Symptom intensity of hospice patients : A longitudinal analysis of concordance between patients' and nurses' outcomes

Nearing death, hospice patients are increasingly unable or unwilling to self-report their symptom intensity and rely on nurses' assessments. We hypothesized that concordance between patients' and nurses' assessments of symptom intensity improves over time. METHOD: A prospective longitudinal study was conducted from January 2012 to June 2015 using dyads of patients' and nurses' reported outcome measures, collected in daily hospice practice in the first three weeks after admission. Main outcomes were symptom intensity and well-being, measured using the Utrecht Symptom Diary (USD) and USD-Professional (USD-P). Absolute concordance was the proportion of dyads with no difference in scores between USD and USD-P per week after admission. For agreement beyond chance, the squared weighted Kappa for symptom intensity, and the one-way agreement intraclass correlation coefficient for well-being were used. RESULTS: The most prevalent symptoms, fatigue, dry mouth and anorexia, also had the highest intensity scores assessed by patients and nurses. Symptom intensity was underestimated more frequently than overestimated by the nurses. The absolute concordance was fair to good (35% - 69%). Agreement beyond chance was low to fair (0.146 - 0.539) and the ICC for well-being was low (0.25 - 0.28). Absolute concordance and agreement beyond chance did not improve over time. CONCLUSION: Concordance between patients' and nurses' assessment of symptom prevalence is good and both patients and nurses reveal identical symptoms as most and least prevalent and intense. However, nurses tend to underestimate symptom intensity. Concordance between patients and nurses symptom intensity scores is poor and does not improve over time