Sampling difficulties in palliative care research

AIMS: To analyze the sample attrition in research with cancer patients at home. METHODS: Descriptive and transversal study. Multicentre study with purposive sampling. Inclusion criteria: palliative oncological, ECOG 2-3, older than 18 years, with family caregiver at home, informed consent. Variables: sociodemographic, sample attrition and other factors related to recruiting difficulties, participation and time delay between stages of research. Descriptive and inferential analysis. RESULTS: One hundred and thirty four patients met the inclusion criteria, 69.97 years old on average.50% men, 58.2% were married and 29.9% widowed. Only 57 patients completed the study. Fifty-three percent of patients were not included in the study. The sample attrition was 57.5%, 4.5% surveys were not completed.35.2% patients declined to participate (57.7% men) and 42.3% caregivers (80.3% women). The team excluded 22.5% patients for clinical reasons. The causes of nonparticipation were: 29.6% unspecified, 25.4% symptomatic exacerbation, 23.9% patient worsening, 8.5% died, 5.6% had no food problems, 4.2% lack of time, 1.4% did not sign consent and 1.4% patients did not want to involve their caregiver. All of the uncompleted surveys were due to symptomatic exacerbation. The time between recruitment and surveying on participation was not statistically significant between participants and nonparticipants, or between the completed and uncompleted surveys. CONCLUSIONS: The design of any investigation in palliative care should consider the high sample attrition in this type of research, factors that could be easily modifiable by the researcher to reduce it were not found