SNAP AF-52 is an ambidirectional observational registry conducted in family health centers (primary care) across Ordu, Türkiye, enrolling adults ≥65 years with a pre-existing diagnosis of atrial fibrillation. The study assesses (a) label-concordant dosing of oral anticoagulants (DOACs/warfarin) using drug-specific criteria, and (b) medication adherence via Proportion of Days Covered (PDC) over the prior 12 months (good adherence defined as PDC ≥80%). Unsafe findings (e.g., suspected under-/overdosing, critical drug-drug interactions, very low renal function) trigger same-day referral to tertiary cardiology for evaluation and management. The retrospective window is Dec 1, 2024-Nov 30, 2025; the prospective single-visit inclusion window is Dec 1, 2025-May 31, 2026. No experimental treatment is administered; all care is routine.

Family physicians will screen their ≥65-year attendees with documented AF and record a minimum dataset during a single routine visit: demographics, body weight, most recent serum creatinine (date/value) to compute Cockcroft-Gault creatinine clearance (CrCl), current oral anticoagulant and regimen, common interacting drugs, and pharmacy dispensing dates with days' supply for the last 12 months. The system auto-classifies dose appropriateness per drug label criteria (including apixaban reduction rules; rivaroxaban/dabigatran/edoxaban CrCl thresholds and P-gp/CYP3A4 interactions) and computes PDC with carry-over of early refills; primary nonadherence is defined as no fill within 30 days of the first prescription. Retrospective capture of events includes ischemic stroke/TIA/systemic embolic event (SEE) and hemorrhagic stroke/ISTH major bleeding, plus on-treatment status at the event date.

Participants with potentially unsafe dosing or clinically high-risk interaction (e.g., strong inducers; strong dual inhibitors) or CrCl <15 mL/min are same-day referred to tertiary cardiology. Identifiable data will be stored securely per ethics approval and Turkish data protection (KVKK); only aggregated results will be reported.

Inclusion Criteria:

• Age ≥65 years.
• Established atrial fibrillation (any type) documented in the medical record.
• Presenting to a family health center (primary care) in Ordu province during the prospective window or having records within the retrospective window.
• On an oral anticoagulant (DOAC) at any time in the 12-month window.

Exclusion Criteria:

• Mechanical prosthetic heart valve or moderate-to-severe rheumatic mitral stenosis
• Participation in an interventional drug study affecting anticoagulation
• Lack of minimal data elements required for dose classification (age, sex, body weight, serum creatinine with date, current OAC and regimen)
• Investigator judgment that reliable data cannot be obtained from records

Adults aged ≥65 years with established atrial fibrillation (any type) receiving care in family health centers across Ordu, Türkiye. The prospective cohort enrolls consecutive eligible attendees during Dec 1, 2025-May 31, 2026 at a single routine visit; the retrospective cohort includes patients with AF who received an oral anticoagulant during Dec 15, 2025-May 31, 2026 identified from primary care records. Exclusions: mechanical prosthetic heart valve or moderate-severe rheumatic mitral stenosis, and cases lacking the minimal data needed for dose/adherence classification. No investigational treatment is delivered; care is routine with same-day tertiary referral when unsafe dosing, high-risk drug interactions, or very low renal function are detected.

Individual participant data will not be shared because identifiers are retained in the working dataset to enable same-day clinical referral and safety follow-up, and the ethics approval (and patient permissions) limit use to quality-of-care evaluation within participating centers. De-identification sufficient for public sharing would require extensive linkage files and could still pose re-identification risk in a single-province cohort under local data-protection law (KVKK). We will share only aggregated, de-identified results (tables/figures) upon reasonable request and after ethics approval; the study protocol, CRF, and PDC calculation SOP can be provided as supporting documents.