Study Design
This retrospective, non-interventional study was based on data from existing health insurance claims in Japan. The Japanese Medical Data Vision (MDV) clinical database was used. The MDV database contains data from hospital information systems and administrative claims covering more than 12.94 million patients attending 280 acute hospitals as in- or out-patients within secondary medical care blocks in Japan. These hospitals used the Diagnosis Procedure Combination (DPC) case-mix classification system for inpatient reimbursement claims. Approximately 1500 hospitals are designated as DPC hospitals in Japan [17]. The database contains anonymous information from health insurance claims for outpatients, DPC claims for inpatients, prescriptions, operations and medical procedures, hospitalization and results of laboratory tests from some of the participating hospitals [18].
The study protocol was reviewed and approved by the Keio University’s Faculty of Pharmacy Ethics Committee for Research involving humans, which also waived the informed consent requirements on October 11, 2017 in accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects and the Declaration of Helsinki (1964) and its later amendments. The study was registered with ClinicalTrials.gov (NCT03254147).
OAC-naive adult patients with NVAF who initiated anticoagulant therapy with dabigatran, warfarin, apixaban, rivaroxaban, or edoxaban between March 14, 2011 and June 30, 2016 were included in the study. The main inclusion criteria were: patients (age > 18 years) with confirmed diagnosis of NVAF (International Classification of Diseases [ICD] 10 I48), having a first prescription (index date) of any one of the OACs (dabigatran, warfarin, rivaroxaban, apixaban, or edoxaban), and having no prescription of OACs for 6 months prior to the index date (baseline period), as well as enrolment in the MDV database for at least 6 months. The main exclusion criteria were: patients on OAC therapy and patients who underwent dialysis or kidney transplant during the baseline period; patients with atrial flutter, valvular atrial fibrillation (AF), mechanical valve placement, rheumatic AF, and/or mitral valve prolapse/regurgitation/stenosis during the baseline period; patients with a record of deep vein thrombosis or pulmonary embolism within 6 months of AF diagnosis during the baseline period; and patients receiving treatment with two or more OACs at index date.
To calculate the incidence rate of clinical events during OAC treatment, treatment duration was expressed in patient-years of follow-up. The follow-up period was defined from the day after the index date to the date of any OAC treatment discontinuation, end of continuous enrolment in the database, end of the study period, first occurrence of an event of interest, or death, whichever came earlier. On OAC observation time after switching from one OAC to another was included in the primary analysis, and a sensitivity analysis was undertaken that censored for events after the switch.
Patients were stratified on the basis of prescribed OAC and age (≤ 64 years, 65–74 years and ≥ 75 years).
Outcomes
The primary objective of the study was to assess the overall and age-stratified incidence rates of emergency surgery and major bleeding due to fracture and trauma in adult patients initiating OAC therapy for NVAF. Emergency surgery was defined as any surgical procedure (ICD10 code K000-879) performed on the day of hospital admission with additional claims associated with urgent fee MDV codes (114701370, 160000210, 180709110). Major bleeding due to fracture was defined as any bleeding associated with hospitalization or blood transfusion (ICD10 code K920) in patients with fracture (ICD 10 codes M484, M80, S021, S12, S22, S32, S42, S52, S62, S72, S82, S92, T02, T08, T10, T12, T142) listed as the primary diagnosis (i.e., the diagnosis requiring the most medical resource). Major bleeding due to trauma was defined as any bleeding associated with hospitalization or blood transfusion (ICD10 code K920) in patients with any kind of trauma (ICD 10 codes S00-S09, I620, M125, M164, M165, M172, M173, M191, M483, M872, S065, S066, S081, S089, S130, S18, S230, S480, S481, S532, S533, S580, S581, S589, S633, S634, S680, S681, S684, S688, S689, S780, S789, S889, S980, S981, S982, T050, T053, T058, T059, T116, T136, T147, T794, T796, T797, S271, S272, S330, S334, S382), listed as the primary diagnosis.
The secondary objective of the study was to assess the overall and age-stratified incidence rates of cardiac tamponade and pericardiocentesis. All cardiac tamponade (ICD 10 code 4200001) and pericardiocentesis cases (MDV procedure code 140010510) on the same or next day as catheter ablation or percutaneous coronary intervention (MDV procedure codes 150153910, 150267810, 150263310, 150284310, 150303310, 150345710, 150374910, 150375010, 150375210, 150375310, 150375410) were included. The types of emergency surgeries were also identified in the study population.
Statistical Analysis
All MDV data used during the study were in electronic format and were managed by Milliman Inc. Statistical analysis was performed using SAS version 9.4 and Microsoft Excel 2010. Descriptive statistical methods were used for analyzing baseline and outcomes data. The primary outcomes, further outcomes and patient characteristics were determined both in total and in age-stratified groups.