Journal of Thrombosis and Thrombolysis

, Volume 22, Issue 3, pp 185–190

A “two-step” educational approach for patients taking oral anticoagulants does not improve therapy control


  • Doris Barcellona
    • Dipartimento di Scienze Mediche InternisticheUniversity of Cagliari
  • Paolo Contu
    • Dipartimento di Igiene e Sanità PubblicaUniversity of Cagliari
    • Dipartimento di Scienze Mediche InternisticheUniversity of Cagliari

DOI: 10.1007/s11239-006-9027-2

Cite this article as:
Barcellona, D., Contu, P. & Marongiu, F. J Thromb Thrombolysis (2006) 22: 185. doi:10.1007/s11239-006-9027-2


Background and objective

The educational aspect of oral anticoagulant treatment is considered a possible cause of variability in anticoagulation levels. The aim of this prospective study was to investigate whether the time spent in the therapeutic range (TSTR) by patients taking oral anticoagulants could be improved by two different, consecutive educational approaches on the crucial aspects of oral anticoagulant therapy.

Design and methods

Between May and June 2004, validated interviews were conducted with 240 patients (128 male and 112 female, mean age 50 ± 12 years) enrolled in the study. Three months later, the patients were randomly allocated to three groups. A course that focused on the questions in the interview was followed by the first group (n = 80); a brochure containing the correct answers to questions was given to the second (n = 81); nothing was provided for the third (n = 79).


A significant difference was found in the TSTR between the quarters preceding and following the interview. Mean TSTR increase was 13%. Patients that were randomly selected to attend the educational course, read a brochure or do nothing showed similar TSTR percentages in the quarter following the interview. A good control of the anticoagulant therapy (TSTR >70%) was maintained, with no significant variation during the following three-quarters.

Interpretation and conclusions

A two-step educational approach for patients on long-term oral anticoagulation does not improve TSTR percentages in the short term.


Oral anticoagulantsPatient education


Oral anticoagulants are widely used in the prevention of thromboembolic events, and periodic international normalised ratio (INR) checks and contextual therapy adjustments are required to monitor oral anticoagulant therapy [1]. The risk of minor and major haemorrhage and thrombotic episodes increases in patients with poor anticoagulation control [2, 3]. The time spent by a single patient in the assigned therapeutic range [time spent in the therapeutic range (TSTR)] is an indirect measurement of the efficacy of the treatment and is considered good when it is at least 70% [4]. It is well known that intercurrent diseases, drugs [5], irregular vegetable intake [6, 7], seasonal variation [8], poor compliance [9, 10], doctors’ experience in handling oral anticoagulants [11] and genetic variants of cytochrome P450 CYP2C9 [12] may affect INR values and anticoagulation levels, thereby increasing the risk of haemorrhagic or thrombotic complications.

Patients are generally given a short educational course on how to manage oral anticoagulants correctly before starting treatment. This course, covering the indications, risks and benefits of oral anticoagulants, administration, dietary behaviour, influence of other drugs and intercurrent diseases, and general life-style, is intended to educate patients and increase their adherence to therapy.

In recent years, several studies have dealt with the educational aspects of oral anticoagulant treatment and have considered patients’ poor comprehension of doctors’ explanations to be a possible risk factor for the variability in oral anticoagulation levels.

Estrada et al. [13] demonstrated that brochures related to oral anticoagulants are incomprehensible for most elderly patients affected by atrial fibrillation and taking oral anticoagulants, suggesting the need for such material to be written at a clear and simple level. It has also been suggested that more importance be given to educational courses for patients on anticoagulation, as the TSTR was low in patients who forgot to take their daily tablets and in those who did not know why they were taking oral anticoagulants [14].

The aim of this prospective study was to investigate whether the TSTR could be improved by an interview followed by a further, different, randomly assigned educational approach on the crucial aspects of oral anticoagulant therapy.

Design of the study and patients

Design of the study

The aim of the study was to demonstrate whether a randomised allocation of patients to three different educational approaches, after an interview on the crucial aspects of oral anticoagulation, could lead to higher TSTR percentages. The number of patients included in the study was based on a previous estimation of the sample size needed to identify a significant difference in TSTR percentages before and after the second educational intervention. The minimum sample size required to detect a 10% mean difference in TSTR was 63 patients per group, considering an α level of 0.05 (two-tailed) and a power of 80%.

A pilot study was conducted on a group of 76 (34 male and 42 female, mean age 49 ± 5 years) patients on long-term anticoagulation therapy to verify the comprehension level of the interview. This was administered while they were awaiting INR results and therapy adjustments. Patients were then asked to explain the meaning of a few randomly selected questions. Ninety-five percent of them understood the questions and declared that the interview was clear. This allowed us to obtain a final version of the interview, which was then validated. All 76 patients were excluded from the study.

