Background
The prevalence of patients receiving anticoagulants in Germany is high, with about 930,000 people taking coumarins daily [
1]. Adequate oral anticoagulation management is therefore a key challenge in the medical care of elderly patients in general practices [
2]. In the general adult population in Germany, the 1-year prevalence of atrial fibrillation (AF), the most common indication for long-term oral anticoagulation [
3], has recently been estimated at 2.3%, and it rises considerably with age [
4]. Oral anticoagulation therapy (OAT) is also used for secondary prevention, e.g. in patients with recurrent thrombosis [
5].
Vitamin K antagonists (VKA), which are still the oral anticoagulants that are most often prescribed in Germany [
1], take fourth place in the List of High-Alert Medications consisting of drugs that are most likely to cause harm to patients [
6]. The management of coumarin therapy is often challenging because the narrow therapeutic index (for patients with AF a target Internationalised Normalised Ratio (INR) of 2.0 to 3.0 is recommended) makes it difficult to achieve optimal doses of the medication, and food-drug interactions are common [
7]. Improper handling can thus lead to dangerous bleeding complications or thromboembolic events. About 20% of hospitalisations in Europe as a result of acute stroke are related to atrial fibrillation [
8].
The introduction of direct oral anticoagulants (DOACs) has had a considerable impact on the world of oral anticoagulation: In Germany, dabigatran, a direct thrombin inhibitor, and rivaroxaban, a factor Xa inhibitor, were the first DOACs to be approved for the prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation and at risk of stroke [
9]. Since receiving approval in 2011, prescriptions of DOACs have risen strongly, with 38 million defined daily doses (DDDs) of DOACs prescribed in Germany in 2012 (vs. 389 mio. DDDs of VKAs), and 183 mio. DDDs prescribed in 2014 (vs. 364 mio. DDDs of VKAs) [
10]. According to the GARFIELD registry (Global Anticoagulant Registry in the FIELD), which includes patients with newly diagnosed atrial fibrillation and at least one additional risk factor, the proportion of patients on DOACs in Germany increased from 4.3% in 2011 to 24.8% in 2014 [
11]. Although DOACs are considered to be an effective treatment choice for long-term anticoagulation therapy, there have been a number of concerns [
9]. For instance, the lack of a readily available monitoring test and the absence of specific antidotes (apart from dabigatran, for which an antidote was approved in 2015 [
9]) are a problem in the event of DOAC-associated bleedings [
12]. DOACs are not suitable for all patients: For instance, patients with mechanical heart valves should not use them [
13]. Moreover, DOACs cost significantly more than VKA [
10]: In Germany, the mean net cost of vitamin K antagonists is €0.18 per DDD, compared to €3.75 for dabigatran, and €3.45 for factor Xa antagonists.
Oral anticoagulation treatment by caregivers in clinical practice is often unsatisfactory. A systematic review indicates under-treatment of high-risk atrial fibrillation patients in two thirds of analysed studies [
14]. Despite the need to maintain the INR within the guideline-recommended therapeutic target range in accordance with the indication [
5], INR monitoring and documentation quality in German general practices have often been reported to be inadequate [
15]. A cross-sectional survey by Chenot et al. indicates gaps in patient knowledge that reduce the safety and effectiveness of OAT. High age and low education levels are associated with lower overall safety-relevant knowledge about OAT [
16,
17]. Insufficient patient knowledge about OAT is related to poorer anticoagulation control [
18,
19]. An additional predictor of poor anticoagulation quality includes long intervals between measurements [
20].
