From Omnichannel Journey Design to Recommendation Intention: Psychological Needs and Trust in Dental Care

Introduction

Digitalization is steadily dissolving the boundary between online and on-site channels in healthcare delivery and reshaping patient expectations regarding the seamlessness, transparency, and personalization of the care journey [1-3]. Within Vietnam's private healthcare sector, dentistry is among the segments with the fastest pace of digital transformation. A patient may book through an application, receive advice through Zalo or a Facebook page, obtain appointment reminders and post-treatment guidance by message, and then attend the in-chair visit, with every contact linked inside one integrated journey [4-6]. As the number of touchpoints grows, each touchpoint becomes both an opportunity to strengthen the brand relationship and a risk that may erode it when consistency is lacking. Because dental care couples high perceived risk with strong dependence on professional judgment, patient trust has long been treated as central to the dentist-patient relationship [7-9].

This research utilizes customer journey design—defined as the strategic orchestration of touchpoints spanning from initial information gathering to post-service engagement—as a framework to analyze these professional dynamics [2, 10, 11]. Within a dental context, patients navigate a tripartite risk structure: physical hazards from invasive procedures, psychological distress regarding dental phobias or overtreatment, and informational vulnerabilities stemming from significant expertise disparities between practitioners and patients. Such compounding risks necessitate trust as a fundamental prerequisite for sustained patient loyalty and advocacy [12, 13]. However, scholarship concerning customer journey design remains heavily skewed toward retail and e-commerce sectors [2, 7, 10]. Consequently, there exists a notable deficit in empirical evidence regarding the internal psychological drivers influencing omnichannel healthcare journeys, particularly within the distinct socioeconomic landscapes of emerging markets that require further academic exploration and rigorous investigation.

To bridge this gap, this research utilizes the Stimulus-Organism-Response framework [14, 15] as a foundational architecture, incorporating self-determination theory’s three psychological needs—autonomy, competence, and relatedness [16, 17]—alongside commitment-trust theory [18, 19]. By positioning trust centrally within the causal sequence from journey attributes to recommendation intention, we frame design features as stimuli, psychological satisfaction and trust as internal organism states, and behavioral intent as the response. Consequently, this study provides a patient-centered perspective, shifting the focus away from traditional provider-centric models to better understand the dynamics of patient experience and advocacy.

On this basis, the study pursues three objectives: to identify which omnichannel journey design attributes satisfy each psychological need of dental patients; to test the mechanism through which the three needs build trust and, through trust, lead to recommendation intention; and to establish the relative weight of the three needs so that managerial priorities can be focused. The remainder of the paper presents the theoretical background and hypotheses, the materials and methods, the combined results and discussion, and the conclusion.

Theoretical background and hypotheses

S-O-R framework and integrated model

The Stimulus-Organism-Response framework, originating in environmental psychology [14, 20], explains behavior through three layers: an environmental stimulus acts on the internal state of the organism, which then gives rise to a behavioral response. The framework has been applied widely in research on digital behavior, including influencer marketing [21, 22] and the omnichannel customer journey [2, 23], and recent work has extended it to health-related behavior [24, 25]. The present study brings the framework into dental services and integrates self-determination theory with commitment-trust theory to specify the organism layer: the three need satisfactions operate as intrinsic motivational states, whereas trust in the clinic operates as a relational state that crystallizes from those satisfactions. To keep the data structure consistent with this theoretical layering, every indicator was recoded with one of three prefixes: ST_ for the stimulus set, OG_ for the organism set, and RS_ for the response.

Stimuli and autonomy satisfaction

Within self-determination theory, autonomy is characterized as the fundamental necessity for individuals to perceive themselves as the authors of their own behaviors, possessing authentic agency [17, 26]. In clinical settings, fostering autonomy by providing transparent information, acknowledging patient viewpoints, and offering viable decision-making opportunities correlates with increased psychological need fulfillment and superior health results [27, 28]. Three specific journey attributes address this need through varied mechanisms. Transparency (ST_TRA)—the reliable conveyance of costs, care pathways, and provider qualifications via digital platforms—minimizes ambiguity, enabling informed patient choices [29, 30]. Personalization (ST_PER), involving the customization of guidance and alerts based on unique medical histories, bolsters a sense of agency that reflects individual priorities [2, 31]. Finally, seamlessness (ST_SEA), which ensures data synchronization across digital and in-person touchpoints, empowers patients by eliminating repetitive disclosures, thereby maintaining a consistent experience of personal control [2, 32, 33].

H1. Transparency (ST_TRA) positively affects autonomy satisfaction (OG_AUT).

H2. Personalization (ST_PER) positively affects OG_AUT.

H3. Seamlessness (ST_SEA) positively affects OG_AUT.

