A Population-Based Mortality Follow-Back Survey Evaluating Good Death for Cancer and Noncancer Patients: A Randomized...

Evaluation of end-of-life care is a key element in quality improvement, and population-based mortality follow-back designs have been used in several countries. This design was adapted to evaluate a good death in Japan.

This study aimed to explain the scientific background and rationale for assessing the feasibility of a mortality follow-back survey using a randomized design.

We used a cross-sectional questionnaire survey to assess feasibility using response rate, sample representativeness, effect on response rate with two methods, and survey acceptability.

The subjects were 4812 bereaved family members of patients who died from the major five causes of death: cancer, heart disease, cerebrovascular disease, pneumonia, or kidney failure, using mortality data.

Overall, 682 (14.2%) questionnaires could not be delivered, and 2294 (55.5%) family members agreed to participate in the survey. There was little difference in the distribution of characteristics between the study subjects and the full population, and sample representativeness was acceptable. Sending the questionnaire with a pen achieved a higher response rate than without (weighted: 48.2% vs. 40.8%; P < 0.001). In follow-up contact, there was no difference in response rate between resending the questionnaire and a reminder letter alone (weighted: 32.9% vs. 32.4%; P = 0.803). In total, 84.8% (weighted) of the participants agreed with improving quality of care through this kind of survey.

This study demonstrated the feasibility of conducting a population-based mortality follow-back survey using a randomized design. An attached pen with the questionnaire was effective in improving the response rate.