Opioids, Survival, and Advanced Cancer in the Hospice Setting

Objective: Although pain is common among advanced cancer patients, it can be controlled in a large proportion of patients. Several barriers hinder this, including the concern that opioids hasten death. We examined whether opioids influence survival among advanced cancer patients. Design: Retrospective observational study from September 2006 to October 2007. Setting: In-patient hospice unit. Participants: Participants were 114 consecutive hospice patients (mean age 71.7 +/- 13.9 years). Outcome Measurement: Analysis of survival (days) following admission, according to opioid usage. Standardized Oral Morphine Equivalents (OME mg/d) were calculated. Results: On admission 74.6% received opioids, rising to 92.1% at death. Mean opioid dosage was OME of 146 245 mg/d, and mean survival was 12.3 +/- 12.15 days. Mean survival, according to opioid dosage of 0, 1 to 119, and greater than or equal to 120 OME mg/d respectively at admission, was 16.7 +/- 13.4, 11.2 +/- 12.1, 10.0 +/- 10.2 (P =.009), and according to dose at death was 17.0 +/- 15.1, 12.3 +/- 12.1, 11.1 +/- 11.3 (P = ns). Increasing overall opioid dosage was associated with improved survival compared with no change or decreasing overall dosage (mean survival 14.0 +/- 12.7 days versus 9.3 +/- 9.8 versus 9.1 +/- 11.4, days respectively, P = .01). Adjusting for clinical variables in Cox proportional hazards models, no significant association was found between mortality and of the following aspects of opioid usage: (1) dose on admission (Hazard Ratio [HR] 1.009, 95% confidence interval [Cl] 0.999-1.019); (2) dose at death (HR 1.004, 95% CI 0.996-1.013); (3) mean dose (HR 1.006, 95% Cl 0.997-1.016); (4) overall dose increase (HR 0.733, 95% CI 0.417-1.288) and decrease (HR 0.967, 95% CI 0.472-1.984); (5) day-by-day dosage changes (HR 1.005, 95% CI 0.996-1.013). Conclusions: Opioid usage, even at high dosages, had no effect on survival among advanced cancer patients in a hospice setting. (J Am Med Dir Assoc 2011; 12: 129-134)