Abstract
OBJECTIVES: The purpose was to analyze trends in intensity of care at End-of-life (EOL), in two cohorts of patients with lung or pancreatic cancer. SETTING: We used population-based registry data on health care utilization to describe proportions and intensity of care at EOL comparing the two cohorts (deceased in the years of 2010 and 2017 respectively) in the region of Stockholm, Sweden. PRIMARY AND SECONDARY OUTCOMES: Main outcomes were intensity of care during the last 30 days of life; systemic anticancer treatment (SACT), emergency department (ED) visits, length of stay (LOS) > 14 days, intensive care (ICU), death at acute care hospital and lack of referral to specialized palliative care (SPC) at home. The secondary outcomes were outpatient visits, place of death and hospitalizations, as well as radiotherapy and major surgery. A multivariable logistic regression analysis was used for associations. A moderation variable was added to assess for the effect of SPC at home between the cohorts. RESULTS: Intensity of care at EOL increased over time between the cohorts, especially use of SACT, increased with 10%, p<0.001, (n = 102/754 = 14% to n = 236/972 = 24%), ED visits with 7%, p<0.001, (n = 25/754 = 3% to n = 100/972 = 10%) and ICU care, 2%, p = 0.04, (n = 12/754 = 2% to n = 38/972 = 4%). High intensity of care at EOL were more likely among patients with lung cancer. The difference in use of SACT between the years, was moderated by SPC, with an increase of SACT, unstandardized coefficient β; 0.87, SE = 0.27, p = 0.001, as well as the difference between the years in death at acute care hospitals, that decreased (β = 0.69, SE = 0.26, p = 0.007). CONCLUSION: These findings underscore an increase of several aspects regarding intensity of care at EOL, and a need for further exploration of the optimal organization of EOL care. Our results indicate fragmentation of care and a need to better organize and coordinate care for vulnerable patients.