Hospice and pain management in nursing home residents with cancer

Abstract

Background: The prevalence of untreated pain in nursing home residents with cancer is unacceptably high. Hospice may increase the likelihood of receiving pain management at the end of life. Objectives: To estimate whether receipt of hospice in nursing homes increases the receipt of pain management for nursing home residents with cancer at the end of life. Methods: We conducted a cross-sectional study on a national sample of Medicare decedents who had cancer and were nursing home residents during the last 90 days of life in 2011–2012. We used the last Minimum Data Set (MDS) 3.0 assessment before death and the Medicare Beneficiary Summary File to measure hospice use, pain, and pain management at the last MDS assessment. We matched residents with cancer and in pain who received hospice care to residents in pain not receiving hospice care on nursing home facility and time from last MDS assessment to death. The primary outcomes were receipt of pharmacologic pain management including scheduled and PRN analgesics and non-pharmacologic pain management. Conditional logistic models were used to estimate the association between hospice use and pain management. Results: In matched analyses, untreated pain was uncommon (2.9% in hospice users and 5.6 in non-hospice users), though there was an absolute difference of 15.4% in scheduled analgesics use between hospice and non-hospice users (71.5% vs. 56.1%, respectively). Hospice use was associated with receipt of scheduled analgesics (adjusted Odds Ratio(aOR): 1.85, 95% Confidence Interval(CI):1.73–1.97), PRN medication (aOR: 1.31, 95% CI:1.20–1.43), and non-pharmacologic pain management (aOR: 1.18, 95% CI:1.11–1.26). Conclusions: Untreated pain at the end of life among nursing home residents with cancer was unusual. Hospice use was associated with increased pain management in nursing home residents with documented pain. Further work to examine the type and effectiveness of pain management strategies used is warranted.