Telepsychiatry as reliable as face-to-face interviews for symptom assessment

Assessments of depressive, anxiety and psychotic symptoms and suicidal risk by telepsychiatry have good reliability in general when compared with face-to-face interviews, according to researchers from the Queen Mary Hospital (QMH).

“Since the COVID-19 pandemic, interest in clinical outcomes with telepsychiatry has gained momentum and new studies have been conducted,” wrote the researchers. “[However], limitations in sample size, methodology and information technology were found in previous studies, and the reliability of symptom assessment remained a concern.”

To evaluate the reliability of telepsychiatry in carrying out assessments, the researchers recruited 90 adult patients from psychiatry wards in QMH (median age, 39 years; female, 57.8 percent). For each participant, a telepsychiatry interview and a face-to-face interview were conducted within 2 days by two raters with similar psychiatry experience in public service setting. [Psychiatry Res 2022;doi: 10.1016/j.psychres.2022.114728]

The telepsychiatry interview was conducted via Zoom, and the distance between the patient’s seat and iPad was fixed at 45 cm to maximize the rater’s view of the patients’ facial expression and gestures over the upper part of the body.

“As suicidal risk assessment is a major component in psychiatric consultations, the reliability of assessing suicidal risk is a major focus of this study,” wrote the researchers. For Columbia Suicide Severity Rating Scale (C-SSRS), there was a nearly perfect agreement between telepsychiatry and face-to-face interviews in assessing suicidal risk.

The inter-method reliability of individual items in the Hamilton Depression Rating Scale (HRMS) and Hamilton Anxiety Rating Scale (HAM-A) ranged from 0.6–1 based on Kappa statistics, revealing moderate to almost perfect agreement between the two modalities in measuring depressive and anxiety symptoms. [Phys Ther 2005;85:257-268]

Notably, agitation and retardation showed poorer inter-method agreement vs other items, suggesting that telepsychiatry is less reliable for picking up subtle differences in agitation and retardation.

“Some participants were observed to show a lesser degree of agitated movement when they were facing the monitor. It appeared that they limited their actions intentionally to avoid their actions being out of the camera’s capture range,” explained the researchers. Additionally, most participants had a tendency to speak in a slower pace and louder volume during telepsychiatry interviews due to their distance from the monitor. Therefore, raters might score these items inaccurately during telepsychiatry interviews.

For inter-method reliability of assessing psychotic symptoms, all items in the Brief Psychiatric Rating Scale (BPRS) showed no significant difference (all p>0.05) except “blunt affect” (Z=-2.078; p=0.038) in the Wilcoxon signed-rank test. The reliability of other observation-based items, including emotional withdrawal and tension, are lower in telepsychiatry interviews based on the Spearman’s rho correlation coefficient.

“Overall, inter-method reliability in HDRS, HAM-A, C-SSRS and BPRS showed good agreement between telepsychiatry and face-to-face interviews,” summarized the researchers. “However, items based on observation, including agitation, retardation, emotional withdrawal, tension, and blunting of affect, showed relatively lower correlation when compared with those based on verbal report in general.”