Edited by: Gian Mauro Manzoni, University of eCampus, Italy
Reviewed by: Eleonora Volpato, Fondazione Don Carlo Gnocchi Onlus (IRCCS), Italy; Tobias Kube, Harvard Medical School, United States
This article was submitted to Psychology for Clinical Settings, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Cancer patients experience various psychological and social difficulties, the most common being depression and anxiety. The purpose of this study was to develop and evaluate the effectiveness of an app-based cognitive behavioral therapy program for depression and anxiety in cancer patients. For this purpose, 63 participants who met the inclusion criteria were randomly assigned to either a mobile-application-based cognitive behavioral therapy program (HARUToday), a simple information-provision mobile-application-based program (HARUCard), or a waitlist control group. Self-report questionnaires including the Beck Depression Inventory, State-Trait Anxiety Inventory, Health-Related Quality of Life Scale, Dysfunctional Attitude Scale, and two computer tasks including the dot-probe task and the Implicit Association Test, were administered before and after 66 days of intervention. The results showed that the Beck Depression Inventory and State-Trait Anxiety Inventory scores of the cognitive behavioral therapy program (HARUToday) group decreased significantly after the intervention compared to the attention control (HARUCard) and waitlist control groups. However, there were no significant changes in scores of the Health-Related Quality of Life Scale and Dysfunctional Attitude Scale, and the two computer tasks. Such results suggest that a mobile-application-based cognitive behavioral therapy program may be an effective intervention for alleviating depression and anxiety, but not the general quality of life of cancer patients. Taking into consideration that psychosocial problems may not the topmost priority for cancer patients who are facing a chronic and possibly mortal disease, a mobile-application cognitive behavioral therapy program may be a possible solution for the alleviation of depression and anxiety in cancer patients who have many restraints in terms of time and space.
Cancer is one of the most common diseases worldwide and is the second leading cause of adult death (
Cancer is a physical disease that can threaten survival, and the lives of cancer patients are greatly influenced after cancer treatment, not only by the disease itself but also by the consequences and the aftereffects of the treatment (
The most frequently adopted psychosocial interventions for depression and anxiety in cancer patients involve cognitive behavioral therapy (CBT) (
A meta-analysis study has shown that CBT is effective for decreasing depression (
Recently, computer-based (e.g., CD-ROM, DVD, software, or internet-based) CBT programs have gained attention as a promising therapeutic alternative that can spread widely within a very short period. There have been numerous attempts to validate the effectiveness of various computer-based CBT programs because they are more easily accessible and efficient in terms of labor and cost than traditional face-to-face interventions are (
Mobile health is an updated version of computer-based programs that utilize mobile technologies such as smartphones, tablets, and wearable devices to provide interventions related to physical and mental health (
However, few studies have developed and explored the effects of mobile-app-based CBT programs for cancer patients. For example,
These studies suggest the usefulness of a mobile app-based CBT program as a treatment alternative for cancer patients, but more studies are needed to establish its effectiveness. Furthermore, most of these studies are limited in terms of demonstrating their effectiveness, since they included only self-report questionnaires as outcome measures. The need to include both subjective as well as objective outcome measures in a treatment outcome study was repeatedly stressed by several researchers (
The purpose of this study was to develop a mobile-app-based CBT program for reducing depression and anxiety in cancer patients and to test its effectiveness using both self-report questionnaires and computer tasks to sensitively detect possible changes.
The participants were recruited through two channels: referral from the oncologist in charge and advertisements in hospitals and public places. For those patients who were referred to the study by their oncologist in charge from the three major cancer centers in Korea, a research team member met the patients individually face-to-face at each center and obtained a written consent form after providing an explanation of the purpose of the study. Other patients were recruited via internet portal sites for cancer patients, bulletin boards of websites of cancer associations, bulletin boards of the three hospitals, and subway advertisements. In these cases, the cancer patients contacted the research team directly via phone to gather information about the study and signed the consent form when they visited the study site.
