This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), 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.
Internet-based cognitive behavioral therapy (I-CBT) leads to a reduction of fatigue severity and disability in adults with chronic fatigue syndrome (CFS). However, not all patients profit and it remains unclear how I-CBT is best embedded in the care of CFS patients.
This study aimed to compare the efficacy of stepped care, using therapist-assisted I-CBT, followed by face-to-face (f2f) cognitive behavioral therapy (CBT) when needed, with f2f CBT (treatment as usual [TAU]) on fatigue severity. The secondary aim was to investigate treatment efficiency.
A total of 363 CFS patients were randomized to 1 of the 3 treatment arms (n=121). There were 2 stepped care conditions that differed in the therapists’ feedback during I-CBT:
Noninferiority was indicated, as the upper boundary of the one-sided 98.75% CI of the difference in the change in fatigue severity between both forms of stepped care and TAU were below the noninferiority margin of 5.2 (4.25 and 3.81, respectively). The between-group differences on the secondary outcomes were also not significant (
Stepped care, including I-CBT followed by f2f CBT when indicated, is noninferior to TAU of f2f CBT and requires less therapist time. I-CBT for CFS can be used as a first step in stepped care.
Nederlands Trial Register NTR4809; http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=4809 (Archived by WebCite at http://www.webcitation.org/74SWkw1V5)
Chronic fatigue syndrome (CFS), sometimes referred to as myalgic encephalomyelitis (ME), is characterized by medically unexplained, severe, ongoing, and disabling fatigue. It is not alleviated by rest, and according to the revised US Centers for Disease Control (CDC) criteria from 2003, it is accompanied by at least 4 out of the following 8 additional symptoms: postexertional malaise, headache, unrefreshing sleep, muscle pain, joint pain, sore throat, tender lymph nodes, and impaired concentration or memory [
It is helpful to distinguish between precipitating and perpetuating factors of CFS [
CBT for CFS is a time-intensive treatment, requiring 12 hours of therapist contact on average [
In the past years, I-CBT was developed for several mental disorders and chronic medical conditions, and its efficacy has been tested in randomized controlled trials (RCTs) [
We developed I-CBT for adults with CFS and recently tested its efficacy in an RCT. I-CBT led to a reduction of fatigue and disabilities compared with a waiting list [
It is important to uncover how I-CBT can best be embedded in clinical care for CFS. The effect size of I-CBT on fatigue severity was smaller than the effect size previously found in a study investigating the efficacy of f2f CBT delivered in groups, using the same treatment principles (effect sizes 0.6 and 1.1, respectively) [
The primary objective of this study was to determine whether stepped care for CFS, with I-CBT as the first and f2f CBT as the second step, would be as effective as TAU. This randomized noninferiority trial was a follow-up study of the RCT investigating the efficacy of I-CBT for adults with CFS. All patients who were still severely fatigued or functionally impaired after I-CBT or the waiting list were offered f2f CBT, resulting in 2 stepped care conditions and 1 TAU arm. The efficacy in reduction of fatigue and efficiency in reduction of therapist time needed for stepped care were compared with TAU.
Moreover, we explored the role of therapist guidance in I-CBT, as it is not well understood how therapist feedback influences I-CBT. Guided I-CBT was found to be more effective than unguided I-CBT [
This study was registered in the Dutch trial register (NTR4809) and was approved by the medical ethical committee of the Radboud University Medical Center (reference NL42543.091.12). It was a follow-up study of an RCT testing the efficacy of I-CBT for CFS (NTR4013) and was designed as a three-arm, parallel, randomized, noninferiority trial. In 2 arms, the patients received stepped care (SC) consisting of I-CBT, either with protocol-driven feedback (SC-protocol-driven feedback) or with feedback on demand (SC-feedback-on-demand), followed by f2f CBT when necessary, that is, still severely fatigued (CIS fatigue severity >35) or disabled (SIP >700) after I-CBT. The third arm was f2f CBT after a variable waiting period (TAU).
