Validating the working alliance inventory in a mobile health smoking cessation program: A quasi-experimental study...

The therapeutic alliance has been identified as a key factor influencing smoking cessation success. This study aimed to evaluate the psychometric properties - internal consistency, factorial validity, and concurrent validity - of the Spanish version of the Working Alliance Inventory-Short (WAI-S) in a mobile health (mHealth) smoking cessation program for Mexican adults who smoke. Additionally, it examined the association between WAI-S score, smoking-related outcomes and program satisfaction.

A quasi-experimental study was conducted in Mexico between June and October 2021 with 100 adults who smoke intending to quit. The 12-week mHealth cessation program, grounded in Social Cognitive Theory, combined automated text messages and tailored counselor feedback. At program completion, 80 participants completed the 12-item WAI-S. Psychometric analyses included exploratory factor analysis, Kaiser-Meyer-Olkin and Bartlett's tests, Cronbach's alpha, and item-total correlations. Logistic regression models assessed the association of WAI-S scores with program satisfaction and smoking cessation outcomes.

Exploratory factor analysis (n=80) revealed that two negatively worded items weakened internal consistency; their removal produced a refined 10-item scale with a robust two-factor structure and excellent reliability (Cronbach's α=0.91). Higher reported therapeutic alliance scores were associated with greater odds of satisfaction with the overall program (AOR=1.12; 95% CI: 1.04-1.21), the digital application (AOR=1.10; 95% CI: 1.02-1.19), and the text message content (AOR=1.10; 95% CI: 1.03-1.18), compared with participants reporting lower alliance scores. No significant association emerged between WAI-S scores and self-reported or biochemically verified smoking abstinence.

The 10-item Spanish WAI-S demonstrated strong psychometric validity for evaluating therapeutic alliance in an mHealth smoking cessation among Mexican adults. While not predictive of abstinence, higher alliance score correlated with greater satisfaction, underscoring the instrument's potential for monitoring engagement and informing the design of more effective digital cessation programs.