Effects of an Internet Support System to Assist Cancer Patients in Reducing Symptom Distress

A Randomized Controlled Trial

Cornelia M. Ruland, PhD, RN; Trine Andersen, MSc, RN; Annette Jeneson, MSc; Shirley Moore, PhD, RN; Gro H. Grimsbø, MSc, RN; Elin Børøsund, MSc, RN; Misoo C. Ellison, PhD


Cancer Nurs. 2013;36(1):6-17. 

In This Article



WebChoice (www.communicaretools.org) was developed with extensive end-user participation and is described in more detail elsewhere.[28] The content of WebChoice is based on a comprehensive review of the breast and prostate cancer–related scientific literature, and the information content is regularly updated. WebChoice contains the following:

  1. An assessment component that allows patients to monitor and report their symptoms, problems, and priorities for support along physical, functional, and psychosocial dimensions, currently and over time. Patients can use this information in many useful ways, for example, to monitor improvement/deterioration of their condition to know when to alert healthcare providers; to prepare for a hospital/doctor consultation to improve patient-provider communication and to prepare for discussions about treatment/care; or to obtain immediate access to tailored self-management support as described below.

  2. Tailored symptom self-management support. Patients' self-reported symptoms trigger the display of the appropriate subset of self-management activities from which patients can choose. Each message contains an explanation of what the activity is, how to perform it, potential risks, adverse effects, contraindications, when to contact a physician, level of evidence, references to the source of information from where the evidence was obtained, and links to other relevant, reliable Web sites that contain related information. This information can be printed out for further reading or can be used to create an individualized self-management plan that includes selected activities tailored to the patient's individual symptoms and problems.

  3. An information section where patients have access to other reliable, relevant Web resources, such as information about specific tests, treatments and potential adverse effects, lifestyle suggestions, information about patients' rights, and links to support groups. All information complies with the HON Code of Conduct for medical and health Web sites that promote the highest principles for privacy, security, credibility, and reliability of information on Internet health sites.[36]

  4. A communication section where patients can share their experiences with other patients and obtain professional support. It includes (a) an unrestricted support forum for group discussion, allowing users to post messages anonymously, and (b) a question-and-answer area where patients, in private, can ask questions to expert nurses in cancer care. In our study, the expert nurses accessed the communication section daily on weekdays and participated in the group discussions when appropriate. The communication component gives patients an opportunity to ask difficult questions anonymously and to learn how others deal with similar problems—in their own homes and with confidentiality.

  5. In addition, patients have access to a diary, where they can keep personal notes.

All pages in WebChoice are based on language-independent templates that are expanded at run time by separating the content from the system into separate files. The system is generic; content adjustments, for example, to fit it to patients with different diagnoses, can be dynamically performed without making any changes to the system. Strong security measures are implemented to protect patients' submitted information. Patients are authenticated using a smart card–based public key (PKI) solution. All data are submitted to a secure server through an encrypted connection. All procedures comply with the Norwegian Personal Data Act (equivalent of Health Insurance Portability and Accountability Act in the United States). More information about WebChoice is available at www.communicaretools.org.


Inclusion criteria were as follows: being diagnosed with, and undergoing treatment for, breast cancer (surgery plus additional treatment of either radiation, chemotherapy, hormone therapy, or a combination of those), or being diagnosed with, and undergoing treatment for, prostate cancer; older than 18 years; able to speak Norwegian; Internet access at home; and no radiation on the brain as this may influence a patient's ability to reliably report symptoms and to fill out questionnaires.

Sample Size Calculation

The estimated sample size was based on an expected effect size of 0.2 in the difference of the mean changes in the primary outcome of symptom distress between the intervention and control group from baseline through the 1-year study period, divided by its SD. Because no previous data were available from which we could calculate an expected effect size, we chose a small effect size because we expected large SDs due to large variations in stage of disease among study participants. Based on a repeated-measures analysis of variance via a mixed-model approach, we needed to enroll 113 patients per group to achieve 85% power at a 2-sided 5% significance level with 5 repeated measurements.[37,38] To account for an up to 30% dropout rate, estimated from a similar 1-year follow up study,39 we needed to enroll 161 patients per group.

Study Procedures

The study received approval from the Regional Ethical Review Committee (institutional review board) of the Health Region South and the Data Security Inspectorate in Norway. Written informed consent was obtained from all study participants. After pilot testing the intervention and procedures, patients were recruited into this study from May 2006 through July 2007 and followed up for 1 year.

Invitations to participate in this study were disseminated through advertisements in national newspapers, in weekly magazines, on the Norwegian Cancer Society's Web site, as well as through information pamphlets mailed to patients from the Norwegian National Cancer Registry. The advertisements contained information about the purpose of the study, what it involved, study duration, inclusion criteria, the name of the principal investigator, the responsible institution, and a contact number to call. Those who were interested in participating or wanted more information were invited to call. When calling, they were screened for inclusion criteria by a trained research assistant using a screening script. Callers who did not meet the inclusion criteria were thanked for their time and interest. Those who met the inclusion criteria were asked for their mailing address and sent a letter containing information about the study on a consent form and baseline questionnaires, along with instructions for completion and 2 self-addressed, stamped return envelopes. Participants were asked to return the consent form and baseline questionnaires in 2 separate envelopes.

