Warfarin is a well-established oral anticoagulant used for the prevention and treatment of thromboembolism and thromboembolic complications in patients with atrial fibrillation, heart valve replacement, or myocardial infraction [
Several studies have demonstrated that pharmacist-managed anticoagulation clinics improve time in therapeutic range, lower the incidence of adverse events, and reduce the need for frequent office/anticoagulant clinic visits [
Hospitalization is one of the potential causes of INR fluctuations in the postdischarge period, particularly among patients who newly initiated therapy [
This study is a randomized controlled trial with parallel group design and a 1:1 allocation ratio. It was conducted between May and September 2016 in Kamphaeng Phet Hospital, a 410 bed general hospital in the Lower Northern Region of Thailand. The study protocol was approved by Naresuan University Institutional Review Board (NU-IRB no. 119/59). Informed consent was obtained from all patients before their participation in this study.
Patients aged 20 years or older were eligible for inclusion if they (i) were prescribed warfarin upon discharge, (ii) had no scheduled surgery within three months after discharge, (iii) were able to manage their own medication, (iv) could communicate in Thai, and (v) possessed a telephone. Those who were prescribed warfarin prior to admission and had INR values in target ranges throughout the preceding three months were excluded.
As no similar study was found that could be used for sample size calculations, the number of participants in this study was based on a recommendation by Collins et al. for experimental research design [
Participants in the telephone follow-up group and the usual care group received standard pharmacy services for patients discharged on warfarin, which included (i) assessing appropriateness of the warfarin dose, (ii) checking for warfarin drug interaction with other concomitant medicines, (iii) educating patients about the indication and the use of warfarin, (iv) assessing and identifying factors associated with warfarin adherence for those who had been using warfarin prior to admission, and (v) checking for patients’ understanding of the given information.
Three clinical pharmacists who worked in medical wards were informed about the telephone follow-up procedures and were checked for their fidelity to the study protocol on a monthly basis. After discharge of the participants in the telephone group, pharmacists made one phone call before the next in-person follow-up based on the following schedules: day three for a one-week scheduled visit, day seven for a two-week scheduled visit, and day fourteen for a three- or four-week scheduled visit. For the second and subsequent visits within the three-month study period, pharmacists continued to make telephone calls on a regular interval as follows: day three for a 1-week scheduled visit; days three and seven for a two-week scheduled visit; days three, seven, and fourteen for a three-week scheduled visit; and days seven, fourteen, and twenty-one for a four-week scheduled visit.
During each 10–25-minute telephone call, pharmacists performed medicine use review by asking patients about problems/obstacles with managing warfarin including adverse events and complications, assessing medication adherence, and giving reminders for the next scheduled visits. Pharmacists promptly addressed any problems identified during the telephone call with patients.
The primary outcomes were the proportion of INR values in range, time in therapeutic range (TTR), proportion of INR out of range, and the number of patients with one or more out-of-range INR values. Safety outcomes were the frequency of complications or adverse events associated with warfarin, i.e., bleeding events, thromboembolic events, emergency room visits, and hospitalization. Bleeding in the following locations was classified as major bleeding: intracranial, intraspinal, intraocular, retroperitoneal, intra-articular, pericardial, and intramuscular [
All INR testing was performed with ACL TOP 300 CTS machine, which was calibrated by technicians from the instrument’s company every month. The TTR was calculated with the Rosendaal method using INR measurements from all warfarin clinic visits for each patient. The medical technologists were not aware of the study groups when they performed and reported the participants’ INRs.
Demographic characteristics, indication for warfarin therapy, and goal INR were extracted from medical records for each patient. Warfarin clinic records were the source of data on outcomes of interest. At each clinic visit, patients were asked to report complications/adverse events such as bleeding and thrombosis. Pharmacists at the warfarin clinic then checked for signs and symptoms of reported safety outcomes. Participants were also asked to report any emergency room visits or hospitalization, which was subsequently confirmed with hospital records.