Before randomization, all patients completed a baseline assessment (T0). Directly after randomization, the patients in the stepped care arms started with I-CBT for a duration of 6 months. After 6 months, they completed a second assessment (T1). If they started additional f2f CBT after T1, they were assessed again 6 months later (T2). Patients from the TAU group were placed on the waiting list directly after T0, for a duration of maximally 6 months (refer to the section Interventions), after which they completed their T1 assessment. Directly after T1, the f2f CBT was started. Furthermore, 6 months after T1, the patients were assessed again (T2). Assessments were web-based questionnaires.
Participants were all recruited from adult CFS patients consecutively referred to a tertiary CFS treatment center at a university hospital. They comprised the 240 participants of the RCT testing the efficacy of I-CBT [
Eligible patients were informed about the study and were included after giving written informed consent. They were asked to refrain from seeking treatment for CFS elsewhere for the duration of the study. Inclusion criteria were being aged 18 years or above; ability to speak, read, and write Dutch; meeting the CDC criteria, revised in 2003, for CFS [
For patients in the TAU group, the therapy started after a waiting period. During the first part of the study (first 240 randomized patients), the waiting list was 6 months for all patients. In the second part of the study (last 123 patients), the waiting period fluctuated depending on treatment capacity but was not longer than 6 months. For ethical reasons, patients could start sooner if the waiting period for routine clinical care was less than 6 months.
CBT was delivered according to a treatment protocol [
The I-CBT was accessible via a website, with a username and password that the patient received by email. Patients did not have to pay to use the intervention, although internet access was not provided. Patients were not trained to use the platform and could ask for support by email. During the f2f diagnostic sessions, the patient had met the therapist who would deliver the I-CBT. Both versions of I-CBT were based on the treatment protocol for f2f CBT for CFS [
The I-CBT consisted of 7 modules: (1) In “getting started and goal setting” psychoeducation is given, a treatment contract is signed, and goals are set. When goals are submitted, the next 5 modules are accessible: (2) “regulate sleep-wake cycle,” (3) “helpful beliefs about fatigue,” (4) “how to communicate with others about CFS,” (5) “gradually increasing my activity,” and (6) “reaching my goals step by step.” When this module is completed, the last module becomes accessible: (7) “evaluation and the future.” The modules are described in more detail elsewhere [
Within 2 weeks after T1, patients had an f2f evaluation session with their therapist. During this session, it was evaluated whether treatment goals were reached or additional f2f CBT was indicated. Patients were offered f2f CBT after I-CBT if they were still severely fatigued (CIS fatigue severity ≥35) and/or severely disabled (SIP8 ≥700) and/or if not all therapy goals were attained and the therapist expected that additional treatment was necessary to attain them. This decision could only be made after consultation of an experienced supervising CBT therapist (HK). The reason for continuation was, in that case, recorded and reported.
The additional CBT was delivered according to the treatment protocol [
Therapists in all the 3 arms were psychologists, trained in CBT for CFS, both f2f and internet-based. Therapists received weekly group supervision during the study [
The following patient characteristics were recorded: age, sex, duration of symptoms, the presence of each CDC symptom, work status, years of education followed, and the presence of depressive symptoms above a clinical significant cutoff of 4 on the Beck Depression Inventory, primary care version [
Fatigue severity (the primary, noninferiority outcome) was measured with the Checklist Individual Strength (CIS) fatigue severity subscale [
A total score of Sickness Impact Profile 8 (SIP8; [
The Medical Outcome Survey Short Form-36 (SF-36; [
The Symptom Checklist-90 (SCL-90 [
Therapists recorded the time needed per patient. Time spent on I-CBT (in stepped care), on additional telephone calls or email contact (all conditions) were recorded in minutes. Each f2f CBT session counted for 60 min (all conditions). In all conditions, 120 min was counted for the diagnostic assessment, consisting of 2 sessions.
