Impact of Diagnosis Nondisclosure on Quality of Dying in Cancer Patients: A Bereavement Study

Context. Patients should be optimally informed about their illness for patients' autonomy and shared decision-making. However, diagnosis nondisclosure to patients is traditionally widespread in Japanese culture with family-oriented autonomy. There is insufficient research on quality of death (QOD) and quality of care (QOC) among patients who are not told their diagnosis. Objectives. This study aimed to examine the impact of diagnosis nondisclosure on QOD and QOC in cancer patients. Methods. We performed a secondary analysis of a cross-sectional, nationwide mortality follow-back survey. The bereaved families responded to the questionnaire. Measurements included decedents' QOC, QOD, and bereaved families' outcomes. After using the propensity score matching method based on the covariates which can affect nondisclosure actions to compare the " disclosure " group and " nondisclosure " group, we compared differences in QOC, QOD, and bereaved families' outcomes between the two groups. Results. Of the 110,990 family members who were sent the questionnaires, we finally analyzed 46,672 responses. The disclosure group and nondisclosure group included 42,300 (90.6%) and 4,372 (9.4%) decedents, respectively. Most of the QOD domains (14/18) showed significantly higher scores in the disclosure group compared with the nondisclosure group. In terms of QOC domains, all domains showed higher scores in the disclosure group. Respondents in the disclosure group reported higher overall care satisfaction. Conclusion. We demonstrated that overall QOD and QOC in decedents with cancer were significantly higher in decedents with explicit cancer diagnoses. Furthermore, bereaved family members' outcomes were better among the family members of decedents with an explicit cancer diagnosis. J Pain Symptom Manage 2025;69:196-203. (c) 2024 American Academy of Hospice and Palliative Medicine. Published by Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.