Sakiko Fukui1, Tatsuya Morita2, Kazuhiro Yoshiuchi3. 1. Department of Community Health Nursing, Graduate School of Nursing, The Japanese Red Cross University, 4-1-3 Hiroo, Shibuya-ku, Tokyo, 150-0012, Japan. sfukui@redcross.ac.jp. 2. Seirei Hospice, Seirei Mikatabara Hospital, Hamamatsu, Shizuoka, Japan. 3. Department of Stress Sciences and Psychosomatic Medicine, Graduate School of Medicine, The University of Tokyo, Bunkyo-ku, Tokyo, Japan.
Abstract
PURPOSE: The aim of this study was to investigate the predictive value of a clinical tool to predict whether discharged cancer patients die at home, comparing groups of case who died at home and control who died in hospitals or other facilities. METHOD: We conducted a nationwide case-control study to identify the determinants of home death for a discharged cancer patient. We randomly selected nurses in charge of 2000 home-visit nursing agencies from all 5813 agencies in Japan by referring to the nationwide databases in January 2013. The nurses were asked to report variables of their patients' place of death, patients' and caregivers' clinical statuses, and their preferences for home death. We used logistic regression analysis and developed a clinical tool to accurately predict it and investigated their predictive values. RESULTS: We identified 466 case and 478 control patients. Five predictive variables of home death were obtained: patients' and caregivers' preferences for home death [OR (95% CI) 2.66 (1.99-3.55)], availability of visiting physicians [2.13 (1.67-2.70)], 24-h contact between physicians and nurses [1.68 (1.30-2.18)], caregivers' experiences of deathwatch at home [1.41 (1.13-1.75)], and patients' insights as to their own prognosis [1.23 (1.02-1.50)]. We calculated the scores predicting home death for each variable (range 6-28). When using a cutoff point of 16, home death was predicted with a sensitivity of 0.72 and a specificity of 0.81 with the Harrell's c-statistic of 0.84. CONCLUSION: This simple clinical tool for healthcare professionals can help predict whether a discharged patient is likely to die at home.
PURPOSE: The aim of this study was to investigate the predictive value of a clinical tool to predict whether discharged cancerpatients die at home, comparing groups of case who died at home and control who died in hospitals or other facilities. METHOD: We conducted a nationwide case-control study to identify the determinants of home death for a discharged cancerpatient. We randomly selected nurses in charge of 2000 home-visit nursing agencies from all 5813 agencies in Japan by referring to the nationwide databases in January 2013. The nurses were asked to report variables of their patients' place of death, patients' and caregivers' clinical statuses, and their preferences for home death. We used logistic regression analysis and developed a clinical tool to accurately predict it and investigated their predictive values. RESULTS: We identified 466 case and 478 control patients. Five predictive variables of home death were obtained: patients' and caregivers' preferences for home death [OR (95% CI) 2.66 (1.99-3.55)], availability of visiting physicians [2.13 (1.67-2.70)], 24-h contact between physicians and nurses [1.68 (1.30-2.18)], caregivers' experiences of deathwatch at home [1.41 (1.13-1.75)], and patients' insights as to their own prognosis [1.23 (1.02-1.50)]. We calculated the scores predicting home death for each variable (range 6-28). When using a cutoff point of 16, home death was predicted with a sensitivity of 0.72 and a specificity of 0.81 with the Harrell's c-statistic of 0.84. CONCLUSION: This simple clinical tool for healthcare professionals can help predict whether a discharged patient is likely to die at home.
Entities:
Keywords:
Death; Neoplasms; Patient discharge; Place; Predictive to
Authors: Alberto Alonso-Babarro; Eduardo Bruera; María Varela-Cerdeira; María Jesús Boya-Cristia; Rosario Madero; Isabel Torres-Vigil; Javier De Castro; Manuel González-Barón Journal: J Clin Oncol Date: 2011-02-22 Impact factor: 44.544
Authors: Bridget Gwilliam; Vaughan Keeley; Chris Todd; Matthew Gittins; Chris Roberts; Laura Kelly; Stephen Barclay; Patrick C Stone Journal: BMJ Date: 2011-08-25