Assuming a power of 0.80, a one-sided alpha of .0125 (correcting for 2 comparisons), an SD of 13.6 [
Patients were randomly allocated to 1 of the 3 conditions. Randomization was computer-generated, in blocks of 12 patients. The randomization program was programmed by a statistician, not involved in this study. Randomization was performed by an administrative assistant, in the presence of the patient and the therapist. Participants were partly blinded: they were unaware of the existence of 2 I-CBT formats and were told that they could either receive I-CBT followed by CBT if needed or f2f CBT. After randomization, therapists and patients could read the result from the computer screen: “(1) internet therapy” and “(2) internet therapy” or “wait list.” The therapist knew that condition “1” of I-CBT was protocol-driven feedback and “2” was feedback on demand.
Statistical analysis was performed on a data file, which blinded the researcher performing the analysis to patient and allocation condition. Post analysis, allocation to condition was unmasked to enable the authors to interpret the results.
Analyses were done on the basis of intention to treat after imputation of missing primary and secondary outcome measures at postassessment. Postassessment was T1 for patients who stopped after waiting list or after I-CBT and was T2 for patients who received (additional) f2f CBT. When T2 data were needed but only T1 was present, postassessment scores were imputed as well. We used multiple imputation (20 imputed data sets), assuming that data were missing at random. All outcome variables at postassessment were included in the imputation model and were imputed. Baseline variables of all outcomes were only entered as predictors for the imputation model and were not imputed. Imputation was done in IBM SPSS version 22, as were all statistical analyses.
To answer the primary research question, an analysis of covariance (ANCOVA) was performed with postassessment CIS fatigue severity score as the dependent variable, baseline fatigue scores as covariate, and treatment condition as the fixed factor. Noninferiority was assumed when post-treatment CIS fatigue severity was maximally 5.2 points higher for either form of stepped care in comparison with TAU [
Differences in the proportion of patients with clinically significant improvement in fatigue severity were compared with chi-Square tests. Each stepped care group was compared with TAU. Clinically significant improvement in fatigue severity was assumed when there was a statistically reliable change of >1.96 SD in combination with a CIS fatigue scores of <35 on postassessment. The reliable change index (RCI) of the CIS fatigue severity was calculated for each person following Jacobson and Truax [
In addition, 3 sensitivity analyses were performed. First, the noninferiority was tested, whereas missing observations on fatigue severity were not imputed with multiple imputations but replaced in the following manner: (1) it was hypothesized that patients in the TAU group improved, missing scores were replaced by the CIS fatigue severity score at T0 minus the mean change in fatigue of the TAU group and (2) it was hypothesized that patients of the stepped care groups deteriorated, scores were replaced by the maximum CIS fatigue severity score. Second, the main analysis was repeated excluding patients who received CBT by telephone or email, instead of f2f. Third, the main analysis was repeated selecting only patients who met the SEID criteria.
Therapist time of both stepped care conditions were compared with TAU (either means with
For patients in the stepped care groups who received additional f2f CBT, whether treatment outcomes at T2 (after f2f CBT) differed significantly from T1 (after I-CBT) was investigated using paired-samples
This analysis compared 2 groups of patients. The first 240 randomized patients were included between April 2013 and June 2015 [
Patients were recruited between April 2013 and December 2016 and data collection was completed in December 2017. As shown in the flowchart (
In the TAU condition, the waiting list duration was 29 weeks (SD 4) for the first part of the study and 18 weeks (SD 12) for the second part of the study. The Spearman rank correlation between waiting list duration and reduction of fatigue was
In the SC-protocol-driven feedback condition, 116 out of the 121 patients started I-CBT (95.8%). At T1, 87 patients met the step-up criteria (71.9%). Of them, 37 (43%) started f2f CBT. Furthermore, 3 patients, who did not meet the step-up criteria, received f2f CBT because several goals were not reached. In addition, 1 patient received the additional CBT by telephone and email.
