted for further data sharing. The CONSORT‐eHEALTH (Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth) checklist is available as
MM contributed to the study design, data interpretation, editing, critical review, and approval of the manuscript. CB contributed to the study design, data interpretation, critical review, and approval of the manuscript. BF contributed to the study design, data interpretation, critical review, and approval of the manuscript. JK contributed to the study design, data interpretation, critical review, and approval of the manuscript. FC contributed to the study design, data interpretation, critical review, and approval of the manuscript. SS contributed to the study design, data interpretation, critical review, and approval of the manuscript. AE contributed to the study design, data interpretation, critical review, and manuscript approval. CK contributed to the study design; data collection, analysis, and interpretation; and drafting, editing, critical review, and approval of the manuscript. RK contributed to the study design, data interpretation, critical review, and approval of the manuscript. DK contributed to the study design, data interpretation, critical review, and approval of the manuscript.
MM and SS are employees of Takeda Pharmaceuticals USA, Inc. JK, BF, and FC are employees of Cognition Kit. CB was an employee of Advocate Aurora Health at the time of the study and is now an employee of Takeda Pharmaceuticals USA, Inc. CK, RK, and DK are employees of Advocate Aurora Health. DK has received remuneration from Takeda for activities unrelated to the conduct of the study. AE was an employee of Lundbeck LLC at the time of the study and is now an employee of LB Pharmaceuticals.