The steps for selecting the final participants of this study were as follows. In order to screen for participation in the study, the participants had to meet the following criteria: (1) aged 16–65 years, and (2) received a diagnosis of any type of cancer prior to the screening assessment. A total of 89 participants (11 males and 78 females) participated in the screening assessment. Upon screening, the participants had to meet the following criteria in order to participate in the study: (1) 16 points or more on the Beck Depression Inventory-Second Edition (BDI-II) score, and/or 39 points or more on the State-Trait Anxiety Inventory (STAI) for either state or trait anxiety, and (2) no medications (such as antidepressants). A total of 80 participants met the criteria mentioned above. Next, the 80 participants were assigned to three groups (the intervention group, attention control group, and waitlist control group). Of the 80 participants, 17 dropped out of the study due to fatigue, health deterioration, or death related to cancer treatment during the study period. Ultimately, there were 21 participants each in the intervention group, attention control group, and a waitlist control group (
Flowchart of study design.
Characteristics of participants.
3/18 | 3/18 | 3/18 | ||||
41.90 (11.30) | 43.52 (10.37) | 47.10 (11.19) | 1.23 | 2 | 0.299 | |
Age range (years) | 21–65 | 20–60 | 24–64 | |||
1.08 | 2 | 0.344 | ||||
Graduated college | 14 | 17 | 14 | |||
Graduated high school | 6 | 4 | 6 | |||
Graduated elementary school | 1 | 0 | 1 | |||
0.043 | 2 | 0.732 | ||||
Breast cancer | 9 | 10 | 12 | |||
Gynecologic cancer | 4 | 3 | 1 | |||
Thyroid cancer | 0 | 2 | 2 | |||
Sarcoma | 0 | 1 | 2 | |||
Other | 8 | 5 | 4 | |||
0.762 | 2 | 0.683 | ||||
Stage 1 | 5 | 7 | 5 | |||
Stage 2 | 7 | 5 | 8 | |||
Stage 3 | 5 | 5 | 3 | |||
Stage 4 | 4 | 4 | 5 | |||
4 | 5 | 5 | 1.786 | 2 | 0.410 | |
Surgery | 15 | 16 | 18 | 0.033 | 2 | 0.983 |
Radiotherapy | 10 | 9 | 12 | 1.35 | 2 | 0.514 |
Chemotherapy | 20 | 15 | 15 | 0.792 | 2 | 0.673 |
Other treatment | 8 | 4 | 5 | 0.559 | 2 | 0.756 |
To measure the level of depression, the BDI-II, which was developed by
To measure the level of anxiety, the STAI, which was developed by
To measure the health-related quality of life, the SF-36 survey, which was developed by
To measure the dysfunctional attitudes of the research participants, the DAS, developed by
In order to measure the level of satisfaction and to compile the necessary feedback on the program, the program satisfaction questionnaire used in the study by
In order to measure the attentional bias of the participants, a dot probe task used in a study by
For depression, two facial expressions conveying happiness and sadness were presented alongside neutral facial stimuli, based on research results showing that groups with low levels of depression tend to show attentional bias toward positive facial expressions (
In the dot probe task for depression, 15 “happy-neutral” pairs and 15 “sad-neutral” pairs of facial expressions conveying emotional states were used. For anxiety, 15 “threatening-neutral” pairs of pictorial stimuli conveying threatening or neutral situations were used. All stimuli were 5.5 cm × 3.7 cm in size, and each pictorial stimuli pair was presented side by side at a distance of 4.4 cm on a white background. The selection process of the stimuli used in this study was as follows.
For the dot probe task for depression, the facial expression stimuli of happy, sad, and neutral emotions were selected from the Yonsei Facial Stimuli Database (Chung et al., unpublished). Fifteen each of male and female facial stimuli with more than five points out of seven on the Likert scale (1 = “very weak,” 7 = “very strong”) in intensity rating were selected in descending order of intensity, and the “neutral” facial stimuli were the respective neutral expressions of the 15 selected males and females.
