Prevalence of Medication Nonadherence among Osteoporosis Patients: A Cross-Sectional Study in Tabuk, Saudi Arabia

Abstract Background: Despite the wide use of effective anti-osteoporotic therapies, adherence remains suboptimal worldwide, including in Saudi Arabia. Identifying factors associated with nonadherence—such as medication modality and patient characteristics—is vital for informing effective interventions. Objectives: This study aimed to assess the prevalence of osteoporosis medication nonadherence and to examine clinical, behavioral, and treatment-related predictors, including drug type and dosing regimen, among adult patients in Tabuk, Saudi Arabia. Methods: We conducted a cross-sectional study of 137 adults with osteoporosis receiving anti-osteoporotic treatment at a tertiary care center in Tabuk, Saudi Arabia. The data collected included demographic characteristics, comorbidities, concurrent medications, history of fragility fractures, medication type (oral vs. injectable), and dosing regimen. Adherence was assessed using the validated Morisky Medication Adherence Scale (MMAS-8), administered via structured patient interviews in person or via telephone, and verified against medical records. Nonadherence was defined as a <80% medication possession ratio. Multivariable logistic regression with an 80/20 training-test split was performed to develop a risk stratification model. A clinical nomogram was created for practical application of the model. Results: The prevalence of medication nonadherence was 29.9% (95% CI: 22.5–38.2%). The area under the ROC curve of the proposed combined risk model was 0.704 (95% CI: 0.454–0.921). The model successfully stratified patients into three risk categories: high (2.2% of the cohort, 0% adherence), moderate (35.8%, 61.2% adherence), and low risk (62.0%, 77.6% adherence). At the optimal cutoff, the model yielded 73.7% sensitivity and 75.0% specificity. A clinical nomogram was developed for bedside risk assessment. Conclusion: Nearly one-third of osteoporosis patients were nonadherent, indicating a persistent care gap. Although no individual clinical factors independently predicted nonadherence, our risk stratification model showed good discriminability (AUC 0.704) and separated patients into three risk groups with adherence rates ranging from 0% to 77.6%. The clinical nomogram provides a practical tool for rapid bedside risk assessment, enabling 2 targeted allocation of adherence support resources. Multicenter validation and intervention studies are needed before wider use of the proposed model.