After the patients had provided informed consent, the validated interview, which included 20 multiple-choice questions on oral anticoagulant therapy, was conducted with 240 patients in May 2004. The interviewer (D.B.), collected the answers to the questions without influencing the patients. Data related to gender and schooling were also recorded, and each patient signed the interview. Using a computer-generated randomisation scheme, we randomly assigned patients to one of the three interventions: an educational course (n = 80); a brochure (n = 81); no intervention (n = 79).

Patients randomly chosen to attend the educational course were divided into three main subgroups of about 25 persons, based on their clinical indications for oral anticoagulation therapy, and were asked to participate in the course on three different days. The course focused on the questions contained in the interview. Patients were shown the right answer to each question contained in the interview, with the aid of slides, by the doctor who, in a simple manner, explained why it was the right answer. The brochures given to the second group of patients contained the text of the interview with the correct answers to each question. Patients were invited to read them carefully. Characteristics of the randomly chosen patients are shown in Table 1.
Table 1

Characteristics of the patients randomly allocated to the three groups (educational course, brochure or nothing) (OA oral anticoagulant)


Educational course
















65 (55–74)

65 (54–70)

69 (61–74)


Schooling level

    Illiterate–primary school





    Middle-upper school










Duration of ` therapy

    <5 Years





    5–15 Years





    >15 years





Indication for OA therapy

    Atrial fibrillation










    Arterial thrombosis





    Heart valve prosthesis





    Myocardial infarction





    Dilated cardiomyopathy




    Venous thrombosis





aAge is expressed (in years) as median and interquartile range. P value was obtained by the Kruskall–Wallis test. χ2 test was used in the other cases


Patients who had started oral anticoagulant treatment less than 1 year prior to the study, who received their adjusted therapy at home via fax, whose INR values had not always been available in the previous year due to a transitional interruption of anticoagulants or hospitalisation and those whose oral anticoagulant treatment was expected to end within the following 6 months, were excluded from the study.

A total of 240 consecutive eligible patients (128 male and 112 female, mean age 50 ± 12 years) agreed to participate and were enrolled in the study. The clinical indications for oral anticoagulants were: mechanical heart valve prosthesis (n = 27), atrial fibrillation (n = 109), myocardial infarction (n = 16), dilated cardiomyopathy (n = 2), cardio-embolic stroke (n = 12), peripheral arterial thrombosis (n = 6), and venous thrombosis with or without pulmonary embolism (n = 68). A computer-based decision support system was used to manage oral anticoagulant therapy [Program for archive, refertation, and monitoring of anticoagulated patients (PARMA; release 5.0)] [15]. One-hundred and eighty-six patients were treated with acenocoumarol (Sinthrome), and 54 with warfarin (Coumadin).

All prothrombin times were measured in fresh plasma within 1 h of blood collection employing an automatic coagulometer (ACL Futura, IL Milan, Italy) with a commercial thromboplastin (Thromboplastin, IL Milan, Italy) and with an International Sensitivity Index of 1.08.

A total of 15 patients was lost during the follow-up period: seven from the teaching-course group (two died and five transferred to other anticoagulation clinics); six from the brochure group (one died, three were lost due to admission to other hospitals and two were followed by their general practitioners); two from the no-intervention group (transferred to other anticoagulation clinics). Seven patients in the teaching course group did not take part in the course and were not considered in the statistical analysis of the data.

Statistical analysis

TSTR percentage was calculated for each patient on a quarterly basis using the INR Day Program by Rosendaal et al. [16]. TSTR percentage differences in the quarters before and after the educational interventions were also calculated. This variable showed a normal distribution (Kolmogorov–Smirnov test), and, thus, the t-test for paired data was employed. The 95% confidence interval (CI) of the mean of the differences was also computed. The McNemar test was used for a qualitative comparison of the data, considering a 70% TSTR as cut off. The Fisher exact test and the χ2 test were used to calculate the frequency of patients whose time spent in range improved after the interview. Patients were divided into poor (<70%) and good (≥70%) therapeutic control.

The analysis of variance for repeated measurements was used to evaluate possible influence of gender, schooling level and duration of oral anticoagulation on the variation in TSTR before and after the interview. Finally, the Spearman correlation coefficient was computed for the evaluation of a potential association between variables.


Results showed no significant association between TSTR and the percentage of right answers to questions in the interview. The percentage of right answers is reported in Table 2.