The study population was stratified by experience of using warfarin, i.e., individuals who had used warfarin before admission and naïve users who were first prescribed warfarin during the course of the current admission. A random allocation sequence was created using a computer generated random number table. Random numbers were subsequently put into opaque envelope and sealed. Clinical pharmacists who worked on medical wards identified eligible patients and obtained signed informed consent before they were discharged. During the recruitment phase, information about newly enrolled participants was sent to the principal investigator to assign participants to an intervention.
Participant demographic characteristics were compared using either the chi-squared test for nominal data or the
The recruitment period took place between May and June 2016. Once enrolled, each participant was followed for three months with the last follow-up ending in September 2016. A total of two, out of fifty, participants did not complete the study. One participant from the telephone follow-up group no longer required warfarin due to an improvement in indicated condition (deep vein thrombosis), while one individual from usual care group died (Figure
CONSORT flow diagram.
The majority of participants in the telephone follow-up group were female (18/25, 72%), which was different in the usual care group (12/25, 48%). Mean age of participants in both groups was not different, i.e., 56.6±11.9 years and 58.7±10.0 years in telephone follow-up and usual care group, respectively. Participants who were prescribed warfarin for the first time accounted for half of the participants in each group (13/25, 52%). The main indication for warfarin in both groups was atrial fibrillation, i.e., 56% in the telephone follow-up group and 68% in the usual care group. Other prescribed indications included valvular heart disease, mechanical prosthetic valves, deep vein thrombosis, and pulmonary embolism. Participants in both groups shared common comorbidities such as diabetes, hypertension, and dyslipidemia. Almost all participants had the same target INR of 2.0-3.0. Only two (8%) participants in the telephone follow-up group and one (4%) in the usual care group had their INR goal at 2.5-3.5 (Table
Demographic characteristics.
|
|
|
|
---|---|---|---|
Gender, female | 18 (72%) | 12 (48%) | 0.083 |
Age, years | 56.6 |
58.7 |
0.506 |
Experience of using warfarin | |||
Naïve | 13 (52%) | 13 (52%) | 1.00 |
|
5 (20%) | 3 (12%) | 0.440 |
>3 month experienced | 7 (28%) | 9 (36%) | 0.544 |
Indication | |||
Valvular heart disease | 6 (24%) | 8 (32%) | 0.529 |
Mechanical prosthetic valves | 2 (8%) | 1 (4%) | 0.552 |
Atrial fibrillation | 14 (56%) | 17 (68%) | 0.382 |
Deep vein thrombosis | 8 (32%) | 6 (24%) | 0.529 |
Pulmonary embolism | 0 (0%) | 1 (4%) | 0.312 |
Comorbidities | |||
Diabetes | 4 (16%) | 5 (20%) | 0.713 |
Hypertension | 9 (36%) | 15 (60%) | 0.089 |
Dyslipidemia | 7 (28%) | 8 (32%) | 0.758 |
Ischemic heart disease | 3 (12%) | 1 (4%) | 0.297 |
Heart failure | 5 (20%) | 6 (24%) | 0.123 |
Chronic kidney disease | 5 (20%) | 6 (24%) | 0.733 |
Others | 3 (12%) | 2 (8%) | 0.637 |
Warfarin dosing on discharge |
18.2±6.7 | 18.1±4.6 | 0.922 |
Target INR | |||
2.0-3.0 | 23 (92%) | 24 (96%) | 0.552 |
2.5-3.5 | 2 (8%) | 1 (4%) | 0.552 |
INR: international normalized ratio.
