This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
WEB-Based Distress Management Program for Implantable CARdioverter defibrillator Patients (WEBCARE) is a Web-based randomized controlled trial, designed to improve psychological well-being in patients with an implantable cardioverter defibrillator (ICD). As in other Web-based trials, we encountered problems with attrition and adherence.
In the current study, we focus on the patient characteristics, reasons, and motivation of (1) completers, (2) those who quit the intervention, and (3) those who quit the intervention and the study in the treatment arm of WEBCARE.
Consecutive first-time ICD patients from six Dutch referral hospitals were approached for participation. After signing consent and filling in baseline measures, patients were randomized to either the WEBCARE group or the Usual Care group.
The treatment arm of WEBCARE contained 146 patients. Of these 146, 34 (23.3%) completed the treatment, 88 (60.3%) dropped out of treatment but completed follow-up, and 24 (16.4%) dropped out of treatment and study. Results show no systematic differences in baseline demographic, clinical, or psychological characteristics between groups. A gradual increase in dropout was observed with 83.5% (122/146) completing the first lesson, while only 23.3% (34/146) eventually completed the whole treatment. Reasons most often given by patients for dropout were technical problems with the computer, time constraints, feeling fine, and not needing additional support.
Current findings underline the importance of focusing on adherence and dropout, as this remains a significant problem in behavioral Web-based trials. Examining possibilities to address barriers indicated by patients might enhance treatment engagement and improve patient outcomes.
Clinicaltrials.gov: NCT00895700; http://www.clinicaltrials.gov/ct2/show/NCT00895700 (Archived by WebCite at http://www.webcitation.org/6NCop6Htz).
The implantable cardioverter defibrillator (ICD) is a cardiac device that is implanted with leads in and on the heart in patients for the primary and secondary prevention of sudden cardiac death due to life-threatening ventricular tachyarrhythmias [
A subgroup of ICD patients experiences psychological distress after ICD implantation, such as anxiety, depression, post-traumatic stress, and impaired quality of life [
In order to make psychological treatment for ICD patients more patient-tailored, which may reduce dropout, the use of online Web-based interventions has been advocated [
In the current study, we will describe the attrition and adherence issues that we encountered during the “WEB-based distress management program for implantable CARdioverter dEfibrillator” patients (WEBCARE) trial (NCT00895700). The trial design paper was published previously [
Consecutively implanted ICD patients from six hospitals in the Netherlands (ie, Erasmus Medical Centre, Rotterdam; Amphia Hospital, Breda; Catharina Hospital, Eindhoven; Onze Lieve Vrouwe Gasthuis, Amsterdam; Canisius Wilhelmina Hospital, Nijmegen; Vlietland Hospital, Schiedam) were approached for study participation between April 2010 and February 2013. Patients were eligible for participation if they fulfilled the following inclusion criteria: first-time ICD implant, age 18-75 years, proficient in the Dutch language, and with Internet access and a sufficient level of Internet skills. Exclusion criteria were the following: life expectancy less than 1 year, history of psychiatric illness other than affective/anxiety disorders, or on the waiting list for heart transplantation.
Patients were approached by the ICD nurse or ICD technician prior to or briefly after ICD implantation. They were informed both verbally and in writing about the study. If the patient met the inclusion criteria and was willing to participate, informed consent was signed. Patients who could not decide at that time were approached again after ICD implant while still hospitalized. Prior to discharge from the hospital, consented patients were provided with the first set of questionnaires (baseline) and their medical records were accessed for information on their demographic and clinical variables. After completing the questionnaires, patients returned them in a self-addressed and pre-stamped envelope to Tilburg University, Netherlands, which served as the core lab for WEBCARE. If the questionnaires were not returned within two weeks, patients received up to 3 reminder phone calls. Patients who did not want to participate but who were willing to give access to information from their medical records also signed an informed consent form. The study was approved by the Medical Ethics Committee of all participating centers and was conducted in accordance with the Declaration of Helsinki. All patients provided written informed consent.