A significant difference was found in the TSTR between the quarters preceding and following the interview (Table 3). Mean TSTR increase was 13% (CI 9–17%) (Table 4).
Table 2

Oral anticoagulant therapy interview: questions and percentages of correct answers (OA oral anticoagulant)


Percentage of correct answers

(1) Why am I taking OAs?

206/240 (86%)

Free answer

(2) What do OAs protect me from?

195/240 (81%)

(a) Haemorrhage (b) thrombosis (c) illnesses in general

(3) Why are PT INR tests necessary?

212/240 (88%)

(a) To identify blood disorders (b) to identify heart disorders (c) to modify/confirm OA dosage

(4) How many times a day must OAs be taken?

237/240 (99%)

(a) Once (b) twice (c) every 2 days

(5) When must OAs be taken?

77/240 (32%)

(a) At lunch (b) between meals (c) at the most convenient time so do not forget to take it

(6) What happens if OAs are omitted?

142/240 (59%)

(a) The PT increases (b) the PT drops (c) the PT does not change

(7) Can OAs be taken a few hours late?

190/240 (79%)

(a) Yes, always (b) never (c) only after informing the Centre’s doctors

(8) What is vitamin K?

111/240 (46%)

(a) A vitamin for osteoporosis (b) a vitamin that reduces PT (c) a vitamin that increases PT

(9) Which foods contain most vitamin K?

155/240 (64%)

(a) Fish (b) meat (c) vegetables

(10) Can different daily quantities of vegetables be eaten?

157/240 (65%)

(a) Yes (b) no (c) do not know

(11) What should vegetable intake be?

149/240 (62%)

(a) The same in quantity (b) the same in quality (c) as required on a daily basis

(12) What can happen if other drugs are taken or suspended?

166/240 (69%)

(a) Nothing, PT remains the same (b) can be intoxicated (c) PT can increase or decrease

(13) Can other drugs be taken with OAs on top of those recommended?

222/240 (92%)

(a) Yes, after informing the doctors at the Centre (b) never (c) yes

(14) Are intra-muscular injections allowed while taking OAs?

136/240 (56%)

(a) Only occasionally using an ice bag (b) yes (c) never

(15) What can happen after an intra-muscular injection?

129/240 (54%)

(a) Haemorrhage (b) thrombosis (c) infection

(16) Are teeth extractions allowed while taking OAs?

207/240 (86%)

(a) Never (b) yes, without controlling PT (c) yes, after informing doctors at the Centre

(17) What are the risks when PT INR is 1.60?

124/240 (52%)

(a) None (b) thrombosis (c) haemorrhage

(18) What are the risks when PT INR is 6.30?

141/240 (59%)

(a) None (b) thrombosis (c) haemorrhage

(19) What are the side effects of OAs?

134/240 (56%)

(a) Stomach ache (b) headache (c) serious haemorrhage

(20) Are OAs permitted during pregnancy?

12/18 (66%)

(a) Yes, after informing doctors at the Centre (b) no, they can cause foetus malformation

Table 3

Percentage of TSTR (mean ± SD) in the quarters preceding and following the interview

Quarter preceding and following interview

TSTR (%)

February–April 2004

61 ± 25

Interview (May)

June–August 2004

75 ± 19

Frequency of patients (91 out of 225) with good anticoagulation levels (≥70%) was significantly higher in the quarter following the interview than in the preceding one (χ2 McNemar = 15.24, P = 0.00009). No influence of gender (P = 0.97), schooling level (P = 0.31) or duration of anticoagulation therapy (P = 0.11) on TSTR variations was observed. The TSTR percentages of the patients randomly allocated to an educational course, receive a brochure or to act as a control group were similar (Table 5).
Table 4

Mean difference and 95% CI of the percentage of TSTR between the different quarters of oral anticoagulant treatment before and after the educational programmes


Mean (%)

CI 95%


Interview (May)

    June–August 2004 vs February–April 2004




Course, brochure, nothing (September)

    October–December 2004 vs June–August 2004




    January–March 2005 vs October–December 2004




    April–June 2005 vs January–March 2005




The three groups of patients randomly chosen to attend the educational course, read a brochure or do nothing showed similar TSTR percentages in the following quarter (Table 5). A good control of the anticoagulant therapy (TSTR >70%) was maintained, with no significant variation during the following three quarters.
Table 5

TSTR percentages in the three groups of patients after attending an educational course, receiving a brochure or doing nothing


Course (n = 66) (%)

Brochure (n = 75) (%)

Nothing (n = 77) (%)

All patients (n = 218) (%)

