PROGNOSTIC FACTORS IN PATIENTS IN THE TERMINAL PHASE OF HEMATOLOGICAL MALIGNANCIES WHO RECEIVED HOME MEDICAL CARE

ABSTRACT

Background: In the current global COVID-19 pandemic, terminal care in a patient's home has been expanded as a positive choice even for patients suffering from hematological malignancy (HM). Although there are many tools for predicting the prognosis of patients in the terminal phase of solid tumors, there is little information about prognostic factors in patients in the terminal phase of HM, especially patients receiving home medical care (HMC). In comparison to patients with solid tumors, those with HM are more likely to have acute diseases such as acute bleeding and acute infection leading to death. A previous report revealed that HM was a factor associated with aggressive end-of-life care. Because of the various complications associated with HM, it was reported to be difficult to predict the prognosis for patients with HM. Providing patients with accurate information about prognosis is important for them to consider how to spend their remaining life. Aims: In patients in the terminal phase of HM who received HMC, we aimed to validate the usefulness of two prognostic models Palliative Prognostic Index (PPI), which is an established prognostic model for patients in the terminal phase of a solid tumor, and the prognostic model reported by Kripp et al., which is a prognostic model for patients with HM in a palliative care unit. In addition, we aimed to determine prognostic factors for patients in the terminal phase of HM who received HMC and to develop a more detailed prognostic scoring system. Methods: We retrospectively evaluated 136 patients in the terminal phase of HM who were receiving HMC provided by 6 clinics between 2008 and 2022. Medical records relevant to prognosis were collected by a chart review. The effects of possible factors associated with overall survival (OS) were determined by the Kaplan-Meier method and univariate and multivariate Cox regression models. This study was approved by the IRB of Hokkaido University Hospital. Results: Patients characteristics were as follows male/female, 78/58;age, 25 to 94 years;median age, 79 years;AML, 50 patients;B-NHL, 32;MDS, 24;MM, 13;T-NHL 6;ALL, 5;ATL 2, CMML 2, and PV, 2. According to PPI, there was no significant difference in OS between the intermediate-risk group and the low-risk group (panel A;P = 0.15). By using the prognostic model reported by Kripp et al., we could stratify the patients into 3 risk groups with significantly different survival times (panel B;P < 0.01). However, there was a wide range of survival times in the high-risk group (OS, 0 to 125 days;median OS, 24 days). In our investigation of factors associated with OS, multivariate analyses revealed that there were 7 factors associated with poor OS (panel C). For the development of our prognostic scoring system, each variable was weighed according to the value of the hazard ratio (panel C) and 4 risk groups were shown to clearly discriminate survival (P < 0.01) low-risk group (n = 25, median OS of 434 days), intermediate-low risk group (n = 60, median OS of 112 days), intermediate-high risk group (n = 31, median OS of 31 days), and high-risk group (n = 20, median OS of 9 days). (Figure Presented ) Summary/Conclusion: This is the first investigation of prognostic factors that influence the overall survival of patients in the terminal phase of HM who received home medical care. In comparison to previously reported prognostic models, our scoring system could stratify patients in more detail. Providing patients and medical staff with accurate information about prognosis will lead to a higher quality of life in the terminal phase and better support by medical staff.