Statisticians have been keen to critique statistical aspects of the “replication crisis” in other scientific disciplines. But new statistical tools are often published and promoted without any thought to replicability. This needs to change, argue Anne-Laure Boulesteix, Sabine Hoffmann, Alethea Charlton and Heidi Seibold

Nonadherence to medicines is a widely acknowledged problem that limits the effectiveness of treatments and results in poor patient outcomes, including increased mortality, 1-4 hospitalization, 5,6 and higher healthcare costs. [6][7][8] Direct oral anticoagulants (DOACs) are increasingly used for the prevention of stroke in patients with nonvalvular atrial fibrillation (NVAF). 9,10 In addition to the potential clinical implications of nonadherence, when compared to vitamin K antagonists (VKAs), 11,12 adherence to DOACs is a subject of particular interest in light of their reduced monitoring requirements and shorter half-lives.The use of pharmacy claims databases has expanded our ability to conduct large-scale real-world adherence studies.Thus far, however, drug utilization research on DOACs has mainly focused on other adherence measures such as persistence, [13][14][15][16][17][18][19] proportion of days covered, 16,17,[20][21][22][23][24] or medicine possession ratio. 15,25,26 Primary nonadherence, defined as failing to collect the first prescription of a medicine, is the first step to achieving effective treatment. Estimates of primary nonadherence to medicines vary widely. 27 One large study found that 28.3% of patients were primary nonadherent to new medicines, with medication class as a strong predictor of primary adherence. The study also found that medicines for chronic diseases have particularly poor primary adherence rates. 28 Primary nonadherence has been associated with worse patient outcomes, including higher rates of mortality following myocardial infarction 2 and worse control of hemoglobin A1c in diabetes. 29 The reasons for not collecting a prescription may be complex, with possible causes including cost, concern about the medicine, and the perception that the medicine is not necessary. 27 Primary nonadherence to DOACs remains relatively unexplored. Although, one recent study found the rate of nonadherence to DOACs to be 10.8%. 30 The same study, which included VKAs and DOACs, found that primary nonadherence was associated with the type of anticoagulant prescribed, foreign nationality, dementia, and higher coinsurance in older age groups. Lower odds of primary nonadherence were found for patients with liver disease, those taking over six concomitant medicines, or whose prescriptions were transmitted electronically.Information about the rate of primary nonadherence to DOACs can help elucidate the real-world cost-effectiveness of these medicines and highlight a key area for adherence interventions. Furthermore, identifying the determinants of nonadherence can serve as the basis for designing interventions to improve adherence and patient outcomes. This study uses pharmacy claims data from a large primary care database in Catalonia to measure the rate of primary nonadherence to newly prescribed DOACs in NVAF patients and to identify associated factors. Methods DATA SOURCEAll data for this study were obtained from the Information System for the Development of Research in Primary Care (SIDIAP), which co...

Statisticians play a key role in almost all scientific research. As such, they may be key to solving the reproducibility crisis. Heidi Seibold, Alethea Charlton, Anne‐Laure Boulesteix and Sabine Hoffmann urge statisticians to take an active role in promoting more credible science

Statisticians play a key role in almost all scientific research. But are they also the key to solving the reproducibility crisis? Heidi Seibold, Alethea Charlton, Anne-Laure Boulesteix and Sabine Hoffmann urge statisticians to take an active role in promoting more replicable and more credible science.