A Predictive Model for Patient-Centered Informed Consent in Dentistry

Thesis Abstract

This study addresses a critical gap in dentistry by examining how patient-centered informed consent can be effectively operationalized through a predictive modeling approach that integrates clinical, communicative, and psychosocial factors to enhance patient understanding, autonomy, and treatment adherence. The aim is to develop and validate a predictive model that accurately forecasts patient comprehension, satisfaction, and engagement with the informed consent process across diverse dental procedures. Specific objectives include (1) identifying key determinants of patient-centered informed consent from patient, clinician, and contextual dimensions; (2) developing a theoretical framework that links communication theories with consent outcomes using constructs from the Theory of Planned Behavior and the Health Belief Model; (3) constructing a predictive model using hierarchical logistic regression and machine learning techniques to classify cases with high versus low consent quality; (4) validating the model across multiple dental clinics representing private, public, and teaching hospitals; and (5) proposing an evidence-based protocol for implementing patient-centered consent practices in routine dental care. The methodology adopts a mixed-methods sequential explanatory design. The research will be conducted in three dental care networks within a metropolitan region, encompassing 12 clinics. The target population includes adult patients (n = 1,200) undergoing procedurally informed treatment planning and dental clinicians (n = 120) who obtain consent. A stratified random sample will be drawn to include adults aged 18–75, with proportional representation of age, gender, education, and socio-economic status, ensuring inclusion of vulnerable groups. Data collection instruments comprise (a) a patient questionnaire capturing comprehension of procedures, perceived autonomy, satisfaction with the consent discussion, and demographic variables; (b) a clinician checklist capturing consent delivery practices, use of explainable aids, and time allocated for discussion; (c) a standardized informed consent quality index scored by independent observers using audiorecordings of consent conversations; (d) qualitative semi-structured interviews with a purposive subsample of 40 patients and 20 clinicians to explore perceptions and barriers. Instrument validity will be established through content validity indexing with a panel of five dental ethicists and clinicians, and reliability will be assessed using Cronbach’s alpha and inter-rater reliability (? > 0.80 for consent coding). Data analysis will proceed in two stages. Quantitative data will be analyzed using descriptive statistics, bivariate correlations, and hierarchical logistic regression to identify independent predictors of high-quality consent, followed by machine learning approaches (random forest and gradient boosting) to optimize predictive accuracy and accommodation of non-linear relationships. Model performance will be evaluated using AUC-ROC, calibrations plots, Brier scores, and k-fold cross-validation. The qualitative data will be analyzed using thematic analysis, coded by two researchers and reconciled to yield themes related to communication strategies, information transparency, and patient empowerment; findings will be triangulated with quantitative results to enhance interpretability and model face validity. The theoretical framework integrates elements from the Theory of Planned Behavior, the Health Belief Model, and Communication Accommodation Theory to explain how patient and clinician variables interact to influence consent quality, culminating in a simplified conceptual model of patient-centered informed consent in dentistry. Expected findings include that patient literacy, use of visual aids, clinician communication style, consultation length, and procedural complexity collectively predict higher consent quality, with the predictive model achieving an AUC-ROC of at least 0.85 and good calibration. The study anticipates differential effects across patient subgroups, suggesting the need for tailored communication strategies for older adults, individuals with limited health literacy, and linguistically diverse patients. The contribution to knowledge lies in providing a rigorously tested, generalizable predictive framework that integrates theory-driven constructs with empirical evidence to guide the design of patient-centered consent processes in dental settings, as well as a practical protocol for clinics to implement enhanced consent practices, including training modules and decision-support tools. The main conclusion is that patient-centered informed consent in dentistry can be significantly improved through a validated predictive model that informs targeted communication interventions and supports shared decision-making. Recommendations include integrating the model into electronic health records to prompt clinicians, developing standardized consent aids with adjustable levels of information, and conducting ongoing audits of consent quality to sustain improvements across diverse dental care contexts.