Upon return of participants' signed consent forms and completed baseline questionnaires, they were randomized into either the experimental or control group, using a computerized minimization algorithm to balance covariates.[40] This algorithm simultaneously capitalizes on the benefits of random assignment and, at the same time, equalizes the proportion of cases on covariates, so that differences between treatment groups on covariates do not emerge based on the "randomness" of random assignment. In this study, participants were stratified on cancer diagnosis and stage of disease (primary, recurrence, metastases).

Participants who were included were informed about their group assignment by receiving a letter that also included information about follow-up study procedures. All participants received questionnaires by mail at 3, 6, 9, and 12 months of the study. Participants were not paid for study participation, but received at the 6-month data collection point a scratch-off lottery ticket worth Norwegian Kroner 25 (approximately US $4) as a token of appreciation for their continued participation.

Experimental Group Protocol. Because of the nature of the study, intervention participants were not blinded. Participants who were randomly assigned to the experimental group received a letter informing them of their group assignment, a user manual with instructions to use WebChoice, and a contact address and phone number for technical support. Participants were told that they could use WebChoice and any of its components as they liked during their 1-year study participation and that the use of the system was entirely voluntary. Participants were instructed not to identify themselves by their real names when using the communication components in WebChoice.

Control Group Protocol. In addition to the letter informing them of their group assignment, participants who were assigned to the control group received an information sheet with suggestions for publicly available, cancer-relevant Internet sites that could be useful to them.


The primary study outcome was symptom distress. Secondary outcomes were depression, self-efficacy, HRQoL, and social support. Symptom distress was measured 5 times: at baseline and at 3, 6, 9, and 12 months. Secondary outcomes were measured only at baseline and at 3, 6, and 12 months because of institutional review board's concerns about response burden.

Symptom Distress. Because of the subjective nature of the experience of symptoms, self-reports are today considered the criterion standard for symptom assessment. Consistent with this approach, symptom distress was measured using the Memorial Symptom Assessment Scale–Short Form (MSAS-SF).[41] A physical symptom subscale, a psychological symptom subscale, and a global distress index (GDI) that is considered to be a measure of overall symptom distress can be derived from the MSAS-SF.[42] Respondents are asked to indicate how bothersome each of the symptoms are, on a 5-point rating scale from "not at all" (0) to "very much" (4). Higher scores indicate greater symptom distress. Cronbach α as a measure of reliability for our sample at baseline was .91.

Depression. Depression was measured using the Center for Epidemiological Studies–Depression Scale,[43] a 20-item instrument where respondents answer brief questions by responding to 5-point rating scale statements ranging from "rarely or none of the time" to "most or all of the time." Higher scores indicate greater depression. Cronbach α for our sample at baseline was .74.

Self-efficacy. Self-efficacy was measured using the Cancer Behavior Inventory version 2.0,[44] a 33-item scale measuring self-efficacy for coping with cancer-related stress that comprises 7 dimensions: maintenance of activity and independence, seeking and understanding medical information, stress management, coping with treatment-related adverse effects, accepting cancer and maintaining a positive attitude, affective regulation, and seeking support. Respondents answer to 9-point rating scale questions from "not at all confident" to "totally confident." Higher scores indicate greater self-efficacy. Cronbach α for our sample at baseline was .96.

Health-related Quality of Life. Health-related quality of life was measured using the 15D HRQoL instrument, a generic, comprehensive, 15-dimensional, self-administered instrument measuring HRQoL among adults. It combines the advantages of a profile and a preference-based, single-index measure. Respondents respond to 5 ordinal levels on each dimension, by which more or less of the attribute is distinguished. The respondent chooses from each dimension the level that best describes his/her present health status. Higher scores indicate greater HRQoL.[45] Cronbach α for our sample at baseline was .70.

Social Support. Social support was measured using the Medical Outcomes Study Social Support Survey, a 20-item instrument with 2 subscales addressing instrumental and emotional support.[46] Respondents respond on a 5-point rating scale ranging from "none of the time" to "all of the time." Higher scores indicate more social support. Cronbach α for our sample at baseline was .81.

Statistical Analyses

Analyses were performed using intention to treat, regardless of whether intervention group participants had logged on to WebChoice or not. A linear mixed-effects model methodology,[37,38,47] which accounts for both the correlation between the repeated measurements across times within each subject and the variability between the subjects, was applied to compare groups on slopes over time on primary and secondary outcomes. Assuming that missing data were randomly distributed and because sufficient data were available, all included participants were kept in the final analyses. Because "time since diagnosis" showed close-to-significant group differences (P = .07) at baseline (Table 1), and there were large variations among patients, we included this variable as a covariate in the analyses and controlled it statistically. We performed all statistical analyses using SAS release 9.1.3 (SAS Institute, Inc, Cary, North Carolina).