Scalable Health Labs

Mobile Bio-Behavioral Sensing


Smartphone- and OnLine-Usage-Based Evaluation for Depression

Concordance Of Clinical & Electronic Data In Assessment of Depression: Findings From The Smartphone And Online Usage Based Evaluation For Depression (SOLVD) Study,” APA 2016. See also news story. 


Jian Cao, Graduate student, ECE, Rice University

Anh Truong, Baylor College of Medicine

Dr. Nidal Moukaddam, Assistant Professor, Baylor College of Medicine

Dr. Ashutosh Sabharwal, Professor, ECE, Rice University

THE PROBLEM: Depression is one of the most common mental disorders that carries significant emotional and financial burden for modern society. At any point in life, 3 to 5 percent adults suffer from major depression, and the lifetime risk is about 17 percent. The situation is even worse for teens and young children, with 2 out of 100 young children and 8 out of 100 teens having serious depression. To ensure successful prevention and treatment, the early diagnosis and continuous monitoring of one’s depression state are critical. However, depression is often monitored through clinician psychometric instruments, and there is still a lack of an effective method to automatic and continuously track the patient’s mental health status.

OUR SOLUTION: We propose to develop the Smartphone- and OnLine-usage-based eValuation for Depression (SOLVD), which is a new tool for continuous and real-time monitoring of the patient’s depression state. We develop the SOLVD mobile app and the backend cloud platform, for data collection, storage, analysis and sharing.

In the study, we collect three types of data from clinically depressed patients:

  1. Smartphone sensor and usage data, including accelerometer, location, steps, screen status, call log, text messages, and apps.
  2. Self-reported mood and activity level on daily basis.
  3. Psychometric data from biweekly in-clinic exams, including PHQ-9 (Patient Health Questionnaire), HamiltonD (Hamilton Rating Scale for Depression) and HamitonA (Hamilton Anxiety Rating Scale).

We conducted a pilot clinical trial on 25 patients with major depressive disorder, with an average age of 50.28 years (std: 10.07 yrs, max: 66 yrs and min: 22 yrs). Three patients data was excluded due to software errors (which led to loss of data). The patients were grouped for analysis by first week PHQ9 score into two groups:  (i) Normal/mild group: phq9<=14, and (ii) Moderate/severe group: phq9>14.

After a 2-month data collection for each subject, we studied the correlation between smartphone data and psychometric scores. The results showed that the adherence rate to the daily self-reported mood input was about 82%, and the attendance rate for clinic visits was 95%. The correlation between self-reported mood score averaged over a 2-week period and the bi-weekly PHQ-9 score is 0.73 in the moderate/severe group and 0.36 in the normal/mild group. The passive phone sensor and usage data, including their number of steps, quantity of text messages and the amount of time spent messaging, also correlate with clinical assessments. For example, when depression worsens, the frequency of phone calls and text messages will drop.


  1. The preliminary findings indicate that the SOLVD app could be a reliable approach to tracking moderate-to-severe depression.
  2. Informal finding: The patients (average age of 50+ years) were comfortable with smartphone tracking, especially if only their doctor can view the summary statistics.