June–August 2004 vs October–December 2004

75 ± 20

73 ± 23

77 ± 18

75 ± 20

January–March 2005 vs October–December 2004

71 ± 19

76 ± 19

78 ± 20

75 ± 20

April–June 2005 vs January–March 2005

74 ± 22

73 ± 20

78 ± 20

75 ± 21


The results show that the use of an interview does improve TSTR, irrespective of schooling level, gender or therapy duration. This finding confirms our previous report [14], even though the questions in this study dealt with general topics concerning oral anticoagulants and did not cover personal behaviour related to the anticoagulant therapy. Recently, Palareti et al. [17] also found that carrying out a questionnaire on oral anticoagulant therapy had a beneficial effect on patients with poor anticoagulation.

On the basis of an interesting study by Laporte et al. [18], showing that there were no differences in therapy control between an intensive educational course and the standard one normally offered, we have suggested that doctors at anticoagulation clinics should give explanations and advice on oral anticoagulants gradually, focusing the attention of patients with poor anticoagulation on the crucial concepts of oral anticoagulants during the entire period of therapy [19].

Surprisingly, our study demonstrates that a two-step educational approach is useless over a short period of time. In fact, the main result of this study is that a 13% improvement in TSTR is maintained for three quarters in patients who had neither attended a course nor read an illustrated brochure after the interview.

All the patients submitted for interview received general information covering the indications, risks and benefits of oral anticoagulants, administration, dietary behaviour, influence of other drugs and intercurrent diseases, and general lifestyle at the start of anticoagulant therapy. Long-term treatment with oral anticoagulants could lead patients to forget or to underestimate some of the information received from the anticoagulation clinic doctors, especially if haemorrhagic or thrombotic events did not occur. Moreover, the computerised programs used in the management of oral anticoagulant therapy and the over-crowded anticoagulation clinics today have probably determined less contact between doctor and patient, and, therefore, the information given at the start of the therapy is not reinforced.

A study conducted in two different Italian anticoagulation clinics (Cagliari and Padua) showed that 83% and 39%, respectively, of the patients taking oral anticoagulants preferred to speak to a doctor, suggesting that patients in some way could resent this new method of monitoring oral anticoagulants [20].

One can therefore presume that the use of the interview alone could have reminded patients of the crucial aspects of anticoagulant treatment, and that the second educational intervention does not lead to a further increase in TSTR, as no new information capable of modifying patient behaviour is added. The interview probably had a role in increasing patients’ motivation rather than improving their understanding of oral anticoagulation. Both second educational steps, the teaching course and the brochure, probably failed to increase patient motivation further. On the other hand, the lack of correlation between the percentage of right answers to the interview and TSTR in the quarter preceding the interview indicates that poor adherence to the therapy is not dependent on patient comprehension of the oral anticoagulant mechanisms or related clinical problems. With regard to some questions, about 50% of patients ignored some fundamental aspects of oral anticoagulation, which could give rise to legal implications. In fact, in the general management of oral anticoagulant therapy, not knowing what vitamin K is and in which foods it is primarily present is not so dangerous, but not knowing the main side effects of coumarins, i.e. haemorrhages, their teratogenic effects and the correlation between PT values and possible complications, is.

TSTR improvement and therapy adherence seem to be linked to an active patient role in their relationship with anticoagulation clinic doctors. These findings question whether patients on anticoagulant therapy should receive a traditional teaching approach alone or should be periodically stimulated to be directly involved by means of a questionnaire or an interview after an individual educational course at the start of the oral anticoagulant therapy. The use of a questionnaire is so easy and cheap that it is suitable even in crowded anticoagulation clinics and for patients who do self-testing. Contact between patients and the anticoagulation clinics can be created via the Internet by sending the questionnaire to be filled in at home. In the future, this clinical practice may also be suitable for patients treated with new anti-thrombotic drugs who do not require laboratory monitoring, as recently suggested [21].

This study does not allow definitive conclusions on the educational approach to be drawn, but, for the first time, at least to our knowledge, it challenges the traditional educational approach to patients on anticoagulant therapy. A much larger and prolonged randomised study could indicate the best way to arrive at a standardised educational programme. Finally, other limitations of this study could be: (a) therapy duration could have affected the results, since we only considered patients on long-term oral anticoagulation. Results might be different if a new study were performed on patients for whom a shorter oral anticoagulation therapy was planned; (b) the results could not be applied to centres with small numbers of patients or to patients followed by general practitioners; (c) the variability in preparing interviews, teaching courses and brochures.

Copyright information

© Springer Science+Business Media, LLC 2006