After receiving the baseline questionnaires and signed informed consent, participants were randomly assigned to either of two conditions: (1) the WEBCARE (WC) group, receiving questionnaires at baseline, 3 months, 6 months, and 12 months by mail, and getting access to the Web-based intervention for a time period of 12 weeks to complete 6 modules online [
Patients were randomized using block randomization by computer, randomizing 20 patients per hospital, at each time point. Randomization lists were generated by an independent, blinded statistician and sealed by a research assistant. For the current analyses, we will only focus on patients who were randomized to the WC group.
Patients who signed the informed consent form but who decided to quit the intervention and/or the study prematurely were contacted by telephone 12 weeks after randomization and asked why they had decided to quit. This time interval was chosen in order to not interfere with possible intervention effects (patients were allowed to work at their own pace, some chose to finish the intervention within the first two weeks, while others decided to do the 6 lessons within the last two weeks. For that reason, it was clear at 12 weeks who had quit or finished the intervention). Hence, patients were contacted at the time that they should have received their 3-month follow-up and finished the 6-module online course.
The intervention was based on the previously developed Web-based treatment “Alles Onder Controle”
Patients were allowed to work at their own time and pace; however, if a lesson was not finished within two weeks, a reminder email was sent, with up to 3 reminders per lesson. Patients could proceed to the next lesson only when the previous one was finished and the homework assignment was sent to the therapist. If patients did not log in within the first two weeks, a reminder email was sent. Twelve weeks after receiving the log-in information, patients’ accounts were automatically closed.
Information on demographic variables (ie, age, gender, working status, marital status, education level) was collected through purpose-designed questions in the questionnaires, while information on clinical variables (ie, left ventricular ejection fraction [LVEF], QRS-width [electrocardiogram reading], New York Heart Association functional class [NYHA-class], presence of heart failure, use of cardiac and psychotropic medication) were extracted from patients’ medical records at the time of implantation by the implanting electrophysiologist or research nurses at the participating centers. The Charlson Comorbidity Index [
The Generalized Anxiety Disorder scale (GAD-7) was used to assess anxiety [
The Patient Health Questionnaire (PHQ-9) is a 9-item self-report measure of depression (eg, “Having little interest or pleasure in doing things”) that taps into the 9 diagnostic criteria for DSM-IV depressive disorder [
Type D personality was assessed with the 14-item Type D scale (DS14) [
The Life Orientation Test (LOT) was used to assess optimism and pessimism [
Continuous variables were compared using the Student’s
A detailed description of the patient recruitment for WEBCARE is displayed in
Demographic, clinical, and psychological characteristics of patients who were randomized to the WC group are shown in
Flowchart of patient recruitment.
Baseline demographic, clinical, and psychological characteristics of patients randomized to the WEBCARE treatment condition (WC; n=146).
Characteristic | WC group, mean (SD) or n (%) | |
|
||
|
Age, mean (SD) | 58.23 (9.87) |
|
Gender (male) | 120 (82.2%) |
|
Employed | 68 (46.6%) |
|
Has partner | 124 (84.9%) |
|
High education | 106 (72.6%) |
|
||
|
LVEF≤35a | 87 (59.6%) |
|
QRS>120msb, n=144 | 59 (41%) |
|
NYHA class III/IVc, n=122 | 20 (16.4%) |
|
Heart failure | 78 (53.4%) |
|
CCId, mean (SD) | 1.60 (1.06) |
|
Beta-blockers | 117 (80.1%) |
|
ACE-inhibitors | 82 (56.2%) |
|
Diuretics | 72 (49.3%) |
|
Psychotropic medication | 13 (8.9%) |
|
||
|
Anxiety, mean (SD) | 4.57 (5.02) |
|
Depression, mean (SD) | 5.93 (5.11) |
|
Psychological treatment | 8 (5.5%) |
|
Cardiac rehabilitation, n=145 | 20 (13.8%) |
|
Type D personality | 24 (16.4%) |
|
Optimism, mean (SD) | 11.23 (2.68) |
|
Pessimism, mean (SD) | 5.73 (3.52) |
aLVEF: left ventricular ejection fraction
bQRS-width: electrocardiogram Q,R, and S waves
cNYHA: New York Heart Association functional class
dCCI: Charlson Comorbidity Index
Patients who did not return the baseline questionnaires (n=51) and were excluded (not randomized) from current analyses did not differ systematically on demographic variables. However, significant differences on clinical variables were observed with patients who were not randomized, more often having a NYHA Class III/IV (
Of the 146 randomized patients to the WC group, 34 (23.3%) completed the treatment and filled in the follow-up assessment (completers), 88 (60.0%) patients dropped out of the treatment but remained in the study and filled in the follow-up assessments (treatment dropouts), and 24 (16.4%) patients dropped out of the treatment and the study (dropouts). Focusing on the treatment,
Adherence and attrition during the intervention (n=146).
