excluded. Participants were compensated for their time in the form
of gift cards redeemable at nationwide and regional merchants.
Data collection
Interviews were conducted, using a semi-structured interview
guide, by one of the study’s principle investigators, who is
experienced in qualitative data collection, the project coordinator,
and a research assistant who were both trained by the
study PI. in qualitative interview techniques. The initial interview
was the lengthiest at 45e60 min, followed by monthly interviews
ranging in length from 15 to 30 min. Post-death interview
lengths varied based caregiver wishes and ranged from 10
to 60 min. The caregivers were interviewed while their loved
ones were hospitalized, during a HF clinic visit, or via telephone
for post-death interviews. Caregivers were followed for up to two
years or one follow-up interview if the patient died prior to
study completion.
During the early interviews (1 month through 3 months),
caregivers were asked to describe the patient’s heart failure (what
they understood about it, its progression, how the patient was
doing) and where they saw the patient in one-year. In the parent
study, monthly interviews continued through the two-year study
enrollment period or until the patient died. However, the results
reported from this subset analysis are based only on initial interviews
to assess the caregivers’ understanding of HF severity
and terminality and post-death interviews to assess service
utilization.
Post-death interviews were completed with 34 of the 40 caregivers
whose loved ones died during study enrollment. Interviews
were conducted between one and three months following the
death of the patient. These interviews discussed palliative care,
end-of-life service utilization, and caregiver experiences of caring
for a patient with heart failure who died. Service utilization was
confirmed through medical chart review.
Data analysis
All the interviews were read and individually analyzed by the
study’s PIs and four PhD students. The PIs each analyzed 40 caregivers’
interviews and the PhD students each analyzed 20 caregivers’
interviews. An independent qualitative content analysis15
was done by each team member. During these individual analyses
team members were instructed to determine if they thought the
caregiver understood the severity of their loved one’s HF and if they
believed their loved one was dying based on the question Where do
you see [patient] in one-year? Quotes were required to justify
categorization. The individual analyses were followed by a group
comparative analysis, where each interview was discussed to reach
a consensus as to whether caregivers understood the severity and
terminality of HF based on their description of their loved one’s
heart failure.
Several strategies were employed to ensure the trustworthiness
of the qualitative data and subsequent findings throughout
the analysis process. Recorded interviews were transcribed by an
independent, professional transcriptionist and then cleaned and
checked for accuracy by the project director. To assure dependability,
an audit trail was documented during the team analysis
sessions. Any discrepancies in coding were reviewed by the
research team and discussed among the coders and investigators
until a consensus was met. To assure credibility, the final categorizations
and subsequent content justifications were agreed
upon by all members of the research team which included the
two PIs of the parent study, four PhD students, and the project
director