In the SC feedback-on-demand condition, 113 patients started I-CBT (93.3%). Of the 85 patients who met the step-up criteria (70.2%), 48 (56%) started f2f CBT. Moreover, 6 patients who did not meet the step-up criteria received f2f CBT because not all treatment goals were attained.
The difference in the proportion of patients meeting the step-up criteria that did actually step up to f2f CBT between both stepped care conditions was not significant (χ21=3.3;
At the end of the study, 1 patient was still in therapy and completed the postassessment 6 months after the start of f2f CBT. None of the patients were excluded from the main analysis.
Data on the primary outcome were missing for 6 patients (all in the TAU condition). The upper boundary of the one-sided 98.75% CI of both forms of stepped care was below 5.2, indicating noninferiority. The mean difference in fatigue severity for the SC-protocol-driven feedback condition with TAU was −0.04 points; the CI upper bound was 3.81. For the SC-feedback-on-demand condition, the difference was 0.41 points; the upper bound of the CI was 4.25.
In all sensitivity analyses, both stepped care conditions remained noninferior to TAU (upper boundary CI for SC-protocol-driven feedback ranging from 3.34-3.95; upper bound CI SC-feedback-on-demand ranging from 4.12-4.50).
There were no significant differences between either stepped care condition or TAU for all secondary outcomes, that is, level of disabilities (SIP8), physical functioning (SF-36 physical functioning), and psychological distress (SCL-90). Within-group effect sizes are provided in
Post stepped care, 49 out of 121 (40%) patients in SC-Protocol-driven feedback during I-CBT showed clinically significant improvement in fatigue severity. In the TAU group, 53 out of 115 patients improved. Assuming that patients with missing data did not improve, this was 44%. The difference between both groups was not significant; χ21(N=236)=0.7;
In SC-feedback-on-demand, 61 patients improved (50%). The difference between SC-feedback-on-demand and TAU was also not significant: χ21(N=236)=0.4;
As the data on therapist time were not normally distributed, medians were compared with Mann-Whitney
The analysis including only patients who started treatment showed different results. Median therapist time of both the stepped care formats differed significantly from TAU (
CONSORT flow diagram of eligibility criteria. CBT: cognitive behavioral therapy; CDC: Centers for Disease Control and Prevention; CFS: chronic fatigue syndrome; f2f: face-to-face; I-CBT: internet-based cognitive behavioral therapy.
Baseline characteristics.
Baseline characteristic | Stepped care | Treatment as usual | |
Protocol-driven feedback | Feedback on demand | ||
Age in years, mean (SD) | 36.6 (12.8) | 37.2 (12.3) | 38.7 (12.5) |
Female (N=121), n (%) | 78 (64.5) | 69 (57.0) | 74 (61.2) |
Paid job (N=119), n (%) | 82 (68.9) | 79 (65.8) | 77 (64.7) |
Education level in years, mean (SD) | 15.4 (1.9) | 14.8 (2.3) | 15.7 (1.5) |