For the dot probe task for anxiety, pictorial stimuli were found using search terms (e.g., “threat” or “fear”) from an internet site providing pictures free of charge with appropriate citation.
This task was developed using JavaScript and was carried out using personal laptop computers. The stimuli were presented on a 13-inch screen on laptop computers set 65 cm away from the participants. Participants were instructed to respond to the stimuli using the laptop keyboard. All participants performed the task in a separate laboratory space blocked from external stimuli, and the experimenters supervised the participants’ responses from the right side of the participants.
The dot probe task consisted of three sets: (1) “happy-neutral,” (2) “sad-neutral,” and (3) “threat-neutral.” Four practice trials were carried out prior to each set in order to fully ensure that the participant had understood the task. Each set consisted of two blocks of 60 trials each, and the ratio in which the target stimuli appeared on the right and left and the ratio in which the dot appeared on the right and left were equal. There was a 1-min resting time between each block, and the entire task, consisting of three sets, amounted to a total duration of approximately 10 min.
The task started with the appearance of a fixation point (+) for 500 ms in the middle of the screen. After the fixation point disappeared, a 14 ms interstimulus interval (ISI) was given, followed by the appearance of the pair of pictorial stimuli on each side of the screen for 500 ms. After the disappearance of the pictorial stimuli pair, a dot (0.5 cm × 0.5 cm) appeared randomly on one side of the screen in which the pictorial stimuli had previously appeared, and at this time, each participant was instructed to quickly respond to the location of the dot using a keyboard key that was indicated by an alphabet sticker (left side = “L,” right side = “R”). In all the trials, the dot was presented until the participant responded, and the response initiated the next trial. In the practice trials, each trial was followed by feedback so that the participant could fully understand the procedure of the task, but in the experimental trials, the participants were not given any feedback on their responses. The practice trials were repeated until the participant was correct in at least three trials out of four, and if the practice trials were repeated for more than six times, it was assumed that the participant had not understood the task procedure and the experiment was terminated. The experimental procedure diagram of the dot probe task is shown in
Experimental procedure diagram of the dot probe task.
The dependent variable was the attentional bias score, calculated using the method proposed in a study by
In order to measure the implicit associations of the participants toward positive and negative word stimuli, an IAT used in a study by
In the present study, the categories “self” and “others” were presented as the target categories, and after selecting “positive” and “negative” adjectives as characteristic categories, the implicit attitudes between the two categories were measured through reaction times, based on a study showing that groups with a high level of depression and anxiety had a stronger level of implicit associations between “self” and “negative” words than between “self” and “positive” words (
In this study, four word stimuli belonging to the “self” and “others” categories and four word stimuli belonging to the “positive” and “negative” adjective categories were used. The process by which the word stimuli were selected is described below.
First, for the “self” and “others” categories, the word stimuli “I,” “mine,” “I am,” “my,” “others,” “for others,” “they are,” and “their” were used, as in the study by
As with the dot probe task, this test was developed using JavaScript and conducted using personal laptop computers. The stimuli were presented on a laptop screen in a size 26 font. An example screen of the IAT is shown in
Example screen of the IAT.
The IAT developed in this study consisted of three blocks (Blocks 1, 2, and 4) of exercise trials to distinguish the target categories and characteristic categories, and two blocks (Blocks 3 and 5) of experimental trials to measure the association between the target and the characteristic category, amounting to a total of five blocks. Participants were instructed to place a given word at the bottom of the screen quickly into the correct category out of the two that were presented at the upper left and upper right sides of the screen using the keyboard keys labeled with alphabet stickers (left = “L,” right = “R”). The task lasted for approximately 15 min.