Stimuli and competence satisfaction

Competence reflects the desire to feel effective and capable of mastering a situation [17], and it is especially salient in dentistry because of the pronounced information asymmetry. Evidence in healthcare indicates that competence support, achieved by communicating expectations clearly and providing guidance for self-management, matters as much as autonomy support during care interactions [27, 34]. Thematic coherence (ST_COH), the consistency of imagery and ethical commitments from the online space to the clinic, helps the patient form a coherent and confident understanding [35, 36]. Convenience (ST_CON), the simplification of booking, look-up, and payment on digital platforms, lowers cognitive load and raises the sense of mastering the journey [2, 37]. Complementarity (ST_COM), online self-care guidance that complements the clinician's chairside instructions, increases the patient's capacity for self-care [2, 38].

H4. Thematic coherence (ST_COH) positively affects competence satisfaction (OG_CMP).

H5. Convenience (ST_CON) positively affects OG_CMP.

H6. Complementarity (ST_COM) positively affects OG_CMP.

Stimuli and relatedness satisfaction

Relatedness is the need to feel connected, cared for, and that one belongs [17]. In care settings marked by high anxiety, the bond with the clinical team is closely tied to a sense of safety [9]. Interactivity (ST_INT), the real-time two-way exchange between the patient and the dental team through digital channels, does more than convey information—it creates the feeling of being heard and personally attended to [39-41] and thereby satisfies the need for relatedness.

H7. Interactivity (ST_INT) positively affects relatedness satisfaction (OG_REL).

Need satisfaction and trust

When a dental practice simultaneously fulfills the three fundamental psychological requirements, the clinician-patient dynamic transcends mere commercial exchange, evolving into a foundation of trust. Prior studies indicate that digital healthcare interactions significantly influence this perception, ultimately dictating a patient’s inclination toward future physical appointments [42-44]. According to the commitment-trust theory, trust serves as the cornerstone of enduring professional relationships, rooted in benevolence, integrity, and competence, acting as a critical asset for fostering long-term loyalty [18, 45, 46]. Within the dental environment, relatedness is anticipated to be the primary driver of trust; strong interpersonal connections with clinicians directly address this need [17], further supported by research linking relatedness to relational trust development [47, 48]. Although autonomy and competence remain positive contributors, they exert a comparatively minor influence, as the final diagnostic and technical authority naturally remains with the professional practitioner, limiting the patient’s role in clinical decision-making processes.

H8. OG_AUT positively affects trust in the clinic (OG_TRU).

H9. OG_CMP positively affects OG_TRU.

H10. OG_REL positively affects OG_TRU.

Trust and recommendation intention

Once trust is established, the patient is willing not only to return but also to vouch for the clinic through social media, review platforms, and word of mouth [22, 49]. Recommending a healthcare provider carries a high level of committed responsibility because the recommender stakes personal reputation and the health of close others; it is therefore activated only when trust reaches a sufficiently firm threshold [18]. This expectation is consistent with evidence that motivational, need-satisfying customer experiences drive engagement and advocacy [50, 51].

H11. OG_TRU positively affects service recommendation intention (RS_REC).

Conceptual model

Figure 1 depicts the theoretical model following the Stimulus-Organism-Response framework. At the stimulus layer, the seven journey design attributes are grouped into three clusters corresponding to the three needs (H1 to H7); at the organism layer, the three need satisfactions jointly direct toward trust (H8 to H10); and at the response layer, trust leads to recommendation intention (H11).

Materials and Methods

Study design and data collection

The study used a cross-sectional design with a quantitative survey. An omnichannel dental clinic was operationally defined as a private dental practice that simultaneously operates at least two distinct patient-facing digital channels, such as a booking application, a social messaging platform (e.g., Zalo or Facebook Messenger), or a clinic website, and integrates patient information and communication across these channels within a single service journey. The target population was patients who had used services at such clinics within the previous six months and who had interacted through at least one online channel combined with an on-site visit on the same service journey. Two screening questions at the start of the questionnaire confirmed this condition, and only respondents who satisfied both criteria proceeded to the main assessment.

Data were collected from patients of omnichannel dental clinics in Ho Chi Minh City between March and April 2026. Clinics were recruited through purposive sampling, targeting private practices that demonstrably operated across at least two digital touchpoints combined with on-site visits within an integrated journey. The online component was distributed via the clinics' social media channels and messaging platforms; the on-site component involved self-administered questionnaires completed by patients in clinic waiting areas. Of the 750 questionnaires distributed, 573 met the eligibility and completion criteria and were retained for analysis, yielding a valid response rate of 76.4%.

Among the 573 valid patients, women accounted for 57.9% (332) and men for 42.1% (241). The largest age group was 26 to 40 years (44.5%, 255), followed by 18 to 25 years (27.5%, 158), 41 to 55 years (20.3%, 116), and over 55 years (7.7%, 44). For education, 33.5% (192) held a postgraduate qualification, 53.1% (304) a university degree, and 13.4% (77) a high school diploma. Notably, 77.8% (446) confirmed interaction through at least two different channels during their most recent service journey, and 30.4% (174) had experienced three or more channels. Table 1 summarizes the sample profile (Table 1).

Table 1. Demographic profile of respondents.

Measures

The questionnaire comprised 36 indicators measuring 12 constructs on a five-point Likert scale (1 = strongly disagree; 5 = strongly agree). Scales were adapted from peer-reviewed sources and contextualized for the omnichannel dental clinic setting through independent back-translation by two bilingual experts, expert panel review, and a pilot test. Indicator codes carry the prefix ST_ (stimulus), OG_ (organism), or RS_ (response), consistent with the theoretical architecture.

