Author + information
- Zhen Zhou, MD∗ ()
- ↵∗Menzies Institute for Medical Research, University of Tasmania, 17 Liverpool street, Hobart, Tasmania 7000, Australia
I have read with great interest the paper by Kim et al. (1) that reported a significant association between the use of off-label direct oral anticoagulants (DOACs) and reduced rates of ischemic stroke and systemic embolism in patients with mitral stenosis and atrial fibrillation (AF) compared with the use of conventional warfarin treatment. Despite the important findings, several issues related to the study methods need to be highlighted and further improved.
First, most of the baseline covariates in this study were not necessarily constant through the entire follow-up (referred to as time-dependent covariates), including hypertension, diabetes, congestive heart failure, previous vascular disease, dyslipidemia, chronic obstructive pulmonary disease, and chronic kidney disease. It is necessary to adjust for these time-dependent covariates in the Cox model because they are relevant to the study endpoints.
Second, the investigators explained that the use of off-label DOACs was in part due to the bleeding events that occurred during warfarin use or the diagnosis of mild mitral stenosis. It is likely that a patient who experienced a bleeding event by using warfarin was more likely to experience intracranial hemorrhage with the subsequent use of DOACs compared with those who received DOACs but who never experienced a warfarin-induced bleeding event before. Therefore, the risk of DOACs might have been overestimated. Equally, the efficacy of DOACs on thromboembolism and all-cause mortality might also have been overestimated because patients who received DOACs were more likely to have mild mitral stenosis.
Furthermore, Kim et al. (1) did not describe in their methods about how they dealt with the patients who switched or discontinued medication during the follow-up period in the Cox models. Because both treatment switching and discontinuation might have affected the study results, it is necessary to specify these details.
In addition, there are 2 minor errors in the paper by Kim et al. (1). As per the International Classification of Diseases-10th Revision, code I60 is defined as intracerebral hemorrhage rather than intracranial hemorrhage (I60, I61, I62). Similarly, ischemic stroke is coded as I63 rather than I64 (stroke not specified as hemorrhage or infarction) and I67 (other cerebrovascular diseases) (2).
Please note: Dr. Zhou has reported that she has no relationships relevant to the contents of this paper to disclose.
- 2019 American College of Cardiology Foundation
- Kim J.Y.,
- Kim S.H.,
- Myong J.P.,
- et al.
- ↵(2016) ICD-10 Version: 2016. World Health Organization. Available at:. https://icd.who.int/browse10/2016/en. 342019. ICD-10 Version: 2016. World Health Organization, 2016. Available at: https://icd.who.int/browse10/2016/en. Accessed April 3, 2019..