First, in Block 1, the participant categorized positive and negative adjectives that were presented at the bottom of the screen into the categories “happy” or “sad,” presented in the upper left and upper right sides of the screen. Over a total of 10 exercise trials, four positive adjectives and four negative adjectives were presented once, and two words were randomly presented from each category.
In Block 2, participants were instructed to categorize the four words belonging to the “self” category and the four words belonging to the “others” category into the correct, corresponding category. As with Block 1, there were 10 exercise trials. In Block 3, the participants were instructed to categorize the presented word at the bottom of the screen into a pair of target and characteristic categories that were presented at the upper left and upper right sides of the screen. Over a total of 60 trials, the words belonging to each of the four categories were repeated three times, and three words from each category were randomly presented. Block 4 was composed of exercise trials with the “self” and “others” target categories presented in the opposite locations to those in Block 2. Block 5 was similar to Block 3, but the target category locations were changed, as in Block 4, in order to eliminate any bias between the left and right hands. As in Block 3, there were a total of 60 trials.
In order to rule out the order effect in Block 3 and Block 5, the IAT was designed to have two sets. In set A, the positive adjectives were first associated with the words in the “self” category, and in set B, the negative adjectives were first associated with the words in the “self” category. The two sets were counter-balanced according to recruitment order.
Each exercise block was repeated until the accuracy was at least 70%, and participants who repeated an exercise block more than six times were excluded from the analysis on the assumption that they did not understand the task. The IAT composition and trial numbers of each block can be seen in
The dependent variable of this task was the response time in the main trials (Block 3 and Block 5). Faster reaction time in the block in which the self and positive adjectives are paired indicates a strong association between the self and positive adjectives. The log-transformed values for the “self-positive adjective” reaction times and the “self-negative adjective” reaction times were used in the statistical analysis, as suggested in a study by
A randomized controlled trials design was used to determine whether the intervention was effective for the alleviation of depression and anxiety in cancer patients. All participants went through the following steps: screening, pre-intervention assessment, intervention in the intervention or attention control group or waiting in the waitlist control group, and post-intervention assessment. A simple randomization method was used to randomly assign each participant into the three groups (HARUToday, HARUCard, and waitlist control group). Each participant drew a card from a shuffled deck of three cards reflecting the three groups, and was immediately categorized into the drawn group. Participants were not told about which treatment they were receiving and what type of groups were being compared in the study. A detailed description of each step is shown in
Figure of overall research procedure.
After providing a brief explanation of the present study to prospective participants, those who agreed to participate in this study filled out the BDI-II (
Participants in all groups completed the same pre-intervention assessment package. This consisted of four questionnaires that measured the participants’ depression, anxiety, quality of life, and dysfunctional attitudes, along with two computer tasks. Trained research assistants administered both the questionnaires and computer tasks. All assessments were carried out in laboratory spaces within the present institution or in empty spaces within hospitals.
An app-based CBT program, HARUToday, which was developed by the authors for the purpose of this study, was provided to the participants in the intervention group. The HARUToday was named after the first letter of the four goals of this application, Habituation, Autonomy, Routinization, and Utilization. The last word, Today, was added with the hopes that the participants would use the CBT methods introduced in the application to live a better today. The participants installed the HARUToday
HARUToday is composed of five zones: (1) psycho-education, (2) behavioral activation, (3) relaxation training, (4) cognitive restructuring, and (5) problem-solving. The program was developed in the form of contents-based e-learning that minimizes text and takes advantage of visual and auditory examples, taking into account the age range and interests of the participants. The accuracy and appropriateness of the program were checked via professional consultation and feedback from prospective users. For example, the accuracy of the medical contents was checked by three oncologists. In addition, two focus-group interviews with cancer patients (
The HARUToday program is composed of 48 sessions, each of which takes approximately 10–15 min to complete. All sessions are composed of four phases: (1) Mood rating, (2) Lesson, (3) Summary, and (4) Quizzes. In the “Mood rating” phase, participants rate their mood from 0 points to 10 points. In the “Lesson” phase, the core skills of the day are introduced via case examples in which the main character experiences the most commonly reported depressive and/or anxiety symptoms in cancer patients and practices adequate skills to modify their thoughts and overcome emotional difficulties. A concise summary of the session is then provided in the “Summary” phase, followed by the “Quiz” phase to check whether the participants have become familiar with the session’s contents. Each session ends automatically when the two quiz questions are completed. During the course of the program, participants were prevented from changing the sequence order of the program and were unable to proceed to the next session without completing the previous session. On the “home screen,” participants could check their mood ratings, session progress, and score. They could also set an alarm for the time at which they wished to receive a pop-up notification to start the next session. A few screen samples are shown in
Example screen of the HARUToday program.