Among the stimulus constructs, transparency (ST_TRA) was adapted from Gao et al. [29]; personalization (ST_PER), convenience (ST_CON), complementarity (ST_COM), and interactivity (ST_INT) were adapted from Zheng and Li [2]; seamlessness (ST_SEA) was adapted jointly from Gao et al. [29] and Chang and Li [32]; and thematic coherence (ST_COH) was adapted from Kuehnl et al. [36] and Zheng and Li [2].

The three basic psychological need satisfactions—autonomy (OG_AUT), competence (OG_CMP), and relatedness (OG_REL)—were each measured with three items adapted from Zheng and Li [2], who reconceptualized Thomson's [52] attachment items within a self-determination theory framework for omnichannel contexts. The present study further contextualized these items for the dental setting, aligning wording with the three layers of perceived risk—physical, psychological, and informational—consistent with operationalizations of basic psychological needs in service and technology research [51, 53, 54].

Trust in the clinic (OG_TRU) was adapted from Sirdeshmukh et al. [45], capturing credibility, benevolence, and integrity. Recommendation intention (RS_REC) was adapted from Xia and Bechwati [55] and Alcántara-Pilar et al. [22], both of which validated recommendation scales within service and digital contexts grounded in the same theoretical framework adopted here.

Table 2 lists all constructs, indicators, outer loadings, reliability statistics, predictive relevance (Q²), and source attributions.

Table 2. Measurement model: outer loadings, internal consistency, convergent validity, predictive relevance, and scale sources.

Analytical approach

To evaluate the research framework, partial least squares structural equation modeling (PLS-SEM) was performed using SmartPLS 4 [56]. This approach is ideal for the study, as it handles complex serial and parallel mediation pathways, bypasses strict normality requirements, and excels in predictive and explanatory tasks [57]. The analysis was executed in two phases. Initially, the measurement model was validated by examining internal consistency (Cronbach’s alpha and composite reliability ρc), convergent validity (average variance extracted), discriminant validity (Fornell-Larcker criterion and heterotrait-monotrait ratio [58]), and collinearity (variance inflation factor). Predictive relevance was confirmed using the Stone-Geisser Q² metric via a blindfolding procedure with an omission distance of seven [57]. Finally, the structural model was examined through path coefficients, determination coefficients (R²), effect sizes (f²), and bootstrap-derived significance levels utilizing 5,000 resamples alongside bias-corrected confidence intervals.

To address the potential threat of common method variance, all items were collected in a single survey administration. As a post-hoc diagnostic, a Harman's single factor test was performed by constraining an exploratory factor analysis of all 36 items to one factor; the single factor explained less than 50% of the total variance, falling below the commonly used threshold for problematic common method bias [57].

Results and Discussion

Measurement model

Table 2 summarizes the measurement model. The outer loadings of all 36 indicators ranged from 0.789 to 0.873, all above the 0.708 threshold [57]. Cronbach's alpha ranged from 0.776 to 0.819 and composite reliability (ρ_c) from 0.870 to 0.893, confirming internal consistency reliability. The average variance extracted ranged from 0.691 to 0.735, all above 0.50, establishing convergent validity.

Discriminant validity was confirmed in two ways. Under the Fornell-Larcker criterion, the square root of the average variance extracted of each construct exceeded its correlation with every other construct. Under the heterotrait-monotrait ratio reported in Table 3, the highest value was 0.625, between interactivity and relatedness satisfaction, well below the 0.85 threshold [58]. The variance inflation factors of the inner model ranged from 1.00 to 1.26, ruling out collinearity. The standardized root mean square residual of the estimated model was 0.048, below the 0.08 threshold, indicating acceptable model fit [57].

Table 3. Discriminant validity: heterotrait-monotrait ratio (HTMT)

Predictive relevance was assessed through blindfolding (omission distance = 7). Stone-Geisser Q² values exceeded zero for all endogenous constructs: OG_AUT = 0.199, OG_CMP = 0.172, OG_REL = 0.168, OG_TRU = 0.147, RS_REC = 0.147, confirming that the model has predictive relevance for each dependent construct [57]. These values fall in the weak-to-moderate range, consistent with the model's parsimonious specification.

Structural model

Table 4 reports the test of eleven hypotheses based on bootstrapping with 5,000 subsamples. All hypotheses were supported at a significance level of p ≤ 0.002, with 95% bias-corrected confidence intervals that excluded zero. The model explained 29.7% of the variance in OG_AUT, 25.1% in OG_CMP, 24.2% in OG_REL, 20.5% in OG_TRU, and 21.1% in RS_REC. These R² values fall in the moderate range by the benchmarks of Hair et al. [57]. The moderate explained variance for OG_TRU and RS_REC reflects the parsimonious specification of the model, which assigns a single stimulus cluster to each need and excludes known covariates such as dental anxiety severity and prior digital health experience.

Table 4. Structural model Results.