Patient-Provider Matching, Engagement, and Outcomes of a Digital Mental Health Treatment Platform: Real-World Retrospective Cohort Study
Bibliographic record
Abstract
BACKGROUND: Technology-enabled mental health platforms that incorporate user-driven patient-provider matching may offer a novel way to personalize and optimize outcomes. We conducted this study because little is known about the engagement and clinical symptom changes of these newer types of mental health platforms and whether patient-driven selection of their provider's characteristics is associated with either engagement or clinical outcomes. OBJECTIVE: This study aimed to determine the levels of engagement and clinical symptom changes associated with the use of a technology-enabled mental health platform that allows patients to select preferred provider characteristics and to explore whether the selection of a provider characteristic was associated with engagement and clinical outcomes. METHODS: We conducted a real-world, retrospective cohort study using deidentified electronic health data from adult Grow Therapy patients aged 18 years or older with clinically elevated depressive or anxiety symptoms at baseline (PHQ-9 [Patient Health Questionnaire-9] > 9 or GAD-7 [Generalized Anxiety Disorder-7] > 9). Inclusion required 1 provider visit (intent-to-treat cohort) for engagement analyses; clinical outcome analyses required 2 or more provider visits (complete case cohort). Engagement with the platform was measured by the number of provider visits. Clinical outcomes were measured using changes in PHQ-9 and GAD-7 scores and defined as meeting a minimal clinically important difference (MCID). Bivariate associations between selection of provider characteristics and outcomes were measured using chi-square tests, and adjusted associations were modeled using logistic regression (P<.05). RESULTS: Among 159,448 patients with elevated depressive symptoms and 167,356 patients with elevated anxiety symptoms, engagement was high, with 69.4% (95% CI 69.2%-69.7%) and 69.3% (95% CI 69.1%-69.5%) having 3 or more visits, respectively. In the complete case cohort, symptom reductions were significant; 58.9% (95% CI 58.5%-59.2%) met depressive symptom MCID criteria, and 63% (95% CI 62.6%-63.3%) met anxiety symptom MCID criteria after engagement. Although only ≈35% of patients selected a provider specialty and ≈5% selected a provider identity before enrollment, those selecting a provider specialty experienced significantly better outcomes, and those selecting a provider identity engaged significantly more frequently as compared to those who did not select each characteristic. Sensitivity analyses confirmed these findings. CONCLUSIONS: This exploratory, real-world, uncontrolled study provides early evidence that allowing patients to select provider characteristics within a technology-enabled mental health platform may support both engagement and meaningful symptom improvement. The investigation of the relationship between mental telehealth provider selection characteristics and both engagement and clinical outcomes is a novel addition to the peer-reviewed literature. Findings highlight how user-driven, scalable matching features may personalize mental health care in ways that differ from traditional assignment-based models and underscore the need for more rigorous, controlled studies to demonstrate efficacy and test causality and mechanisms.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.013 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".