Toward the third generation of negative symptom assessment in schizophrenia: A systematic review of digital...

Interview-based clinical rating scales are the primary means of assessing negative symptoms in schizophrenia (SZ). However, this methodology has several inherent issues that limit their utility. New methodologies, referred to as digital phenotyping, are emerging in mobile assessment that address these limitations. In this systematic review, we synthesize the literature on the use of digital phenotyping to measure negative symptoms in SZ.

We reviewed available research from the past 27 years on active and passive digital phenotyping in SZ extracted from Pubmed, Web of Science, and PsychInfo. 129 articles were included that evaluated adherence, face and construct validity, convergent and discriminant validity, criterion validity, internal consistency, and temporal stability.

Active and passive digital phenotyping measures have inherent face and construct validity. Criterion validity has been demonstrated via differences between SZ and healthy control groups. Convergent and discriminant validity have been demonstrated in relation to the same/different construct using digital phenotyping and clinical measures. Internal consistency and temporal stability differ across measures, ranging from poor to excellent. Adherence, tolerability, and feasibility are acceptable and generally not associated with negative symptoms.

Active and passive digital phenotyping measures demonstrate promise for measuring negative symptoms in ways that overcome limitations of clinical rating scales. However, the psychometric validation studies conducted to date have been completed in small samples, used disparate methodologies, and not evaluated the full range of negative symptoms or technologies that may be relevant. Recommendations are provided to facilitate progress.