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.
A fully automated computer-tailored Web-based self-management intervention, Kanker Nazorg Wijzer (KNW [Cancer Aftercare Guide]), was developed to support early cancer survivors to adequately cope with psychosocial complaints and to promote a healthy lifestyle. The KNW self-management training modules target the following topics: return to work, fatigue, anxiety and depression, relationships, physical activity, diet, and smoking cessation. Participants were guided to relevant modules by personalized module referral advice that was based on participants’ current complaints and identified needs.
The aim of this study was to evaluate the adherence to the module referral advice, examine the KNW module use and its predictors, and describe the appreciation of the KNW and its predictors. Additionally, we explored predictors of personal relevance.
This process evaluation was conducted as part of a randomized controlled trial. Early cancer survivors with various types of cancer were recruited from 21 Dutch hospitals. Data from online self-report questionnaires and logging data were analyzed from participants allocated to the intervention condition. Chi-square tests were applied to assess the adherence to the module referral advice, negative binominal regression analysis was used to identify predictors of module use, multiple linear regression analysis was applied to identify predictors of the appreciation, and ordered logistic regression analysis was conducted to explore possible predictors of perceived personal relevance.
From the respondents (N=231; mean age 55.6, SD 11.5; 79.2% female [183/231]), 98.3% (227/231) were referred to one or more KNW modules (mean 2.9, SD 1.5), and 85.7% (198/231) of participants visited at least one module (mean 2.1, SD 1.6). Significant positive associations were found between the referral to specific modules (range 1-7) and the use of corresponding modules. The likelihoods of visiting modules were higher when respondents were referred to those modules by the module referral advice. Predictors of visiting a higher number of modules were a higher number of referrals by the module referral advice (β=.136,
The KNW in general and more specifically the KNW modules were well used and highly appreciated by early cancer survivors. Indications were found that the module referral advice might be a meaningful intervention component to guide the users in following a preferred selection of modules. These results indicate that the fully automated Web-based KNW provides personal relevant and valuable information and support for early cancer survivors. Therefore, this intervention can complement usual cancer aftercare and may serve as a first step in a stepped-care approach.
Nederlands Trial Register: NTR3375; http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=3375 (Archived by WebCite at http://www.webcitation.org/6jo4jO7kb)
Recovery from cancer and its treatment can be challenging for cancer survivors. A variety of physical, psychosocial, and lifestyle difficulties might impede the resumption of previous daily life activities [
The Web-based intervention
Previously published Web-based interventions in the areas of lifestyle, mental health, and chronic conditions differ with regard to the number of (cancer-related) topics, the composition of the target group, the intervention components, and the delivery mode [
The design of the KNW portal differs from most of the existing Web-based interventions for cancer survivors by providing personalized self-management training on seven topics and by allowing users to choose which modules they want to use during an intervention period of 6 months. Previously identified effective intervention characteristics of Web-based lifestyle interventions were tailored feedback, the use of theory, interactivity, goal setting, and online or in-person contact [
The main objective of this study is threefold: (1) to describe the use of the KNW modules and to identify predictors of a higher number of modules used, (2) to investigate the adherence to the provided MRA, and (3) to describe the appreciation of the KNW and its predictors. Additionally, to explore how well the tailoring worked and whether the perceived personal relevance might be different among subgroups, we explored possible predictors of personal relevance.
Overview of the scope and sequence of the modules. From Willems et al (2015). Used with permission.
Module Referral Advice that encourages participants to follow relevant KNW modules. Adapted from Willems et al (2015). Used with permission.
This process evaluation was conducted as part of a two-armed randomized controlled trial (RCT) that evaluates the effects of the KNW portal. For the purpose of this report, all respondents of the intervention condition were included in the analyses. The details of the trial design, sample size calculation, participant eligibility, recruitment procedures, and the intervention have been published elsewhere [
A comprehensive description of the intervention, including the eight KNW modules, the underlying theoretical frameworks, and technical features are published in detail elsewhere [
Classification of the green, orange, and red MRA.