Patient education has been proposed as a means of improving adherence to oral anticoagulation [
9,
21]. However, currently existing patient education strategies vary in terms of content, setting, duration, stated goal, structure and involved personnel [
22]. In Germany, where general practitioners (GPs) manage the majority of patients on OAT [
16], large-scale RCTs to test the effectiveness of clearly defined patient education interventions in primary care settings are needed [
23]. To the best of our knowledge, no study has yet analysed the effectiveness of a complex intervention that includes major elements of well-established tools such as practice-based case management, self-management of OAT, and additional education for patients and practice teams to improve their knowledge about OAT. Previous education strategies that have successfully reduced thromboembolic complications and mortality rates have mostly been based on self-management of OAT [
24‐
26]. Nonetheless, self-management may not be suitable for all patients [
27]. Apart from self-management, case management, which involves systematically monitoring patients, encouragement to continue the treatment and action in the case of non-adherence or no improvement [
28], has been successful in patients with other chronic conditions such as chronic heart failure [
29], osteoarthritis [
30], and depression [
31].
Between 2012 and 2015 the Institute of General Practice, Goethe-University Frankfurt, Germany, conducted the cluster-randomised PICANT trial (Primary Care Management for Optimised Antithrombotic Treatment) on patients with a long-term (i.e., lifelong) indication for oral anticoagulants in general practices in Germany, with the primary aim of improving antithrombotic management in primary healthcare by reducing major thromboembolic and bleeding events requiring hospitalisation [
32].
In this paper we investigate whether the complex PICANT intervention that includes practice-based case management, self-management of OAT and additional patient and practice team education leads to an increase in patient knowledge about anticoagulation therapy after 12 and 24 months in intervention recipients, compared to patients receiving usual care (as one of the PICANT trial’s secondary objectives).
Methods
Study design
The cluster-randomised controlled PICANT trial on anticoagulation management was conducted between June 2012 and March 2015 in 52 general practices in the federal states of Hesse and Rhineland-Palatinate. The study was approved on June 26th, 2012, by the ethical review committee of the University Hospital, Goethe-University Frankfurt, Germany. The study is registered at
www.controlled-trials.com (ISRCTN41847489). The study protocol, the presentation of the Coagulation-Monitoring-List (Co-MoL), the practice and patient recruitment process and data on screened patients have already been published elsewhere ([
32‐
34]). Publications are currently being prepared on the primary and secondary objectives.
Study population
Patients of ≥ 18 years of age, with a long-term (lifelong) indication for oral anticoagulation (atrial fibrillation/flutter, recurrent venous thromboembolism or pulmonary embolism, mechanical heart prosthesis, and other conditions, such as hereditary coagulopathy, intracardial thrombosis) and who were taking any kind of OAT (coumarins, antiplatelet therapies, DOACs such as dabigatran) were recruited. Moreover, they had to regularly attend the GP’s practice and sign an information consent form. Exclusion criteria included dementia, a life expectancy of less than 6 months, drug or alcohol abuse, residence in a nursing home, or insufficient German language skills.
Intervention
As described in the study protocol, the complex intervention included the provision of tools for study participants. These included information materials for patients, which depended on the type of oral anticoagulation they were taking, as well as information materials on oral anticoagulation and guidelines for GPs and healthcare assistants employed in their practices [
32].
In Germany, GP practices generally employ one or more health care assistants (HCAs, ‘Medizinische Fachangestellte’). Their role is comparable to healthcare assistants in the UK or to medical assistants in the United States [
35]. They perform basic clinical tasks such as intramuscular injections, ECGs, spirometry, and taking blood samples.
At a 1-day workshop, health care assistants were trained to perform case management and educate patients (including an information brochure and a video developed by Hua et al., both of which contained safety-relevant information relating to OAT) [
36].
HCAs were taught to tell patients about their disease and treatment conditions, to encourage them to perform self-management and to monitor them regularly using the Coagulation-Monitoring-List (Co-MoL) explained in detail elsewhere [
33]. The training also included DOAC-related topics (e.g. monitoring of renal function). HCAs conducted patient interviews following patients’ visits at the GP’s practice. In general, time intervals between monitoring visits depended on the stable adjustment of therapy. Most contacts took place in the practice, as the majority of anticoagulant patients visit the practice regularly anyway. If necessary, patients were contacted by telephone between the practice visits as well. The results of the interviews were reported to the GP who decided whether any further action was necessary. The aim was to assess symptoms and adherence to medication in patients, and to detect complications early and assess adverse effects.
