This study revealed that medical backgrounds of cancer patients receiving home care in Japan showed common clinical features. For example, 322 (93%) of the patients had solid cancers, and the majority of the cancer types were lung (n = 80, 23%) and gastric cancer (n = 47, 14%). Our study reflects the situation in Japan, in a similar way as previous studies conducted in other countries . Lung cancer is the primary cause of cancer death, and gastric cancer is in the second place in Japan . Meanwhile, patients with hematologic malignancies were rarely treated at home. The occurrence of sudden changes due to complications and need for adjuvant therapy such as blood transfusion and palliative chemotherapy were probably the obstacles for these patients .
Most patients were elderly, and had several complications. For example, 232 (67%) of the patients were 71 years old or older, 261 (75%) were PS 3-4, and 121 (34%) were complicated with dementia. The subjects in this study were older than those in previous studies reported from other countries [11, 12]. These findings suggest that even severely ill patients with poor general conditions can be treated at home by providing adequate care.
The therapeutic interventions conducted during home care were mainly symptomatic treatments. In fact, all the patients received supportive medications. Pain control is a major aim in patients with advanced cancer . In this study, 81 patients (23%) complained intense pains. Among them, 74 were given narcotic medications without serious complications. These findings were comparable to previous studies . Since most physicians who provide home care are familiar with palliative medicine, palliative treatment using narcotics can be provided safely at home.
The majority of cancer patients who received home care was financially independent, and did not receive any public financial assistance. Such financial situations were similar to those of previous studies [14, 15]. Home care might be an option limited to patients with sufficient financial resources since it can increase the economic burden on the patients. Meanwhile, 336 (97%) received support from their families or from the staff of the facility. Even patients who had lived alone and died at home received nursing care by their families during their last moments, suggesting that family or social supports are requisite for cancer patients to live their last moments at their own homes.
This study showed that use of home-visit nursing care services, will of the patients, their gender and age played an important role in the continuation of home care. It is reasonable to speculate that the use of home-visit nursing care services contributes to the safe in home care since medical providers visit more frequently and patients are closely monitored. Women were found as a significant factor for inhibiting the continuitation of home care in this study while a previous study from Taiwan showed that being female was an important determinant of hospice home care . In a meta-analysis using studies mostly from the UK, the US, Australia, and Canada, gender factor did not have significant effect in dying at home . Thus, differences of culture among countries might influence the gender disparity. In Japan, elderly care has been traditionally considered as 'women's work'  and national survey reported that 69.4% of caregivers were women in 2010 . An explanation of the gender disparity might be that Japanese women miss the opportunity to receive medical home care in the background of traditional thinking of gender role. Surprisingly, older age was associated with longer continuation of home care. The exact reason remains unknown; however, it is probable that cancer might have been diagnosed by chance at an early stage in elderly patients who had been treated at home for other chronic diseases.
This study provided valuable information on home care for cancer patients; however, it still involves several problematic issues. The most considerable problem is that this was a small-scale retrospective study. For that reason, it might be accompanied by a bias of which researchers were unaware. For example, because this study was conducted with physicians with a rich experience in home care, caution is required when generalizing these results. To overcome these problems, validation by large-scale prospective studies should be performed in the future.