Measurements and classification criteriaa | MRA categories | ||||
Green | Orange | Red | |||
Fatigue | CIS, subscale subjective fatigue (1-56) [ |
<27 | 27-35 | >35 | |
Return to work | Extended CaSUN [ |
No needs | Score on needs 3-12 | Score on needs ≥13 | |
Mood | HADS-A (0-21); HADS-D (0-21) [ |
HADS-A<8 and HADS-D<8 and MAC ≤36 | HADS-A < 8 and HADS-D 8-15 and/ or MAC > 36; HADS-A 8-15 and HADS-D <8 or 8-15 | HADS-A < 8 or 8-15 and HADS-D >15; HADS-D < 8 or 8-15 and HADS-A >15 | |
Relationships | SSL-D (6-24) [ |
SSL-D ≤7 | SSL-D=8 or 9 & needs CaSUN | SSL-D ≥10 & needs CaSUN | |
Physical activity | SQUASH [ |
Meeting both conditions | Meeting 1 out of 2 conditions | Meeting no conditions | |
Diet | Dutch Standard Questionnaire on Food Consumption [ |
Meeting at least 4 out of 5 conditions | Meeting 2 or 3 out of 5 conditions | Meeting 1 or 0 out of five conditions | |
Smoking | Smoking, not smoking, time point of quitting [ |
Never/formersmokers, quit prior to cancer diagnosis | Quit smoking after cancer diagnosis | Current smokers |
aCIS: Checklist Individual Strength; PA: physical activity; HADS: Hospital Anxiety and Depression Scale, HADS-A: subscale anxiety, HADS-D: subscale depression; MAC: Mental Adjustment to Cancer Scale; SSL-D: Social Support List‒discrepancy subscale; SQUASH: Short Questionnaire to Assess Health Enhancing Physical Activity
bNeeds related to sexuality and fertility.
cWhole-grain bread, oatmeal, cereals.
Throughout the different KNW intervention modules, principles of problem-solving therapy, cognitive behavioral therapy, social cognitive theories, and self-regulation theories were applied [
All data were derived from online self-report questionnaires and logging details.
Module use was assessed by using logging data. Actual use was dichotomized (yes/no) for each module separately (in total eight modules). Module use was categorized into “yes” when at least the first three pages of a module were used. These three pages comprised important key information after which participants followed personalized pathways through the modules. The individual pathways were based on the responses to the baseline questionnaire, own preferences and goals, and take into consideration that the amount of needed information and/or support can vary to initiate behavior change [
At 6-month follow-up, the overall rating of the KNW and separate ratings for each of the used module(s) were assessed on a scale ranging from 1 (very poor) to 10 (outstanding) (eg, “Overall, how do you rate the KNW? Select your rating (1-10)”; “How do you rate module mood on a scale from 1 to 10”). Further, four separate items were measured to evaluate whether the provided information and support was understandable, useful, personally relevant, and recommendable to fellow patients, on a 5-point Likert-scale, ranging from 1 (low) to 5 (high). The perceived personal relevance (“Was the information from the Kanker Nazorg Wijzer of personal relevance for you?”) was included in the analysis of this study to explore whether computer tailoring worked well within the KNW. These items correspond to items that were used in other studies to measure the appreciation of Web-based interventions [
Information about demographic and cancer-related characteristics was collected at baseline. Standard questions were used to measure age, gender, and marital status. Marital status was dichotomized into “with partner” (married, cohabiting partners) and “without partner” (single, divorced, widowed). Education level was categorized into “low” (lower vocational education, medium general secondary education), “medium” (secondary vocational education, higher general secondary education), and “high” (higher vocational education, university education). Employment status was dichotomized into “working” (self-employed, in paid employment) and “not working” (unemployed, retired, unable to work). Type of cancer was categorized into breast, colorectal, and other types of cancer (ie, bladder, esophageal, gynecologic, hematologic, kidney, liver, lung, prostate, stomach, testicular, and thyroid cancer). Type of treatment was categorized into surgery and chemotherapy and radiotherapy, surgery and chemotherapy, surgery and radiotherapy, and other types of treatment. Further, aftercare (yes/no) and comorbidities (yes/no) were measured, and height and weight were assessed to determine BMI. The time since completion of primary treatment in weeks was based on registry data from the hospitals.
