Dubenske, L.L., Gustafson, D.H., Namkoong, K., Hawkins, R.P., Atwood, A.K., Brown, R.L., . . . Cleary, J.F. (2014). CHESS improves cancer caregivers' burden and mood: Results of an eHealth RCT. Health Psychology, 33, 1261–1272

To compare the effects of Comprehensive Health Enhancement Support System (CHESS) access intervention versus provision of internet with recommended websites for lung care cancer care on informal caregivers’ burden, disruptiveness, and mood

Patients and their caregivers were provided with laptops with internet that allowed access to CHESS. CHESS is a home-based e-health system that helps users to appraise the controllability of cancer-related stressors and improves their cognitive, behavioral, and supportive coping skills by providing them with a variety of information, communication, and coaching services appropriate for cancer caregivers based on their need and preference. CHESS allows sharing critical information with clinicians via Clinician Report, where caregivers are able to communicate their needs, ask questions, and rate patients’ symptoms on a scale of 0–10. A score of 7 or more triggers an email alert to the patient’s treating clinician so he or she is aware of the patient’s and caregiver’s situation on the next patient visit. Patients randomly were assigned to internet use only or CHESS plus internet use. Internet users were provided with internet access and a list of high-quality websites for information.

• N = 246 (122 in the control [internet] group and 124 in the intervention [CHESS] group)
• AGE RANGE = 18–84 years
• MEAN AGE = 55.56 years
• MALES: 31.7%, FEMALES: 68.3%
• KEY DISEASE CHARACTERISTICS: Patients with advanced non-small lung cancer IIIA, IIIB, and IV with a life expectancy no less than four months
• OTHER KEY SAMPLE CHARACTERISTICS: English-speaking adults who are primary caregivers; participants were primarily females (68.3 %), spouses/partners (72%), and highly-educated (53% had an associate's degree); most patients had advanced disease (IV) (66.4%) and other major diseases (57.7%)

• SITE: Multi-site    
• SETTING TYPE: Outpatient    
• LOCATION: Northeastern, Midwestern, and Southwestern United States

• PHASE OF CARE: Multiple phases of care
• APPLICATIONS: Pediatrics, elder care, palliative care 

• Randomized clinical trial with a randomization ratio of 1:1
  • Participants were stratified by study site, caregiver’s race, and relationship to patient.
  • Measurements were administered at pretest and post-tests every two months up to two years, or 13 months after a patient’s death, whichever came first.
  • Analysis involved data from the first eight months only.

• Caregivers’ demographics
• CHESS use (number of logins, number of pages viewed, and time spent on the CHESS system)
• Disruptiveness was measured by the Caregiver Quality of Life-Cancer Scale (CQOLC) Disruptiveness Subscale.
• Burden was measured by the CQOLC Burden Subscale.
• Negative mood was measured by a 16-item subset of the Short Version Profile of Mood States (SV-POMS) focusing on measurement of Tension-Anxiety, Anger-Hostility, and Depression-Dejection.
• Patient symptom distress was measured by the caregiver’s perception of the patient’s symptom distress using the Edmonton Symptom Assessment Scale (ESAS).

Average use of CHESS among the intervention group was 14.6 logins, averaging browsing 293 pages and spending 177 minutes on the system (almost 50 minutes monthly). Statistically significant improvement was seen in the intervention group over the control group at six months follow-up in burden (p = .02) and negative mood (p = .006), but not in disruptiveness. These effects persisted even when patient symptom distress (on the ESAS) and other covariates were controlled. Analyses also show that these significant differences existed only at six-month follow-up, not at any other times. 

Targeted electronic systems that are designed comprehensively to help caregivers can improve their experience of caregiving by minimizing their perceived burden and improving their mood.

• Risk of bias (no control group)
• Risk of bias (no blinding)
• Risk of bias (no appropriate attentional control condition)
• Intervention expensive, impractical, or training needs
• Subject withdrawals/attrition 10% or greater
• Physician consent was needed first before a candidate could be admitted into the study because the intervention required a clinician’s responsiveness to patients’ and caregivers’ needs. 
• Although training was given to participants who needed training on accessing CHESS and the internet, the study does limit participation of those who may not be computer savvy or would prefer receiving support in means other than electronic.

Targeted e-health systems are more likely to be beneficial to caregivers of patients with cancer when they are self-directed and have a comprehensive approach to caregivers’ needs (e.g., information, communication, coaching, interaction with healthcare providers). The benefits of e-health vary among cancer caregivers and depend on how they use the tools offered by such systems. Which feature or combination of features would be most helpful to this group of caregivers remains to be determined.