HARUCard,
The HARUCard program was designed to provide information and tips on managing depression and anxiety in a simple card format for participants. The information contained on the cards belonged to six categories: (1) information related to depression and anxiety, (2) exercise tips, (3) hobbies and travel, (4) movies and books, (5) famous quotes, and (6) artworks. Each card included an image corresponding to the content, and the references for the provided information or image were included at the end of each card.
The HARUCard program consists of 48 cards, and a single card was delivered to each participant in random order at a time set by a participant. At the set time, the participant received a pop-up notification that today’s card had arrived, and when the app was initiated, the card was viewable after completing a “mood rating.” The mood ratings were provided in a graph format on the home screen in the same way as in the HARUToday program. Once the cards were delivered, cards could be saved onto the participants’ personal phone or shared through SNS. Participants could easily access the wanted cards through the search function in the home screen. A few screen samples are shown in
Example screen of the HARUCard program.
A point-based reward system was also embedded in the HARUCard program. Score points were given for attendance (20 points/1 day), and bonus points were given if attendance was registered 5 days in a row (20 points). In order to improve the participants’ motivation, as in HARUToday, up to two SNS emoticons were sent when 700 points had been reached. Trained research assistants periodically monitored the performance of participants and provided feedback through telephone or text messages if the participant did not access the program for more than 5 days. In this study, four participants who received telephone or text prompts were dropped from the study, and none of the participants who were included in data analysis received telephone or text messaging feedback.
In the waitlist control group, after completing the pre-intervention assessment, the participants waited for the same amount of time (10 weeks), during which the intervention group and attention control group used the corresponding programs, and there was no further contact between the participants and researchers.
All participants from the three groups returned to the research lab or the hospital within 2 weeks of completion of the program and completed the satisfaction survey and post-intervention assessment. The participants in the waitlist control group were provided with either the HARUToday or HARUCard program upon request after the post-intervention assessment was completed. All participants who successfully completed the post-intervention assessment received a monetary reward as indicated on their consent form.
An
Statistical analysis was performed using IBM Statistical Package for Social Sciences Windows ver. 24.0. The dependent variables were change scores from pre-test to post-test for all outcome measures, which were normally distributed (Kolmogorov–Smirnov test > 0.05); therefore, the one-way ANOVA test was used.
The analysis method was as follows. First, the one-way ANOVA test was performed to evaluate whether the pre-intervention scores showed a difference between the three groups. Next, the one-way ANOVA test was performed in order to evaluate whether there were significant differences across the three groups before and after the intervention. Where the interactions were statistically significant, a modified Bonferroni post-test was conducted to determine which groups had significant differences. Furthermore, the formula below was used to calculate effect sizes (Cohen’s
In order to compare the program satisfaction between the intervention group and the attention control group, a Mann–Whitney
The one-way ANOVA test was conducted to compare the three groups on the pre-intervention scores. The results showed that there were no statistically significant differences between the groups on the pre-intervention assessment self-report questionnaires (BDI-II:
Comparison of self-report questionnaires by group in pre-intervention scores.
