This cross-sectional study was done during six months (between September 2014 to March 2015) at outpatient anticoagulant clinic of Shaheed Rajaie Hospital, Tehran, Iran. This is a well-known center for the treatment of cardiovascular diseases in Iran. Patients diagnosed with non-valvular atrial fibrillation were included in the study if they were between 30-85-years-old and had been taking warfarin for >3 months. Patients who did not want to participate in the study were excluded. All study participants signed consent forms after the study procedures were explained. Demographic data such as age, sex, educational level, medical history and current medications were determined for all participants. The INRs of patients were collected during their referral to the clinic where every patient had at least 3 INR measurements taken in total. Each patient’s TTR was calculated using the Rosendaal method. The Rosendaal linear interpolation methodology is based on the INR-DAY software program (Dr. F.R. Rosendaal, Leiden, The Netherlands) that assumes a linear relationship exists between two INR values and allows the researcher to allocate a specific INR value to each day for each patient [8].

Results are reported as mean ± SD. Data were analysed using the Statistical Package for the Social Sciences version 20 (SPSS-20), and p-values less than 0.05 were considered statistically significant. The continuous data obtained in this study were analysed using Chi-square test. Significant univariate predictors were subsequently assessed in the multivariate logistic regression model to determine their independent effect, expressed as Odds Ratio (OR) and 95% Confidence Interval (CI). The p <.05 was considered significant.

This study involved 470 patients receiving warfarin who participated for 6 months. The total number of INR measurements was 1450. The mean (±SD) TTR for 470 patients was 54.9 ± 11.9%. The stratification of patients according to TTR levels was conducted as follows: a TTR level of >70% was considered to represent good control, a TTR level between 50% and 70% was considered to represent intermediate control, and a TTR level of <50% was considered to represent poor control. The good control group contained 175 patients (37.3%), 116 patients (24.6%) were in the intermediate control group, and 179 patients (38.1%) were in the poor control group. The mean age of the sampled patients was 58.0 ± 14.2 years and 283 (60.2%) were women. The patient demographics have been shown in [Table/Fig-1]. As seen in [Table/Fig-1], there is no significant difference across the three TTR categories in terms of age, sex, educational level, or marital status (p = 0.42, 0.38, 0.43, and 0.87, respectively). The TTRs for patients with medical histories such as hypertension, deep vein thrombosis, pulmonary embolism, heart failure, diabetes mellitus, and coronary heart disease were determined. [Table/Fig-2] shows the distribution of sampled patients with medical histories in each category. Analysis showed that there were not any differences between the TTR categories and medical histories. [Table/Fig-3] shows the number of current medications taken in each category of studied patients. The rate of good control is >85% in patients who receive less than 4 medications, but in patients with ≥5 medications, the rate decreases to <20%. Analysis showed that there is a significant difference between TTR categories and number of current medications in the evaluated patients, where, by the number of poor control patients have increased as the number of medications taken increases. To evaluate the independent effects of each variable (including demographic parameters, medical histories, and number of current medications) as a predictor of poor control, we performed a multivariate logistic regression. We identified only numbers of medication >4 as significant predictors of poor control (OR = 2.06; 95% CI, 1.87, 2.23).

Vitamin k antagonists have been shown to be effective in the treatment and prevention of thromboembolic events; however they possess many drug- drug and drug-food interactions, as well as a narrow therapeutic window [10]. The efficacy and safety of oral vitamin K antagonists such as warfarin depend strongly on the percentage of TTR, with the maximum benefits being evident when the TTR is >70% [11,12]. It is well-known that poor control of anticoagulant intensity increases the risks of thrombotic and haemorrhagic events [9]. The consistency of an effective INR is reflected by the TTR, which is a measure of the period in which the patient was in an optimal INR range. Cotte et al., evaluated the TTRs of 6250 patients in four European countries (France, Germany, Italy, and United Kingdom) with atrial fibrillation who had been prescribed vitamin K antagonists. They concluded that 47.8%, 44.2%, 46.1%, and 65.4% of the evaluated patients had TTRs >70% in France, Germany, Italy, and the United Kingdom, respectively [13]. Our results showed that the percentage of good control patients (37.3%) was less than that of each European country as discussed by Cotte et al., [13].

Mark et al., [14] recently analysed data from 272 patients with non-valvular atrial fibrillation in a hospital in Hungary. They did not classify their patients into different TTR categories and only reported the mean TTR, which was found to be 64%. The mean TTR in our study (54.9%) was lower than that reported by Mark et al. It seems that Iranian patients have poorer control of warfarin dosing compared to the patients in European studies [14]. Our results were similar to those from recent research by Zubaid et al., in Kuwait. They evaluated the quality of warfarin therapy for 369 patients with non-valvular atrial fibrillation and estimated TTRs by the Rosendaal method. They reported a mean TTR of 52.6% in their sample, which is close to the mean TTR determined in the present study (54.9%) [15]. Pharmacogenetic and dietary regimens are two important factors to be considered in relation to warfarin [16–18]. Iran and Kuwait are located in western Asia and have similar dietary regimens, cultures and genetic patterns, which may provide reasons for why our results and those in Zubaid’s study are similar. Zubaid et al., had concluded that females and patients with no history of hypertension were more likely to have poor anticoagulation (expressed as Rosendaal TTR < 58%). Unlike Zubaid et al., we did not see any tendencies of poor control among females or patients without hypertension. Melamed et al., [19] studied TTR in 906 patients diagnosed with atrial fibrillation in the United States who were treated with warfarin for at least 6 months. They concluded that poor control (TTR < 60% in their study) was significantly associated with females, advanced age (>75 years), and heart failure [19]. However, in our study, there were no significant differences in TTR between male and females (p = 0.38), patients <75 years, those >75 years (p = 0.42), and patients with and without heart failure (p = 0.58). Previous studies have not referred to the relationship between TTR and the number of patient’s medications. In the present study, a significant decrease in good control rates had been observed when the patients were prescribed more than 4 medications. It seems that the number of prescribed medications may be an important factor influencing patient adherence to warfarin therapy that can affect TTR indirectly. Zulling et al., evaluated adherence barriers among patients with cardiovascular risk factors. The most commonly reported medication barrier was having too many medications to take (31%), in their study [20].

There are no reports in the literature regarding TTR values in Iranian patients and this is the first study that evaluates TTR in the Iranian population. We found a mean TTR of 54.9% in Iranian patients diagnosed with non-valvular atrial fibrillation who were receiving warfarin therapy. Only one factor was significantly related to poor control among patients: increase in the number of medications administered by the patients. In the future, we recommend evaluating factors that could possibly affect INR values and TTR rates, such as drug-warfarin interactions, food-warfarin interactions and patients’ treatment adherence.