The analyses were performed using STATA version 13.1. Descriptive statistics were used to describe demographic and cancer-related characteristics of the module (non-) users and the number of weeks of module engagement among all participants of the intervention condition at baseline. To calculate the appreciation outcomes, participants who completed the relevant questions at the 6-month measurement and who used the corresponding modules were included. Chi-square tests were used to determine the relationships between the MRA and the subsequent module use with a two-sided alpha=.05 level of significance. Negative binominal regression analysis was used to identify the predictors of a higher number of modules used (0-8), due to overdispersed count data. Independent variables (hypothesized predictors) were demographic variables (gender, age, marital status, education, employment), cancer-related variables (cancer type, type of treatment, number of weeks after completing primary cancer treatment, aftercare, comorbidities, BMI), the number of red and orange MRA, ranging from 0-7, and the perceived personal relevance, ranging from 1-5. To examine the predictors of a higher overall appreciation of the KNW, multiple linear regression analysis was applied among participants who completed the follow-up questionnaire after 6 months. The dependent variable was the overall rating of the KNW, measured at 6-month follow-up, ranging from 1-10. The same independent variables as described above were counted as predictors. Furthermore, the number of used modules (sum score 0-8) was added to the multiple linear regression model. To explore possible predictors of perceived personal relevance, ordered logistic regression analysis was conducted, taking into consideration that the dependent variable, perceived personal relevance, was an ordinal variable, ranging from 1-5. Within this analysis, all demographic and cancer-related characteristics were added as independent variables. Dummy coding was used for categorical variables including more than two categories and the continuous and ordinal variables were standardized in all conducted regression analyses. Since filling out all computer-based questions was required, and respondents were reminded automatically if a question was not answered, there were no missing data at baseline. Missing data at 6-month follow-up due to dropout were not imputed when calculating appreciation outcomes.
Baseline characteristics of the intervention participants are displayed in
The majority (80-100%) of the module users continued after reading the first three compulsory pages of the different modules. The numbers and percentages of participants who used the separate modules are displayed in
The relations between the color of MRA (respectively red, orange, green) and module use are shown in
Overall baseline characteristics of the KNW participants and categorized for module use (N=231).
Overall (N=231) | No module |
KNW Modules | |||||||||
Fatigue |
Return to work |
Mood |
Relation-ships |
Physical activity |
Diet |
Smoking |
Residual symptoms |
||||
Female, n (%) | 183 |
26 |
63 |
46 |
41 |
30 |
44 |
106 |
17 |
40 |
|
Age, mean (SD) | 55.6 |
52.5 |
55.1 |
52.8 |
54.4 |
55.9 |
56.3 |
56.0 |
51.6 |
56.2 |
|
With partner, n (%) | 193 |
27 |
65 |
43 |
37 |
31 |
42 |
109 |
16 |
36 |
|
BMI, mean (SD) | 26.0 |
27.2 |
26.2 |
25.7 |
25.3 |
26.1 |
26.1 |
25.4 |
24.8 |
25.4 |
|
Low | 76 |
13 |
23 |
12 |
15 |
12 |
18 |
42 |
9 |
13 |
|
Medium | 76 |
12 |
31 |
20 |
20 |
13 |
18 |
44 |
7 |
14 |
|
High | 79 |
8 |
28 |
21 |
14 |
13 |
15 |
48 |
7 |
20 |
|
Working at baseline, n (%) | 122 |
20 |
40 |
38 |
28 |
18 |
27 |
70 |
13 |
26 |
|
Breast | 162 |
24 |
55 |
40 |
36 |
27 |
41 |
94 |
18 |
32 |
|
Colon | 29 |
4 |
10 |
4 |
6 |
5 |
2 |
19 |
3 |
9 |
|
Other | 40 |
5 |
17 |
9 |
7 |
6 |
8 |
21 |
2 |
6 |
|
Had cancer before, n (%) | 24 |
5 |
8 |
3 |
4 |
3 |
5 |
13 |
2 |
5 |
|
Surgery, chemo, radio | 86 |
11 |
37 |
20 |
20 |
18 |
22 |
53 |
11 |
22 |
|
Surgery, chemo | 61 |
11 |
17 |
16 |
16 |
9 |
12 |
35 |
7 |
15 |
|
Surgery, radio | 46 |
5 |
15 |
11 |
10 |
5 |
11 |
26 |
3 |
8 |
|
Other | 38 |
6 |
13 |
6 |
3 |
6 |
6 |
20 |
2 |
2 |
|
Weeks since completion treatment, mean (SD) | 25.1 |
27.1 |
24.1 |
22.3 |
25.3 |
26.5 |
23.7 |
25.0 |
22.1 |
25.4 |
|
Having comorbidities, n (%) | 62 |
10 |
25 |
14 |
12 |
10 |
15 |
34 |
7 |
8 |
|
Using aftercare, n (%) | 145 |
25 |
46 |
38 |
32 |
29 |
31 |
83 |
12 |
29 |
Provided MRA and subsequent module use.