BDI-II | 25.24(9.83) | 24.33(10.97) | 27.48(10.17) | 0.51 | 0.600 |
State | 55.33(9.96) | 54.43(11.83) | 54.24(10.34) | 0.062 | 0.940 |
Trait | 54.24(7.95) | 55.10(9.78) | 55.24(9.05) | 0.077 | 0.926 |
SF-36 | 47.69(18.42) | 48.71(15.66) | 38.22(15.47) | 2.55 | 0.086 |
DAS | 169.24(24.62) | 162.81(23.41) | 152.00(32.22) | 2.40 | 0.091 |
Comparison of computer tasks by group in pre-intervention scores.
AB score for positive stimuli | 9.25(20.41) | 4.93(16.52) | 5.12(20.56) | 1.096 | 0.578 |
AB score for negative stimuli | 2.32(18.89) | −2.56(25.33) | −5.05(12.29) | 1.319 | 0.517 |
AB score for threatening stimuli | 4.07(24.47) | 0.43(14.65) | −0.13(23.88) | 0.661 | 0.718 |
Self-positive association | 6.84(0.22) | 6.77(0.23) | 6.77(0.25) | 0.954 | 0.621 |
Self-negative association | 7.08(0.26) | 7.04(0.22) | 6.92(0.22) | 4.577 | 0.101 |
The one-way ANOVA test was performed in order to test whether there were any significant differences across the groups before and after the intervention in levels of depression and anxiety, health-related quality of life, and dysfunctional attitudes. The between-group variables were calculated by subtracting the pre-intervention scores from the post-intervention scores. The means and standard deviations of each questionnaire for each group in both the pre- and post-intervention assessments are presented in
Comparison of pre- and post-intervention scores of each questionnaire by group.
BDI-II | 25.29(9.83) | 15.90(8.89) | 24.76(11.30) | 20.19(14.49) | 27.00(9.93) | 25.81(10.72) | 4.74 | 0.012* |
State | 55.57(9.96) | 43.29(9.12) | 54.17(11.80) | 47.33(11.40) | 54.24(10.34) | 54.62(9.32) | 10.44 | 0.001* |
Trait | 54.35(8.03) | 48.25(8.03) | 54.05(9.17) | 50.67(7.13) | 54.75(8.78) | 55.00(7.43) | 3.98 | 0.024 |
SF-36 | 47.69(18.42) | 54.63(19.80) | 48.71(15.66) | 56.87(23.32) | 38.22(15.47) | 45.02(16.05) | 2.098 | 0.132 |
DAS | 169.24(24.62) | 170.65(22.84) | 162.81(23.41) | 169.00(22.93) | 148.00(32.22) | 162.05(28.07) | 0.164 | 0.849 |
Significant differences were found across the groups in the BDI scores (
Significant differences were also found across groups in state anxiety (
No significant group differences in terms of SF-36 and DAS scores were found (SF-36:
One-way ANOVA tests were conducted to determine whether there were any significant differences across groups before and after the intervention in the attention-bias scores of positive, negative, and threatening stimuli, respectively. The between-group variables were calculated by subtracting the pre-intervention attentional bias scores from the post-intervention attentional bias scores.
No significant interaction between the groups regarding the positive, negative, or threatening stimuli in the attentional bias score were found [positive stimuli:
Means and standard deviations of AB scores in dot probe task.
AB score for positive stimuli | 9.28(20.41) | 1.19(13.49) | 4.93(16.52) | 0.35(17.34) | 5.12(20.56) | 2.40(14.87) | 0.921 | 0.631 |
AB score for negative stimuli | −5.05(12.29) | 1.42(12.62) | −2.56(25.33) | 1.40(18.94) | −5.05(12.29) | 0.00(12.91) | 1.426 | 0.490 |
AB score for threatening stimuli | 9.30(24.18) | −6.69(13.35) | 3.43(14.25) | −4.32(18.35) | 3.73(14.05) | −5.80(17.67) | 0.589 | 0.745 |
In order to evaluate whether there were any significant differences across groups before and after the intervention in implicit associations of positive and negative words, one-way ANOVA tests were performed. The between-group variables were calculated by subtracting the pre-intervention reaction time from the post-intervention reaction time and converting them to log values.