Pharmacists made a total of 179 successful telephone calls to participants in the telephone follow-up group with an average of 7.16 calls for each participant (range 7–10 calls). During the three month study period, participants in both groups had their INR measured on average 3.4 times. Specifically, a total of 79 INRs were measured in each group. Thirty-six INRs (45.6%) were in range for the telephone follow-up group, whereas only 19 INRs (24.1%) in the usual care group had INR values in the target range, p=0.005 (Table
Therapeutic outcome measures.
|
|
|
|
---|---|---|---|
INR values in range† | 45.6% (36/79) | 24.1% (19/79) | 0.005 |
TTR‡ | 49.8 |
28.0 |
0.017 |
Out of range INR values | |||
Subtherapeutic† | 43.0% (34/79) | 51.8% (41/79) | 0.265 |
Supratherapeutic† | 11.4% (9/79) | 24.1 % (19/79) | 0.037 |
Patients with |
84.0% (21/25) | 100.0% (25/25) | 0.037 |
Complications reported in both groups were not different. A total of three major bleeding events occurred during the study period; two were from the telephone follow-up group, and the other from the usual care group. Although not statistically significant, thromboembolic events and warfarin-related hospital admissions in the telephone follow-up group seemed to be relatively lower than those in the usual care group, 3/25 (12%) versus 6/25 (24%), p=0.259, and 4/25 (16%) versus 7/25 (28%), p=0.306, respectively (Table
Complications.
|
|
|
|
---|---|---|---|
Bleeding events | |||
Major bleeding | 2 (8%) | 1 (4%) | 0.552 |
Minor bleeding | 3 (12%) | 4 (16%) | 0.684 |
Thromboembolic events | 3 (12%) | 6 (24%) | 0.259 |
Warfarin-related hospital admission | 4 (16%) | 7 (28%) | 0.306 |
To the best of our knowledge, this is the first study in the field using a RCT design to evaluate the impact of a telephone follow-up service in patients on warfarin. Hospitalized patients who were discharged on warfarin and received the telephone follow-up service had a higher rate of INR in target range and a higher TTR percentage than those who received usual care. The telephone follow-up service also decreased the number of patients experiencing out-of-range INRs after hospital discharge. This study also proved that the service is safe for patients as both study groups experienced rare and similar rates of thromboembolic and bleeding complications.
Most previous studies [
The beneficial effects of a telephone follow-up service found in this study are well aligned with findings from Ibrahim et al. which showed that follow-up counseling through phone calls every week and home visit biweekly for three months improved INR control [
Despite patients who received telephone follow-up service having a higher percentage of TTR compared with those in the usual care group, they were only at goal 44% of the time, which does not indicate a good response to warfarin management, i.e., 60% or more [
There are some limitations in this study. First, the duration of the telephone follow-up period was only 3 months, which may not be sufficient to observe effects that require long-term follow-up such as death or rehospitalization related to warfarin therapy. The sample size in this study also appeared relatively low compared with other trials. Nonetheless, it was adequate to statistically identify the difference in INR values in range and TTR between the telephone follow-up and usual care groups. The limitation of this small sample size was that it restricted the ability to undertake further subgroup analysis. Another limitation is that patients were recruited from one hospital, which may not represent the general population. It should be noted, however, that several characteristics of these patients including age, comorbidities, warfarin indication, and target INR suggest that they are comparable with general patients who receive warfarin after discharge.
This randomized controlled trial demonstrated that a telephone follow-up service when combined with usual care in recently discharged patients on warfarin helps patients achieve and maintain their INR target. Further studies with more participants and longer study periods may be needed to explore the impact of this service on complications of warfarin therapy. This anticoagulant supportive service could be a feasible approach to improve clinical outcomes and should be promoted to patients receiving warfarin therapy after discharge.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors declare that they have no conflicts of interest.
This research work received financial support from Naresuan University Postgraduate Student Fund. The authors acknowledge all patients participating in the study for their time. The authors would also like to thank clinical pharmacists who helped perform the interventions and all related healthcare personnel for their support during the course of this study. Special thanks are due to Dr. Alicia Ritscher, Pharm.D., Clement J Zablocki Veterans Affairs Medical Center, for improving the English language of the manuscript.