Duration of complaints in years, median (IQRa) | 4.0 (8.0) | 5.0 (12.0) | 6.0 (9.0) |
Number of CDCb symptoms, median (IQR)c | 6 (2) | 6 (3) | 6 (2) |
Memory and/or concentration problems (N=121), n (%) | 114 (94.2) | 112 (92.6) | 116 (95.9) |
Sore throat (N=121), n (%) | 53 (43.8) | 57 (47.1) | 56 (46.3) |
Tender lymph nodes (N=121), n (%) | 49 (40.5) | 61 (50.4) | 52 (43.0) |
Muscle pain (N=121), n (%) | 91 (75.2) | 97 (80.2) | 99 (81.8) |
Multi-joint pain (N=121), n (%) | 83 (68.6) | 86 (71.7) | 93 (76.9) |
Headaches (N=121), n (%) | 91 (75.2) | 94 (77.7) | 85 (70.2) |
Unrefreshing sleep (N=121), n (%) | 119 (98.3) | 114 (94.2) | 119 (98.3) |
Postexertional malaise (N=121), n (%) | 113 (93.4) | 109 (90.1) | 113 (93.4) |
Meeting SEIDd criteria (N=121), n (%) | 89 (73.6) | 92 (76.0) | 93 (76.9) |
Fatigue severitye, mean (SD) | 50.8 (5.0) | 50.2 (4.8) | 49.7 (5.3) |
Overall impairmentf, mean (SD) | 1488.6 (550.1) | 1534.7 (562.0) | 1599.2 (589.6) |
Physical functioningg, mean (SD) | 62.3 (20.1) | 60.5 (19.4) | 61.0 (20.4) |
Psychological distressh, mean (SD) | 154.4 (31.8) | 160.2 (37.7) | 161.2 (38.0) |
Clinically relevant depressive symptomsi (N=120), n (%) | 42 (35.0) | 39 (32.5) | 44 (37.0)j |
No current psychiatric diagnosisk (N=121), n (%) | 100 (83) | 103 (85) | 99 (82) |
Any depressive disorderk (N=121), n (%) | 11 (9) | 9 (7) | 14 (12) |
Any anxiety disorderk (N=121), n (%) | 11 (9) | 11 (9) | 8 (7) |
Other psychiatric disorderk (N=121), n (%) | 1 (1) | 1 (1) | 4 (3) |
Paing, mean (SD) | 59.3 (25.5) | 59.1 (25.5) | 57.7 (25.2) |
aIQR: interquartile range.
bCDC: Centers for Disease Control and Prevention.
cMemory and/or concentration problems were scored together, so a maximum of 8 symptoms was scored.
dSEID: systemic exertion intolerance disease.
eCIS: Checklist Individual Strength.
fSIP8: Sickness Impact Profile.
gSF-36: Medical Outcomes Survey Short Form-36.
hSCL-90: Symptom Checklist-90.
iBDI-PC: Beck Depression Inventory-PC; total score ≥4.
jN=119.
kMINI: The Mini-International Neuropsychiatric Interview.
Treatment effects.
Outcome measure | Stepped care | Treatment as usual | ||
Protocol-driven feedback | Feedback-on-demand | |||
Pre | 50.78 | 50.20 | 49.69 | |
Post | 35.60 | 35.68 | 34.94 | |
Cohen |
1.44 (1.16-1.73) | 1.50 (1.21-1.79) | 1.41 (1.12-1.69) | |
Pre | 1488.56 | 1534.74 | 1593.20 | |
Post | 822.09 | 797.10 | 961.32 | |
Cohen |
1.09 (0.82-1.36) | 1.22 (0.94-1.50) | 0.91 (0.65-1.18) | |
Pre | 62.27 | 60.54 | 60.95 | |
Post | 75.34 | 77.82 | 76.54 | |
Cohen |
0.58 (0.33-0.84) | 0.86 (0.59-1.12) | 0.72 (0.46-0.98) | |
Pre | 154.36 | 160.20 | 161.22 | |
Post | 137.69 | 140.79 | 143.65 | |
Cohen |
0.42 (0.17-0.68) | 0.46 (0.21-0.72) | 0.40 (0.14-0.65) |
aCIS: Checklist Individual Strength.
bUncontrolled effect size: within-group effect. Cohen
cSIP8: Sickness Impact Profile 8.
dSF-36: Medical Outcomes Survey Short Form-36.
eSCL-90: Symptom Checklist-90.
Therapist time invested in total treatment in hours.