Module | Red | Orange | Green | ||||||
Followed module, % | Followed module, % | Followed module, % | |||||||
% | yes | no | % | yes | no | % | yes | no | |
Fatigue | 34.6 | 58.8 | 41.3 | 19.1 | 38.6 | 61.4 | 46.3 | 16.8 | 83.2 |
Return to work | 3.9 | 55.6 | 44.4 | 18.2 | 52.4 | 47.6 | 77.9 | 14.4 | 85.6 |
Mood | 1.7 | 25 | 75 | 28.6 | 30.3 | 69.7 | 69.7 | 17.4 | 82.6 |
Relationships | 11.7 | 25.9 | 74.1 | 19.1 | 27.3 | 72.7 | 69.3 | 11.8 | 88.1 |
Physical activity | 5.2 | 25 | 75 | 35.9 | 37.4 | 62.7 | 58.9 | 12.5 | 87.5 |
Diet | 53.3 | 50.4 | 49.6 | 42.9 | 68.7 | 31.3 | 3.9 | 44.4 | 55.6 |
Smoking | 11.7 | 48.2 | 51.9 | 3.1 | 42.9 | 57.1 | 85.3 | 3.6 | 96.5 |
Relationship between the MRA and module use (chi-square tests; df=1).
Module (yes/no) | Red compared to orange | Red compared to green | Orange compared to green | ||||||
Odds ratio |
Odds ratio |
Odds ratio |
|||||||
Fatigue | 4.599 | .032a | 2.262 |
35.485 | .000a | 7.042 |
8.332 | .004a | 3.113 |
Return to work | 0.030 | .863 | 1.136 |
10.565 | .001a | 7.404 |
28.920 | .000a | 6.515 |
Mood | 0.050 | .822 | .767 |
0.156 | .693 | 1.583 |
4.680 | .031a | 2.065 |
Relationships | 0.016 | .901 | .933 |
3.810 | .051 | 2.597 |
6.349 | .012a | 2.783 |
Physical activity | 0.696 | .404 | .186 |
1.474 | .225 | 2.333 |
18.60 | .000a | 4.173 |
Diet | 7.553 | .006a | .463 |
0.119 | .730 | 1.27 |
2.182 | .140 | 2.742 |
Smoking | 0.063 | .803 | 1.238 |
58.075 | .000a | 25.204 |
22.400 | .000a | 20.357 |
aStatistically significant result.
From the 231 participants who had access to the KNW intervention, 182 responded to the questions concerning appreciation after 6 months. The overall appreciation of the KNW was high (mean 7.5, SD 1.2) (
Appreciation of KNW after 6 months.
Overall | No module | Fatigue | Return to work | Mood | Relationships | PA | Diet | Smoking | Residual symptoms | ||
Overall KNW (1-10), mean (SD) | 7.5 |
7.1 |
7.6 |
7.6 |
7.4 |
7.4 |
7.6 |
7.5 |
7.8 |
7.4 |
|
Modules (1-10)a, mean (SD) | 7.3 |
7.0 |
7.5 |
7.2 |
7.7 |
7.6 |
8 |
6.4 |
|||
Understandable? | 4.3 |
4.1 |
4.4 |
4.4 |
4.3 |
4.5 |
4.4 |
4.4 |
4.3 |
4.4 |
|
Useful? | 3.7 |
3.7 |
3.8 |
3.7 |
3.7 |
3.7 |
3.7 |
3.7 |
3.8 |
3.4 |
|
Personal relevant? | 3.2 |
2.9 |
3.4 |
3.3 |
3.2 |
3.4 |
3.5 |
3.2 |
3.3 |
3.3 |
|
Recommendable to fellow survivors? | 3.9 |
3.6 |
3.9 |
3.9 |
3.8 |
3.7 |
4 |
3.9 |
4.1 |
3.8 |
aNo module n=18, fatigue n=47, return to work n=27, mood n=13, relationships n=11, PA n=28, diet n=77, smoking n=6, residual symptoms n=14.
bNo module n=18, fatigue n=67, return to work n=46, mood n=45, relationships n=34, PA n=45, diet n=115, smoking n=18, residual symptoms n=39.