The results revealed no significant group differences in terms of the implicit associations of positive and negative words [self-positive:
Means and standard deviations of reaction times in IAT.
Self-positive association | 6.84 (0.22) | 6.66 (0.22) | 6.77 (0.23) | 6.67 (0.20) | 6.77 (0.25) | 6.73 (0.20) | 0.371 | 0.831 |
Self-negative association | 7.08 (0.26) | 6.82 (0.29) | 7.04 (0.22) | 6.94 (0.20) | 6.92 (0.22) | 6.95 (0.22) | 2.358 | 0.308 |
A Mann–Whitney
Results showed that the overall satisfaction of the intervention group (HARUToday) was significantly higher than that of the attention control group (HARUCard) (
Means and standard deviations of the composition-related satisfaction and each item of program satisfaction.
Composition-related satisfaction | 15.22 (2.48) | 15.45 (2.37) | 173.50 | −0.192 | 0.848 |
Overall satisfaction | 4.17 (0.62) | 3.55 (0.94) | 123.50 | −2.185* | 0.029 |
Likelihood of recommending | 4.17 (0.70) | 3.90 (1.25) | 177.50 | −0.380 | 0.704 |
Likelihood of participating again | 4.37 (0.75) | 4.35 (0.93) | 171.00 | −0.612 | 0.541 |
The goals of this study were to develop an app-based CBT intervention for cancer patients and to investigate its effects on depression and anxiety using self-report questionnaires and computer tasks. Eighty participants who met the inclusion criteria were randomly assigned to three groups (HARUToday group, HARUCard group, and waitlist control group), in which the participants trained or waited for 10 weeks (66 days), and 63 participants completed the program. The results showed a significant decrease in change scores from pre-intervention to post-intervention depression and anxiety scores in the HARUToday group, the CBT intervention group, compared to the HARUCard, the attention control group, and the waitlist control groups. On the other hand, there were no significant differences between the groups in terms of health-related quality of life, dysfunctional attitude, and computer tasks. The implications of this study are as follows.
First, the app-based CBT program was found to be effective in reducing depression and anxiety among cancer patients. This result is consistent with a previous finding demonstrating the effectiveness of app-based (
Second, the low dropout rate and high satisfaction found in this study indicate that the app-based CBT program was socially relevant and acceptable. One of the most challenging aspects of treatment outcomes studies is the high dropout rate (
In addition, the composition-related and program satisfaction of participants in this study were both above 80%, suggesting that the app-based CBT program is user-friendly and helpful. For example, participants scored high on “simplicity” in composition-related satisfaction and on “willing to re-participate” in program satisfaction. The lowest scores were found for “period of use” in composition-related satisfaction and “willing to recommend” in program satisfaction. These factors should be considered for future program development.
The low dropout and high satisfaction rates are especially encouraging considering that the cancer patients experience fatigue more easily than others do, especially during treatment; they also have behavioral limitations and may have very low motivation and energy levels to plan and maintain an activity (
Third, the app-based CBT program developed through this study is significant because the program is not limited to cancer type, stage of cancer, type of treatment, whether or not there has been metastasis or recurrence, or other medical variables, and has shown effectiveness in a wide range of cancer patients with various cancer types and stages. Most of the previous psychosocial intervention studies on cancer patients have been conducted on patients with a specific type of cancer or at a particular stage of a certain cancer (
Fourth, although the program had a positive effect on the depression and anxiety symptoms of the cancer patients, there was no significant change seen in the Dysfunctional Attitude Scale and Quality of Life Scale. Given that the basic assumptions of CBT are emotional changes through cognitive restructuring (
Fifth, the usefulness of computer tasks as objective measures was not clearly demonstrated, given the findings that no significant differences were observed in attentional bias and implicit attitudes pre- and post-intervention. These findings also raise a question about the mechanism of an app-based CBT program.