Treatment arm | Intention to treat | Starters only | ||||||
n | Mean (hours) | Median (hours) | Minimum-maximum (hours) | n | Mean (hours) | Median (hours) | Minimum-maximum (hours) | |
Stepped care–protocol-driven feedback | 121 | 09:10 | 08:00 | 2:55-22:20 | 118 | 09:19 | 08:09 | 2:55-22:20 |
Stepped care–feedback-on-demand | 121 | 08:30 | 06:55 | 2:00-21:45 | 117 | 08:42 | 07:25 | 2:00-21:45 |
Treatment as usual | 121 | 08:54 | 09:00 | 2:00-27:00 | 83 | 12:03 | 12:00 | 4:00-27:00 |
The difference in time invested between both the stepped care versions was significant when analyzing data of all patients (
In total, 95 patients received f2f CBT after I-CBT. As shown in
Therapist time in stepped care.
Selected group | Stepped care–protocol-driven feedback | Stepped care–feedback-on-demand | ||||||||||
n | Mean total time (hours) | Mean I-CBTa time (hours) | Percentage of total time | Mean f2f CBTb time (hours) | Percentage of total time | n | Mean total time (hours) | Mean I-CBT time (hours) | Percentage of total time (hours) | Mean f2f CBT time (hours) | Percentage of total time (hours) | |
Total group | 121 | 9:10 | 4:04 | 44 | 2:06 | 23 | 121 | 8:30 | 2:29 | 29 | 2:54 | 34 |
No f2f CBT | 81 | 7:00 | 4:06 | 59 | 0:00 | 0 | 67 | 5:11 | 2:19 | 45 | 0:00 | 0 |
Received f2f CBT | 40 | 13:32 | 3:59 | 29 | 6:18 | 47 | 54 | 12:37 | 2:43 | 21 | 6:30 | 52 |
aI-CBT: internet-based cognitive behavioral therapy.
bf2f CBT: face-to-face cognitive behavioral therapy.
Treatment effect of cognitive behavioral therapy after internet-based cognitive behavioral therapy.
Outcome measure | At T1 (after I-CBTa), mean (SD) | At T2 (after CBT), mean (SD) | ||
CISb fatigue severity (N=95) | 42.99 (9.35) | 37.39 (12.06) | 4.901 (94) | <.001 |
SIP8c (N=91) | 1151.86 (660.76) | 851.18 (673.74) | 4.569 (90) | <.001 |
SF-36d physical functioning (N=91) | 71.54 (21.89) | 77.03 (21.36) | -2.866 (90) | .005 |
SCL-90e (N=86) | 146.56 (33.29) | 144.52 (46.01) | 0.517 (85) | .607 |
aCBT: cognitive behavioral therapy.
bCIS: Checklist Individual Strength.
cSIP8: Sickness Impact Profile.
dSF-36: Medical Outcomes Survey Short Form-36.
eSCL-90: Symptom Checklist-90.
There was a significant difference between the 2 groups regarding reduction of fatigue severity. In the first 240 patients, the change score was on average 16.4 points on the CIS fatigue severity subscale (SD 13.9). In the last 123 patients, the change score was 11.8 points (SD 11.3),
This study showed that I-CBT embedded in stepped care for chronic fatigue syndrome is noninferior to f2f CBT (TAU) in reducing fatigue severity. Treatment outcome of stepped care did not differ from TAU with respect to the level of disability, physical functioning, and psychological distress. The proportions of patients with clinically significant improvement of fatigue severity were equal for stepped care as well as TAU. Interestingly, this was despite the fact that approximately 50% of the patients who met the step-up criteria for f2f CBT after I-CBT did not step up.
For patients who did step up, it was found that f2f CBT after I-CBT led to a significant further improvement in fatigue severity and impairment. This suggests that stepped care with I-CBT as a first step is a viable treatment model for CFS. It was more time-efficient than usual care, as approximately 25% less therapist time was needed to deliver it. This is an important finding from a cost-effectiveness perspective, since therapist time accounts for a large proportion of the treatment costs in mental health care. Previous studies found stepped care, not including I-CBT, to be effective for CFS [
How therapist feedback during I-CBT influenced stepped care was explored. As in our previous study [
We also explored if the proportion of patients willing to step up after I-CBT differed in both the feedback formats. It was found that of the patients who needed to step up after I-CBT, the proportion of patients that received f2f CBT differed in favor of the feedback-on-demand format. However, this difference was not significant. It is important to know what prevented patients from stepping up. One plausible explanation could be that patients became less motivated after unsuccessful I-CBT [
Although this study shows that I-CBT fits well into a stepped care model, it is problematic that many patients do not step up when this is needed. Although this did not lead to a lower proportion of improved patients (than in TAU), and it is uncertain whether these patients would have otherwise started and completed TAU, it is possible that some of these patients would have profited more if they had received f2f CBT. To improve the integration of I-CBT in clinical care, there are some options. An aim could be to increase the number of patients that step up for f2f CBT. For example, by stopping earlier with I-CBT when it appears to be ineffective, demoralization can be prevented. However, what is a reliable indicator of the need to step up is not known. In a process study on f2f CBT for CFS, it was found that in a substantial proportion of patients, symptoms did not decrease until after 3 to 4 months [
It is also possible to further improve I-CBT, so that fewer patients need to step up. An option would be to improve I-CBT by developing a more flexible version, in which the intensity of therapist guidance can be further varied. The integration of videoconferencing in I-CBT makes it possible to combine the advantages of f2f sessions and I-CBT. Perhaps this also can be
A significant limitation of our study was that in TAU, one-third of the patients did not start the therapy. This may have reduced the treatment results in this reference group. The within-group effect size for fatigue severity in the TAU condition in our study fell outside the 95% CIs of 2 other CFS studies that had lower proportions of nonstarters [
Furthermore, a possible limitation is that the waiting list policy had changed during the study. However, retaining an unnecessarily long waiting list duration would not have been ethical. It was found that the duration of the waiting list had not influenced the treatment effect on reduction of fatigue.
Another limitation is that we did not provide data on dropout. It was difficult to produce a dropout definition that allowed for a comparison between all treatment arms. For TAU, termination of treatment before a certain number of sessions is often used as a criterion for dropout [
Recently, new diagnostic criteria were proposed for CFS, by the US Institute of Medicine [
There were important strengths of the study: it is one of the few RCTs that compared stepped care with I-CBT as a first-step treatment with usual care [
In conclusion, this RCT showed that stepped care with I-CBT as the first and f2f CBT as the second step for chronic fatigue syndrome is noninferior to TAU in the reduction of fatigue severity. A substantial part of the patients did not need to step up after stepped care, which made treatment for them less intensive. For patients who needed to step up, f2f CBT led to an additional treatment effect. Stepped care took less therapist time to deliver. However, a substantial proportion of patients did not step up after I-CBT with limited results. Treatment efficacy can probably be improved when relatively more patients step up and by further developing the I-CBT.
CONSORT-EHEALTH checklist (V 1.6.1).
analysis of covariance
cognitive behavioral therapy
Centers for Disease Control and Prevention
chronic fatigue syndrome
Checklist Individual Strength
face-to-face
internet-based cognitive behavioral therapy
myalgic encephalomyelitis
reliable change index
randomized controlled trials
stepped care
Symptom Checklist-90
systemic exertion intolerance disease
Medical Outcomes Survey Short Form-36
Sickness Impact Profile
treatment as usual
The authors thank all the patients for their participation and all therapists involved for delivering CBT. They also thank T Fasotti-Dumont, S van Hees, J Smeets, and L Vermeeren for their help with data collection and administration as well as R Donders and P Nieuwkerk for statistical advice and H Abrahams for performing the statistical analyses. The authors also thank J Rosmalen for her comments on a draft of the manuscript.
HK, AJ, and MW-S designed the study. MW-S and AJ were involved with data acquisition. MW-S conducted the analysis. Interpretation of data was performed by HK and MW-S. All authors were involved in the interpretation of results as well as reviewing and revising the manuscript.
HK receives royalties for the treatment protocol of f2f-CBT, which was published.