Collaborative Care Intervention for Cancer Caregivers

The proposed randomized controlled trial will test the efficacy of a stepped collaborative care intervention, versus enhanced usual care arm, to reduce depression of cancer caregivers. Biobehavioral factors will include assessment of changes in health behaviors and biomarkers of inflammation. The investigators will also include measures metabolic abnormalities and clinical markers of CVD. The investigators expect few cardiac events during the study period. The investigators will also measure other health outcomes but CVD risk factors will be the focus of the study.

The investigators expect to follow the caregivers for 12 months during the caregiving period. Based on our prior experience and that of others, patients with clinical levels of symptoms report the greatest benefit from interventions aimed at reducing depressive symptoms. The investigators will screen caregivers for clinical levels of depression using the CES-D scale. If the caregiver screens positive for clinical levels of depressive symptoms (CES-D\>16) and the patients and caregiver consent to participation, the dyad will be randomized to the intervention or enhanced usual care arm based on block randomization and stratification by gender (male/female) to assure approximate equal gender distribution in each arm of the study. The screening will take place in UPMC Montefiore and Presbyian. The participants of this study will be blinded to which arm of the intervention they received, either the stepped collaborative care or enhanced usual care. Prior to randomization they will be asked to provide information on the intervention that they prefer and at the end of the intervention they will be asked what intervention they think they received. The findings will be analyzed to determine if intervention preference or perception of intervention received affected the outcomes. The care coordinators for the proposed intervention will have Master's degrees in psychology or counseling and training in cognitive-behavioral therapy and have experience working with people diagnosed with cancer and their families. The care coordinators will receive training at the University of Pittsburgh in cultural competence. The therapists will serve as a liaison between the patient and other health care professionals within the spousal and intimate partner caregivers' medical team. A psychiatrist will not be included as part of the study team as many of the spousal and intimate partner caregivers do not reside locally. The care coordinator will work with the spousal and intimate partner caregivers' PCP, psychiatrist, and other health care providers in the spousal and intimate partner caregivers' home community to manage the symptoms and medication if indicated. The care coordinators will be supervised by the PI who is a clinical psychologist who has been working with cancer patients and their caregivers for 20 years. Patient and caregiver outcomes will be assessed using a battery of standardized questionnaires. The investigators have chosen instruments that are as brief as possible to limit burden to patient and spousal and intimate partner caregivers. All instruments have been demonstrated to be valid and reliable. The order of the instruments in the battery of questionnaires will begin with general, not emotionally laden instruments and also end with instruments which may not increase distress in a participant. The proposed time points (baseline, 4, 8, and 12 months) reflect intervals that permit the ability to capture changes in symptoms and inflammatory biomarkers while not burdening the patient or spousal/intimate partner caregivers. The proposed questionnaires were used in prior studies by this team and with advanced cancer patients and their family caregivers. The investigators have shown better adherence to completing the outcomes measures with self-reported pencil and paper questionnaires versus telephone interviews (K07CA118576; R01CA176809). However, the investigators will also offer telephone interviews for spousal and intimate partner caregivers or patients who have low levels of literacy, sensory impairments, or prefer a telephone interview. For all interviews, the research associate will be blinded to the participant's assigned treatment arm to prevent bias. The investigators have found no differences by method of administration (paper and pencil versus telephone interview). Inter-operator variability will be addressed by ensuring that the research staff performing measurement of blood pressure, and waist circumference will be trained together, tested by the PI for inter-operator consistency, and be assessed at intervals every 4 months for continued consistency in measurement of blood pressure and abdominal girth. Consistency in laboratory personnel techniques (inter operator variability) will be monitored by Dr. Butterfield. The frequency distributions of study variables will be examined prior to statistical analyses to provide descriptive data and to identify departures from normality. If nonlinearity is detected, a transformation will be performed using the Box-Cox transformation, which finds the optimal relationship between variables. All measures have established reliability and validity similar to that from which the study sample will be drawn; however, the reliabilities of scales will be assessed using Cronbach's α. Data will be analyzed using SAS (Version 9.3, SAS Institute, Inc., Cary, NC) to conduct preliminary exploratory analyses, missing data testing, and the analyses for the primary aims (linear mixed models). The level of significance for the analyses will be set at .05. The efficacy of the web-based stepped collaborative care intervention compared to enhanced usual care will be examined using the 'intent to treat' (ITT) approach, as well as the actual treatment received. In the ITT approach, subjects will be included in the analyses in the groups to which they were randomly assigned, regardless of treatment, adherence, withdrawal or deviation from protocol. The ITT approach minimizes the introduction of bias into the analyses and most closely aligns with clinical practice; therefore, it is strongly recommended for studies examining intervention efficacy.