Completers and (treatment) dropouts did not systematically differ on any baseline demographic, clinical, or psychological measures (
Baseline demographic, clinical, and psychological characteristics stratified by group.
Characteristic | Completers, n=34 | Treatment dropout, n=88 | Dropout, n=24 |
|
|
mean (SD) or n (%) | |||||
|
|||||
|
Age, mean (SD) | 57.91 (9.82) | 58.84 (9.84) | 56.42 (10.20) | .56 |
|
Gender (male) | 28 (82.4%) | 74 (84.1%) | 18 (75.0%) | .59 |
|
Employed | 15 (44.1%) | 38 (43.2%) | 15 (62.5%) | .23 |
|
Has partner | 30 (88.2%) | 73 (83.0%) | 21 (87.5%) | .71 |
|
High education | 26 (76.5%) | 65 (74.7%) | 15 (62.5%) | .43 |
|
|||||
|
LVEF≤35a | 21 (61.8%) | 52 (59.1%) | 14 (58.3%) | .96 |
|
QRS≥120b | 12 (36.4%) | 37 (42.5%) | 10 (41.7%) | .83 |
|
NYHA III/IVc | 4 (13.3%) | 13 (20.0%) | 3 (15.0%) | .69 |
|
Heart failure | 18 (52.9%) | 46 (52.3%) | 14 (58.3%) | .87 |
|
CCId, mean (SD) | 1.88 (1.15) | 1.56 (1.10) | 1.38 (0.65) | .16 |
|
Beta-blockers | 28 (82.4%) | 69 (78.4%) | 20 (83.3%) | .81 |
|
ACE-inhibitors | 18 (52.9%) | 52 (59.1%) | 12 (50.0%) | .66 |
|
Diuretics | 16 (47.1%) | 43 (48.9%) | 13 (54.2%) | .86 |
|
Psychotropics | 1 (2.9%) | 9 (10.2%) | 3 (12.5%) | .36 |
|
|||||
|
Anxiety, mean (SD) | 5.46 (5.18) | 4.21 (5.08) | 4.63 (4.62) | .47 |
|
Depression, mean (SD) | 6.53 (4.40) | 5.79 (5.38) | 5.58 (5.17) | .73 |
|
Type D | 7 (20.6%) | 14 (15.9%) | 3 (12.5%) | .70 |
|
Optimism, mean (SD) | 11.12 (2.71) | 11.41 (2.74) | 10.75 (2.44) | .55 |
|
Pessimism, mean (SD) | 6.15 (4.16) | 5.56 (3.40) | 5.79 (2.99) | .71 |
aLVEF: left ventricular ejection fraction
bQRS-width: electrocardiogram Q,R, and S waves
cNYHA: New York Heart Association functional class
dCCI: Charlson Comorbidity Index
The reasons given by patients for not completing the treatment are displayed in
Time constraints were an issue in 15.2% (17/112) of patients (
Other reasons for dropout included treatment as being too confronting (4.5%, 5/112): “It was too personal, too confronting. I realized that I had more problems than I thought”, and feeling too sick (4.5%, 5/112): “It was too much. I had two surgeries in the past half year and I’m now on the waiting list for heart transplantation. I’m feeling sick all the time”. There were also patients who experienced the treatment as too negative: “I was feeling fine about my ICD, but when I started reading the content of the online course, I started feeling unhappy and therefore I decided to quit”, and “The homework assignments and questionnaires are too negatively worded, while you expect us to start thinking positive. I didn’t want to proceed as I didn’t wanted to start thinking negative about the ICD and how I’m feeling”.