Using a higher number of modules was predicted by a higher number of red/orange MRA (
A higher appreciation with the overall KNW was significantly predicted by a higher perceived personal relevance (
None of the demographic and cancer-related characteristics significantly predicted the perceived personal relevance of the KNW content, indicating that the KNW content was rated comparably personal relevant among individuals with different demographic and cancer-related characteristics (
This process evaluation of the Web-based KNW evaluated the automated guidance toward the KNW modules and subsequent module use, and the appreciation of this intervention. Despite the noncommittal nature of the KNW, more than 85% of the participants used one or more of the eight modules, and there was clear interest in all eight modules. This result confirms the need for wide-ranging support among early cancer survivors. Interestingly, automated referrals to specific modules were related to a higher number of modules used. Moreover, the complex KNW was highly appreciated and perceived as personal relevant by early cancer survivors.
The MRA aimed to guide the respondents toward the appropriate modules by giving feedback about current problem areas and needs. Cancer survivors might not have noticed some of these needs, and the MRA may have raised awareness about these topics. The importance of increasing awareness is theoretically grounded as described by Weinstein and Sandman [
Within the KNW, participants were referred on average to 2.9 modules, while on average 2.1 modules were used. The appreciation rates were high, and the results showed that a higher number of modules used did not contribute to a higher appreciation. However, a higher perceived personal relevance did contribute to a higher appreciation. This is in line with Wilson et al [
Some limitations need to be addressed. First, providing data on completion of the separate themes and specific activities within the modules, and on completion of the evaluation sessions was not possible due to the module design. This information might be interesting for future studies; therefore, we recommend future interventions to study in more detail participation of intervention modules. Second, within our study, it was not possible to compare the relationships between the MRA and module use to a control group not receiving the MRA. Consequently, these associations need to be interpreted with caution, as it is conceivable that without the MRA, some of the same modules would have been used. Future experimental research might explore the specific effects of a similar automated referral system on subsequent choices. Third, this eHealth intervention requires respondents to have computer skills and health literacy, such as competence at accessing, understanding, appraising, and applying the health information provided [
The general KNW and the KNW modules were substantially used and highly appreciated by early cancer survivors, thus confirming the need for wide-ranging support among this target group. Results indicate that the MRA may be seen as a meaningful key component of the fully automated KNW intervention by guiding users to follow a preferred selection of modules, given their current complaints and identified needs. Moreover, the overall intervention and separate modules were highly appreciated, which could be explained by a higher perceived personal relevance. We can conclude that computer tailoring worked well and that the range of topics, design, and personalized information suited the needs of early cancer survivors. This process evaluation adds meaningful information on the use and appreciation of Web-based cancer aftercare interventions and confirms that the KNW offers valuable and appropriate support for early cancer survivors to complement usual cancer aftercare and may serve as a first step in a stepped-care approach.
Determination of the Module Referral Advice categories (red, orange, green).
Overview of cancer diagnoses among the KNW sample.
Predictors of a higher number of followed KNW modules (N=182).
Predictors of a higher appreciation of KNW (N=182).
Predictors of a higher perceived personal relevance of KNW content (N=182).
Body Mass Index
Cancer Survivors’ Unmet Needs questionnaire
Checklist Individual Strength
Hospital Anxiety and Depression Scale
Kanker Nazorg Wijzer (Cancer Aftercare Guide)
Mental Adjustment to Cancer Scale
Module Referral Advice
physical activity
quality of life
randomized controlled trial
Short Questionnaire to Assess Health Enhancing Physical Activity
Social Support List‒Discrepancy Subscale
Our sincere thanks go to the medical staff of the Dutch hospitals, who were involved in recruitment of participants: Beatrix Hospital Gorinchem, Bernhoven Hospital Uden, Bronovo Hospital The Hague, Catharina Hospital Eindhoven, Diakonessenhuis Hospital Utrecht, Elkerliek Hospital Helmond, Flevo Hospital Almere, IJsselland Hospital Capelle aan den Ijssel, Ikazia Hospital Rotterdam, Laurentius Hospital Roermond, Lievensberg Hospital Bergen op Zoom, Maasstad Hospital Rotterdam, Sint Anna Hospital Geldrop, Sint Jans Gasthuis Hospital Weert, Slingeland Hospital Doetinchem, Spaarne Hospital Hoofddorp, University Medical Center Groningen, University Medical Center Maastricht, VieCuri Medical Center Venlo, Langeland Hospital Zoetermeer, Zuyderland Medical Center Heerlen, and Sittard-Geleen.
Special thanks to Audrey Beaulen who provided helpful assistance in the recruitment of the subjects, and Linda Küsters for her contribution in the development of the lifestyle modules.
This research project was funded by the Dutch Cancer Society (grant number NOU2011-5151).
None declared.