Abstract:

The COVID-19 pandemic has led to a shift from in-person face-to-face counseling (F2F) to online videoconferencing counseling (VCP), which poses the question: how does VCP affect treatment outcomes compared to F2F? Existing research has demonstrated the equivalence of VCP and F2F in terms of effectiveness. However, the working alliance, a key common factor in F2F, has been found to be lower in quality in VCP than in F2F in a recent meta-analysis. Moreover, only one study has examined the reciprocal relationship between working alliance and treatment outcomes in VCP at the within-patient level. The present study aims to (a) compare the treatment outcomes between VCP and F2F using longitudinal data from a naturalistic setting; and (b) explore the mutual influence of working alliance and treatment outcomes in VCP and F2F at the within-patient level.

This study was conducted in a counseling center of a university in central China, and participants were arranged to receive VCP or F2F. The final sample consisted of 525 college students, of whom 117 received VCP and 408 received F2F. The only difference between the two conditions was the mode of delivery (VCP vs. F2F). Participants completed the CORE-OM-10 before each session and the Session Alliance Inventory (SAI) after each session. They also completed the PHQ-9, GAD-7, and CORE-OM-34 at pre- and post-treatment. The data from sessions 1 to 6 were analyzed using the Random Intercept Cross-Lagged Panel Model (RI-CLPM). A multi-group RI-CLPM comparison was conducted to examine the alliance-outcome relationship in VCP and F2F at the within-patient level.

The within-patient analysis (Figure 1) revealed that SAI was a significant predictor of CORE-OM in the subsequent session (β = -0.07, SE = 0.03, p = 0.02), and CORE-OM was a significant predictor of SAI in the same session (β = -0.07, SE = 0.03, p = 0.03). The multi-group comparison (Figure 2) indicated that the predictive effect of SAI on CORE-OM did not differ significantly between VCP and F2F (b = -0.05, se = 0.14, p = 0.724); the predictive effect of CORE-OM on SAI had no significant difference between tow groups either (b = -0.01, se = 0.03, p = 0.645). However, the working alliance quality in VCP was significantly lower than that in F2F after the first (M_diff = 0.77, SE = 0.35, 95% CI [0.08,1.46], t(523) = 2.18, p = 0.03) and the fourth sessions (M_diff = 0.74, SE = 0.37, 95% CI [0.02,1.46], t(442) = 2.03, p = 0.04), but not after the other sessions. The post-treatment analysis, using Propensity Score Matching with pretest CORE-OM34, PHQ-9 and GAD-7 as predictor variables, showed no significant difference between VCP (N = 89) and F2F (N = 330), in PHQ-9 (M_diff = -0.19, SE = 0.56, t(419) = -0.34), GAD-7 (M_diff = -0.68, SE = 0.45, t(419) = -1.51), and CORE-OM34 (M_diff = -2.30, SE = 2.36, t(419) = -0.97).

These findings indicate that VCP is as effective as F2F in reducing psychological distress, and that clients can establish a stable working alliance in VCP over time, even if they initially experience difficulties in adapting to the online mode. Moreover, the reciprocal influence of working alliance and treatment outcomes in VCP is similar to that in F2F. This study offers empirical support for the use of VCP, especially in the context of the COVID-19 pandemic.

Key words: online videoconferencing counseling (VCP), in-person face-to-face counseling (F2F), working alliance, treatment outcomes, Random Intercept Cross-Lagged Panel Model (RI-CLPM)