Reasons for dropout (n=112).
This is the first behavioral randomized controlled trial to address the adherence and dropout issues of a Web-based intervention in the ICD population. Our findings show that 23.3% of patients randomized to the treatment arm completed the full treatment (six lessons), while 16.5% never logged on to the intervention. A gradual decline in adherence was observed with more patients dropping out as the lessons proceeded. The three groups (completers, treatment dropout, and dropout) did not differ systematically on any demographic or clinical baseline characteristics and their psychological profile. The top 3 reasons given for dropping out of the treatment were: technical issues with the computer/website, time constraints, and feeling fine/not needing additional support.
The findings of this study are generally in line with previous findings from the “Alles Onder Controle” Web-based intervention for individuals from the general population with increased levels of anxiety or depression who volunteered to undergo the intervention. Although generally higher percentages of completers were reported (38-55%), the rate of patients who never logged on was between 9% to 16% [
In general, higher attrition rates are reported in open access Web-based trials without therapist guidance [
Studies of Web-based interventions have also shown that the duration of treatment [
A number of limitations of this study must be acknowledged. First, current analyses are based on a relatively small sample and should be replicated in larger studies in the future. Second, results on reasons for dropout are based on descriptive data that were obtained via a telephone call to patients. A structured interview or validated questionnaire would perhaps provide more valid information. Third, unfortunately we were not able to reach all patients at 12 weeks by telephone; hence, current findings are based on patients who answered the phone and were willing to provide us with information regarding their reasons for dropout.
This study also has several strengths. To our knowledge, it is the first behavioral intervention trial in ICD patients to have used a Web-based approach and the first study in ICD patients overall to address issues of adherence and attrition. The information from WEBCARE adds to our knowledge about factors that may influence adherence in trials using a Web-based approach, which can be used when designing Web-based behavioral intervention trials for ICD patients in the future, in order to increase the number of patients enrolled in the study and their treatment adherence.
When offering a Web-based intervention to ICD patients, it seems of great importance to make the intervention as patient-tailored as possible. Not all patients have the same needs at the same time. Thus, giving them time and space to complete the lessons when needed is an important factor as is making it possible to select which lessons to complete (some patients are more interested in technical aspects of the ICD, which would give them more reassurance, while others prefer psychological support in dealing with this new situation). As a proportion of the ICD patients indicated that they were feeling fine and did not need any additional support, it might be more important to focus on patients who have higher distress levels post implant. These patients could be identified using brief and standardized questionnaires that are designed to assess psychological distress. Close monitoring of patients’ psychological needs is warranted as it has been associated with morbidity and mortality [
In conclusion, as Web-based treatments are increasingly being implemented in clinical practice, knowing how to keep patients motivated and compliant with treatment becomes more important every day. Our findings indicate that more attention should be paid to the technical aspects of Web-based treatment and making it more user-friendly. In addition, to overcome the barrier of home computers not working as they should, future studies should examine whether a similar intervention could be delivered using smartphones or tablets in order to decrease dropout. Also, future studies should examine the relationship between adherence and outcomes, as the results to date are inconclusive [
analysis of variance
Charlson Comorbidity Index
Generalized Anxiety Disorder scale
implantable cardioverter defibrillator
left ventricular ejection fraction
New York Heart Association functional class
Patient Health Questionnaire
electrocardiogram Q, R, and S waves
randomized controlled trial
usual care (group)
WEBCARE (group)
WEB-based distress management program for implantable CARdioverter dEfibrillator patients
This work was in part supported with grant no. 300020002 (with support of the Dutch Heart Foundation) and a VIDI grant (91710393) from the Netherlands Organization for Health Research and Development (ZonMW), The Hague, The Netherlands, to Dr Susanne S Pedersen. We would like to thank all the ICD and research nurses from the participating centers for their assistance with patient recruitment. In addition, we would like to thank Sophie Truijens, Harriët Abrahams, Leonie Visser, Eva Broers, Ferry van Ekelen, Annemiek de Wit, and Annick van Manen for their help with data management.
None declared.
CONSORT-EHEALTH checklist V1.6.2 [