GPs were provided with detailed explanations of what to expect from case management. Quality meetings were held three times during the course of the trial to discuss the newly approved DOACs, as well as practical problems involved in anticoagulation and individual case reports.
For patients, the complex intervention consisted of practice-based case management involving treatment monitoring, patient education, the provision of individual OAT-specific information, and encouragement to perform self-management where applicable. The control group received treatment as usual according to the current evidence-based guideline for oral anticoagulation therapy [
37,
38].
Data collection and patient knowledge questionnaire
The assessments occurred at baseline, after 12 and after 24 months. Patients completed a knowledge questionnaire developed by Hua et al. for a previous trial on education for patients receiving OAT [
36,
39]. The questionnaire contains 13 items covering safety-relevant knowledge about OAT. It aims to measure what patients learned from the intervention. Five questions relate to the patient’s individual OAT therapy, such as the indication for oral anticoagulation and the expected duration of treatment. Four questions relate to drug and other interactions. The remaining four questions are about safety precautions, such as the recognition of complications or emergencies. Open-ended questions allow free-text answers.
After extracting the main information from the free-text answers given by patients, categories were constructed manually, and answers with similar content summarised. Key facts emphasised in the intervention were used to build the categories. Following the evaluation scheme of Vormfelde at al., each of the thirteen questions included in the questionnaire was weighted with one point. Wrong answers, replies such as “I don’t know”, missing answers, or answers that were not evaluable within the context of the question, received 0 points. The sum-score therefore ranged from 0 to a maximum of 13 points, with higher scores indicating greater knowledge about OAT [
13]. The questionnaire was handed out to all patients (regardless of their OAT), because most questions were about general safety-relevant aspects of OAT. However, the evaluation scheme was modified to take into account particularities of DOACs, using current guidelines as a basis [
21]. For instance, when asked about their individual monitoring intervals, patients who had switched to DOACs and gave answers such as “I no longer need monthly monitoring of coagulation in the general practice” were given one point in the corresponding item, whereas patients taking vitamin K antagonists were expected to report their individual monitoring intervals.
Statistical analyses
Means and standard deviations of patients’ total sum-scores at baseline and at the two follow-ups were calculated for each group. Differences in the development of patient knowledge between the intervention and control groups compared to baseline were assessed separately for each follow-up assessment using linear mixed-effects models, and practices were included as random effects to account for the clustered structure of the data. A p-value of less than 5% was considered significant. SPSS version 19 and SAS version 9.4 were used for data analysis.
Conclusion
Our study showed that patient knowledge of INR target ranges, the proper use of medication to avoid drug interactions (e.g. with over-the-counter analgesics), and the recognition of emergencies such as strokes or bleeding complications, is often lacking. A complex intervention including general practice-based case management, self-management of OAT and additional education for patients and practice teams can lead to growth in patient knowledge. Since this intervention is only dependent upon the involvement of the entire general practice team and requires no additional interface, it would be possible to establish it as a measure to improve oral anticoagulation management in the German primary care sector. The need for education remains, even when vitamin K antagonists are replaced by direct oral anticoagulants. Since a readily available monitoring test does not exist for DOACs, and dosing depends on renal function, patients should be informed about safety-relevant aspects of their anticoagulant therapy. Educational programmes should contain standardised, clearly defined content focusing particularly on the safety-relevant gaps in knowledge that were identified in the baseline population. Patient education needs to be comprehensive and comprehensible, especially for older patients. Moreover, it can be presumed that patients should be educated more than once if a satisfactory level of patient knowledge is to be maintained.
Acknowledgements
The authors would like to thank Phillip Elliott for proof-reading the manuscript. We would like to express our gratitude to Prof. Dr. med. Chenot who provided us with the questionnaire developed by Dr. med. Duc Hua. We would also like to thank the practice teams and patients of the 52 general practices participating in the PICANT trial.