Title | Benefits of Early Palliative Care in the Community: A Propensity Score Matched Cancer Cohort |
Publication Type | Journal Article |
Year of Publication | 2018 |
Authors | Seow H, Barbera L, McGrail K, Burge F, Lawson B, Guthrie D, Chan K, Peacock S, Sutradhar R |
Journal | Journal of Pain and Symptom Management |
Volume | 56 |
Issue | 6 |
Pagination | e42 |
Date Published | Dec |
ISBN Number | 0885-3924 |
Accession Number | WOS:000451633700052 |
Abstract | ObjectiveRandomized trials of early palliative care (PC) showed patient benefits. However these benefits have not been validated in community-based cohorts. The objective is to assess the impact of early vs not-early PC among cancer decedents on the risk of receiving aggressive care (ED/hospitalization), supportive care (home care/physician home visit), or hospital death.MethodsWe took a retrospective cohort of Ontario decedents between 2004 and 2014. We identified those who were ‘early' PC users (i.e. used a validated PC service between month 12-6 before death [exposure]). We used propensity score matching to identify a control group of ‘not-early' PC users. We hard matched on age, sex, cancer type and stage. The propensity score included region, income, year, radiation, etc. Among those with InterRAI assessments in exposure period, we additionally controlled for health instability, and dependency, depression, cognitive performance, pain, and caregiver presence via propensity score. McNemar test used to examine differences between pairs.Results51,001 decedents received early PC vs 85,979 not-early PC (i.e. late or none). After matching among those with no interRAI assessments, we found 34,184 pairs of early and not-early PC users. Both groups had equal distributions of age, sex, cancer type (e.g. 25% lung cancer) and stage (e.g. 24% stage 3 or 4). Compared to not-early PC users, early PC users had a 10% lower absolute risk to die in hospital and have any aggressive care respectively and a 24% higher absolute risk to any receive supportive care. In a mutually exclusive cohort, matching among those with interRAI assessments, we identified 3,419 pairs of early and not-early PC users. Both groups had equal scores in health instability (e.g. 12% moderate-high CHESS score) and other interRAI scales.The outcome results were similar to the non-interRAI analysis.ConclusionUsing propensity score matching, decedents receiving early PC are likely to receive more supportive care and less aggressive care compared to not-early PC users. Our study uniquely focuses on a population receiving community palliative care. In a distinct sample using interRAI assessments, we control for several confounders previously unmeasured.Article info |
DOI | 10.1016/j.jpainsymman.2018.10.464 |
Short Title | J Pain Symptom Manag |
Alternate Journal | J Pain Symptom Manag |