Tanvi Ranjan1, Jennifer Melcher2, Matcheri Keshavan2, Maurice Smith3, John Torous4. 1. John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA; Division of Digital Psychiatry, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02115, USA. 2. Division of Digital Psychiatry, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02115, USA. 3. John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA. 4. Division of Digital Psychiatry, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02115, USA. Electronic address: jtorous@bidmc.harvard.edu.
Abstract
BACKGROUND: Smartphone assessments and sensors offer the ability to easily assess symptoms across environments in a naturalistic and longitudinal manner. However, the value of this new data to make inferences about personal vs population health and the role of environment in moderating symptoms in schizophrenia has not been fully explored in a scalable and reproducible manner. METHODS: Eighty-six adults with a diagnosis of schizophrenia were recruited from the Greater Boston Area between August 2019 and May 2021. Using the open-source mindLAMP app in an observational manner, smartphone surveys and sensors (GPS, accelerometer, screen on/off and call and text logs) were collected for up to six months. RESULTS: Sixty-three participants were analyzed, who had at least completed one survey in the app. App-based self-reported symptom surveys were highly correlated with scores on gold standard clinical assessments (r = 0.80, p = 10-11 for mood and r = 0.78, p = 10-12 for anxiety). For these app-based assessments, inter-individual differences account for a larger proportion of the correlations in longitudinal symptoms as compared to intra-individual differences. Mood, sleep, and psychosis symptoms reported on app surveys were more severe when taken at home as determined by the smartphone's GPS sensor. DISCUSSION: The intra-individual symptom correlations and the stratification of symptoms by home-time highlight the utility of digital phenotyping methods as a diagnostic tool, as well as the potential for personalized psychiatric treatment building on this data.
BACKGROUND: Smartphone assessments and sensors offer the ability to easily assess symptoms across environments in a naturalistic and longitudinal manner. However, the value of this new data to make inferences about personal vs population health and the role of environment in moderating symptoms in schizophrenia has not been fully explored in a scalable and reproducible manner. METHODS: Eighty-six adults with a diagnosis of schizophrenia were recruited from the Greater Boston Area between August 2019 and May 2021. Using the open-source mindLAMP app in an observational manner, smartphone surveys and sensors (GPS, accelerometer, screen on/off and call and text logs) were collected for up to six months. RESULTS: Sixty-three participants were analyzed, who had at least completed one survey in the app. App-based self-reported symptom surveys were highly correlated with scores on gold standard clinical assessments (r = 0.80, p = 10-11 for mood and r = 0.78, p = 10-12 for anxiety). For these app-based assessments, inter-individual differences account for a larger proportion of the correlations in longitudinal symptoms as compared to intra-individual differences. Mood, sleep, and psychosis symptoms reported on app surveys were more severe when taken at home as determined by the smartphone's GPS sensor. DISCUSSION: The intra-individual symptom correlations and the stratification of symptoms by home-time highlight the utility of digital phenotyping methods as a diagnostic tool, as well as the potential for personalized psychiatric treatment building on this data.