Longitudinal pain reports in a palliative care population

Background: The experiences of people with pain under palliative care is not well documented longitudinally. This work aims to describe the prevalence, intensity, progression, and associations of self-reported pain scores of people under a community palliative care service, by retrospectively examining how routinely collected pain scores changed over time. Methods: Pain was subcategorized depending on intensity, allowing group differences to be explored using analysis of variance. Bivariate analysis was used to explore the relationship between pain scores and other symptoms using Spearman's Rho correlation. Results: Over two years, 3027 patients were seen, with data collected an average of 17.3 times (52,387 data points). At referral, 50% (n=1402) were pain free. Of those with pain, mean score was 2 (range 0-10). The majority had cancer (90%), with no significant difference between the severity of reported pain scores between cancer and nonmalignant diagnoses (P=0.27). A significant gender difference was noted, with females statistically more likely to report severe pain (χ2 (3026)=5.61, p=0.018). Similarly, those <60 years were more likely to report pain χ2(3025)=3.07; p=0.022). Longitudinal changes in pain revealed the pain scores of people admitted with pain <7 always reported lower pain scores than those with severe pain on admission (90 days, F(1, 654)=55.72, p<0.001; 60 days, F(1, 1008)=48.62, p<0.001; 30 days, F(1, 1522)=60.36, p<0.001; 7 days, F(1, 1897)=15.4, p<0.001). However, pain scores of those with pain