This study showed that an application-based CBT is effective for relieving depression and anxiety among cancer patients. The strengths of this study are as follows. First, this study is one of the very few studies which have applied CBT to the depression and anxiety of cancer patients, and is also one of the few studies to apply CBT using a mobile application platform. As mentioned previously, CBT is known to be effective, but is costly and time-consuming, which would make it harder for cancer patients to receive, given that they are most likely already having to cope with the medical costs and time-consuming cancer treatment. This study indicates that a mobile app-based CBT treatment specifically designed for cancer patients has an effect in reducing depression and anxiety levels compared to when they have not received CBT treatment. This is important in that mobile app-based CBT can lessen the constraints of space and time of traditional CBT, making CBT more available to cancer patients who naturally consider psychosocial problems as secondary problems to their cancer.
The limitations of this study and future research directions are as follows. The first and largest limitation of this study is the small sample size. As described in the data analysis section, an
The dataset are available from the first author upon request. Email KH at
This study was carried out under the approval of the Institutional Review Boards (IRBs) of Yonsei University, the National Cancer Center, and Ulsan University Hospital in South Korea. All subjects provided written informed consent.
K-MC designed the experiments, managed the experiments, and wrote the manuscript. KH designed the experiments, collected the data, analyzed the data, and wrote the manuscript. SC and YS collected the data, analyzed the data, designed and built the computer task, and wrote the manuscript. MR collected the data and wrote the manuscript. E-SY, HL, J-HK, SK, and S-JK recruited the participants, collected the data, and wrote the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
IAT composition and trial numbers of each block.
1 | Distinguishing the characteristic category (“happy” or “sad”) | 10 |
2 | Distinguishing the target category (“self” or “others”) | 10 |
3 | Characteristic category (“happy” or “sad”) + target category (“self” or “others”) | 60 |
4 | Distinguishing target category (opposite location: “others” or “self”) | 10 |
5 | Characteristic category (“happy” or “sad”) + target category (“others” or “self”) | 60 |
Total number of trials | 150 |
Contents of the sessions in the HARUToday program.
HARUToday | Psycho-education | Sessions 1–6 (total of six sessions) | Introducing depression and anxiety symptoms CBT program overview Familiarization with mood rating scale |
Behavioral activation | Sessions 7–13 (total of seven sessions) | Introduction to behavior activation techniques Learning how to write a behavior record Planning activities Checking and evaluating activities | |
Relaxation training | Sessions 14–24 (total of 11 sessions) | Introducing relaxation techniques through video and audio Introducing systematic desensitization techniques | |
Cognitive restructuring | Sessions 25–38 (total of 14 sessions) | Introducing the A-B-C model Familiarization with how to write an A-B-C record Fixing cognitive errors | |
Problem solving | Sessions 39–48 (total of 10 sessions) | Learning coping strategies |
This study was supported by a grant from the National R&D Program for Cancer Control, Ministry of Health and Welfare, South Korea (HA16C0021).
The source of the pictorial stimuli was referenced on the last page of each task.
The example screens of the IAT have been translated into English for publication. The actual experiments were conducted in Korean.
The Korean version of this program is available for use for research purposes. The introductory video can be found at the following address. To download the application, search “HARUToday Depression and Anxiety” in Korean on the Google Playstore or Apple App Store.
Contents have been translated into English for the purposes of this article, and images have been modified due to copyrights. Image:
The Korean version of this program is available for use for research purposes. The introductory video can be found at the following address. To download the application, search “HARUCard” in Korean on the Google Playstore or Apple App Store.
Contents have been translated into English for the purposes of this article, and images have been modified due to copyrights. Image: